HowSocialMedia Changes User-Centred Design - MAFIADOC.COM

D e p a r t me nto fC o mp ut e rS c i e nc ea ndE ng i ne e r i ng

Mikae lJ o h nso n

H o wSo c ialMe dia Ch ange s U se rCe nt re dD e sign

H o wSo c ialMe dia Ch ange sU se rCe nt re dD e sign

C umul a t i v ea ndS t r a t e g i cU s e rI nv o l v e me ntw i t h R e s p e c tt oD e v e l o p e r – U s e rS o c i a lD i s t a nc e

M i k a e lJ o h ns o n

CURRENT USER INVOLVEMENT

☕ Informal Meeting  Community 📈📈 Web Analytics PROJECT DIMENSIONS

👤👤 Interview TAP 👥👥 Focus Group TO REFINE 👍👍 Poll YOUR STRATEGY ▶ DevÐUser Social Distance

SWIPE Developer–User Social Distance

TIMELINES

◀

FOR OTHER DIMENSIONS

▶ User Involvement

👍👍 📈📈 📈📈 👍👍 📈📈 👍👍 📈📈 👍👍 📈📈 DRAG TO MOVE ☕   ☕ 👥👥  👥👥FORWARD

▶

                      A a l t oU ni v e r s i t y

D O C T O R A L D I S S E R T A T I O N S

Aalto University publication series DOCTORAL DISSERTATIONS 46/2013

How Social Media Changes User-Centred Design Cumulative and Strategic User Involvement with Respect to Developer–User Social Distance Mikael Johnson

A doctoral dissertation completed for the degree of Doctor of Science (Technology) to be defended, with the permission of the Aalto University School of Science, at a public examination held at the lecture hall AS1 of the school on 15 March 2013 at 12.

Aalto University School of Science Department of Computer Science and Engineering

Supervising professor Prof. Marko Nieminen Thesis advisors Prof. Marko Turpeinen Prof. Sampsa Hyysalo Preliminary examiners Prof. Robin Williams, University of Edinburgh, UK Prof. Dag Svanæs, Norges teknisk-naturvitenskapelige universitet, Norway Opponent Prof. Jan Gulliksen, Kungliga Tekniska högskolan, Sweden

Aalto University publication series DOCTORAL DISSERTATIONS 46/2013 © Mikael Johnson ISBN 978-952-60-5072-0 (printed) ISBN 978-952-60-5073-7 (pdf) ISSN-L 1799-4934 ISSN 1799-4934 (printed) ISSN 1799-4942 (pdf) http://urn.fi/URN:ISBN:978-952-60-5073-7 Images: Cover design by Mikael Johnson, who used free and opensource image elements: Entypo pictograms by Daniel Bruce, Signika font by Anna Giedryś, and Lorenz attractor image by User:Wikimol / Wikimedia Commons. Habbo images used with permission. Unigrafia Oy Helsinki 2013 Finland

Abstract Aalto University, P.O. Box 11000, FI-00076 Aalto www.aalto.fi

Author Mikael Johnson Name of the doctoral dissertation How Social Media Changes User-Centred Design: Cumulative and Strategic User Involvement with Respect to Developer–User Social Distance Publisher School of Science Unit Department of Computer Science and Engineering Series Aalto University publication series DOCTORAL DISSERTATIONS 46/2013 Field of research T011Z Usability Research Manuscript submitted 17 April 2012

Date of the defence 15 March 2013 Permission to publish granted (date) 20 December 2012 Language English Monograph

Article dissertation (summary + original articles)

Abstract The aim of user-centred, participatory, and lead-user design approaches is to raise the quality of products and services through methods that aid developers in user involvement. In the literature, the design context is often assumed to be 'one-off projects', which limits the applicability of the guidelines for further service design after market launch. Other challenges concerning social media include ambiguities in the role of informal engagement, the abstraction processes between millions of users and working user categories, and criteria for involving users. This study investigated user involvement strategies and practices in the construction of a social media service. The research questions were (1) how do users' actions in and around social media shape its design after market launch, (2) how do social media developers' user involvement practices evolve over time, and (3) how does user categorisation change with social media? This thesis adopted an exploratory case study approach and the data was collected during 2003–2010. While the study is grounded in usability research and human–computer interaction, the theoretical and methodological framework leaned on science and technology studies. The site of investigation was one of the world's largest social game and online communities for teenagers, Habbo Hotel, operated by Sulake Corporation. The data was gathered from developers and users through a multi-method approach, using traditional qualitative and quantitative methods as well as online data sources. While this study offers unique insight into the interaction between a particular social media company and its users, the implications go beyond the studied target group and games to social media in general. The key findings demonstrate how collaboration and feedback loops between developers and users change over time. In particular, this study highlights the effects of changes in the target group, the broad variety of applied user involvement methods, rhythms in development, and users' contributions after market launch. The author developed two concepts, 'developer–user social distance' and 'content creation capacity', to help designers and researchers communicate previously neglected dimensions of user involvement. Results from analysis of the accumulation of user knowledge in the development organisation criticise assumptions in the literature on when and how designers categorise users as well as the functions of these user categories in the design process. The results further suggest that guidelines and other advice on user involvement should be uncoupled from the assumption of stable and orderly project phases. This case contributes to user-centred design guidelines, process guidance for user involvement, and research on social media development and developer–user collaboration. Keywords user-centred design, strategic user involvement, participatory design, user innovations, social media, virtual worlds ISBN (printed) 978-952-60-5072-0 ISBN (pdf) 978-952-60-5073-7 ISSN-L 1799-4934 ISSN (printed) 1799-4934 ISSN (pdf) 1799-4942 Location of publisher Espoo Pages 347

Location of printing Helsinki Year 2013 urn http://urn.fi/URN:ISBN:978-952-60-5073-7

Sammandrag Aalto-universitetet, PB 11000, 00076 Aalto www.aalto.fi

Författare Mikael Johnson Doktorsavhandlingens titel How Social Media Changes User-Centred Design: Cumulative and Strategic User Involvement with Respect to Developer–User Social Distance Utgivare Högskolan för teknikvetenskaper Enhet Institutionen för datateknik Seriens namn Aalto University publication series DOCTORAL DISSERTATIONS 46/2013 Forskningsområde T011Z Användbarhetsforskning Inlämningsdatum för manuskript 17.04.2012

Datum för disputation 15.03.2013 Beviljande av publiceringstillstånd (datum) 20.12.2012 Språk Engelska Monografi

Sammanläggningsavhandling (sammandrag plus separata artiklar)

Sammandrag Genom att inkludera användare i designprocessen kan systemutvecklare höja kvaliteten på produkter och tjänster. I de främsta metoderna på området, användarcentrerad och deltagande design samt engagemang av "lead-users", utgår man ofta ifrån att designprocessen gäller enskilda projekt, vilket begränsar tillämpningen av metoderna för fortsatt tjänstedesign efter marknadslansering. Vidare utmaningar inom sociala media är oklarheter kring informellt engagemang, praktiska abstraktioner av miljoner användare samt kriterier för användarmedverkan. I denna studie undersöktes strategier för användarmedverkan och praxis i konstruktionen av sociala media. Frågeställningen var (1) hur användares aktiviteter i och kring sociala media formar dess design efter marknadslansering, (2) hur användarmedverkan utvecklas över tid och (3) hur användarkategorisering förändras i och med sociala media. Avhandlingen är en undersökande fallstudie och materialet sammanställdes under 2003–2010. Medan studien är förankrad i forskningsområdet människa–datorinteraktion och begreppet användbarhet, utgår den teoretiska referensramen också från samhällsvetenskapliga teknik- och vetenskapsstudier. Platsen för undersökningen var en av världens största sociala spel och mötesplats på nätet för tonåringar, Habbo Hotel, som drivs av Sulake Corporation. Uppgifterna samlades in från utvecklare och användare genom olika metodansatser. Både traditionella kvalitativa och kvantitativa metoder samt källor på nätet användes. Trots att fallstudien ger en unik insyn i interaktionen mellan ett specifikt företag och dess användare inom sociala media, går implikationerna längre än till den undersökta målgruppen och spel, dvs. till sociala media i allmänhet. De viktigaste resultaten påvisar hur samarbete och feedback mellan utvecklare och användare förändras över tid. I synnerhet belyses effekterna av förändringar i målgruppen, bredden i tillämpningen av metoder för användarmedverkan, rytmer i utvecklingsprocessen samt användarnas insatser. Författaren utvecklade två begrepp, "socialt avstånd mellan utvecklare och användare" och "kapacitet att skapa innehåll", för att hjälpa utvecklare och forskare att kommunicera tidigare försummade dimensioner inom användarmedverkan. Resultaten från analysen av hur kunskapen om användare byggs upp kritiserar ogrundade premisser gällande när och hur utvecklare kategoriserar användarna samt användarkategoriernas funktion i designprocessen. Resultaten påvisar också att riktlinjer och rekommendationer om användarmedverkan bör utvecklas utan premisser om stabila och välordnade projektfaser. Denna fallstudie bidrar till riktlinjer inom användarcentrerad design, processvägledning för användarmedverkan, forskning om utveckling av sociala media och samarbete mellan utvecklare och användare. Nyckelord användarcentrerad design, strategisk användarmedverkan, deltagande design, användarinnovationer, sociala media, virtuella världar ISBN (tryckt) 978-952-60-5072-0 ISBN (pdf) 978-952-60-5073-7 ISSN-L 1799-4934 ISSN (tryckt) 1799-4934 ISSN (pdf) 1799-4942 Utgivningsort Esbo Sidantal 347

Tryckort Helsingfors År 2013 urn http://urn.fi/URN:ISBN:978-952-60-5073-7

Preface

The point of no return passed in the fall of 2005. I realized that I had read enough theoretical texts to put my mind at ease about ‘the user’1 and that I had enough data to complete my thesis work. However, the research project was going to give me yet another Easter egg: instead of finalising my thesis during the third year of the project, I got an opportunity to engage even more deeply with my case organisation and its service users. The benefit was access to additional research data about user feedback in the organisation’s internal processes, but the drawback was having less dedicated time to write up my insights. I was uncertain about continued funding, but my research ambition, privileged data access, and prolonged commitment had fused into a mix of passion and obligation: I could not not write my thesis anymore. This turn of events eventually led to my participation in a series of research projects with interesting people and topics, but where my thesis work often remained a sidetrack. With ‘too much data’ on my hands I then pursued occasional collaborative writing projects that helped me sort out my thoughts and explore different angles to the research phenomenon. The year 2010 became a good year as I received a small scholarship to complete the thesis and a more than permissive research project started soon after, which let me continue the finishing touches for as long as I needed. After many more months, even years, than originally planned, this thesis has finally found its shape. I am indebted to many who have helped me on the way. My first research influences came from the (no longer active) Information Ergonomics Research Group that mixed software engineering, user interface design, usability research, and cognitive and social psychology. I arrived at a time of change and inherited the research projects that my supervisor, Marko Nieminen, had set up for himself, as he became acting professor at the usability and user interfaces professorship at Helsinki University of Technology. Thank you for letting me on-board and into the fascinating research projects. I am particularly grateful for the support from my advisors. Marko Turpeinen organised the Mobile Content Communities (MC2 ) research project, which opened the doors to the software development organisation that became my case study, and he supported me with clear thoughts on social media and research funding. Sampsa Hyysalo took me under his wings and made it his business to see that my thesis got done. Thank you for your early and continuous interest and encourage1 At that time I was puzzled with the mismatch between engineering talk about the user and most of the available social and behavioral science theories for understanding people that I had learned of in my studies, projects, and various academic reading circles.

Preface

i

ment, as well as the research funding and patience in the end. Parts of this thesis could not exist without my dear co-authors, whom I wish to thank for their contributions. Kalle Toiskallio shared the early Habbo experiences with me and guided me through the sociology of community. Vili Lehdonvirta showed me why I could not ignore the economics of virtual worlds and got me acquainted with the relevant economic theories and models together with TerhiAnna Wilska. Tanja Sihvonen and I explored whether a ‘nice’ place like Habbo could be gothic at the same time, while she enlightened me about subcultural studies. Sakari Tamminen and Sampsa Hyysalo form a fantastic duo on any social theory, both with their own well-argued perspectives, and this time we managed to write up our thoughts on the relevance of symbolic interactionism to virtual worlds. I wish to acknowledge the participation of all the kind Habbo people who let me interview and survey them. Both the ones working at Sulake, who created Habbo, and the active Habbo users who voluntarily took part in this research. For reasons of research anonymity I cannot mention you by name, but I wish to express my appreciation for letting me interview you, politely considering my questions and providing me with answers that always enlightened me and exceeded my expectations. My special thanks are extended to the research coordinators, who managed our collaboration: Reetta Pietiläinen, Sampo Karjalainen, Emmi Kuusikko, and Marjoriikka Ylisiurua. During the research process I had the opportunity to exchange thoughts with fellow Habbo researchers—Sebastian Sihvola, Aleksi Koskinen, Vili Lehdonvirta, Jani Merikivi, and Minna Ruckenstein—thank you for the insightful collaboration. I have been fortunate in having a good academic entourage, who taught me the academic ropes. I thank my former IERG colleagues at Software Business and Engineering Institute—especially Sakari Tamminen, Toni Koskinen, Teemu Seppälä, Kati Hyyppä, Viljami Koskela, Kalle Toiskallio, and Päivi Pöyry—for all the fascinating discussions in the early days. I also appreciated the wider collaboration at the institute, including our ‘scientific debate’ meeting and the reading circles on information systems as well as postmodern IT design, which helped me avoid so many theoretical pitfalls. My first research project had a strong connection to Oulu University and I am grateful for the lessons on user-centred design and usability in organisational processes and cultures that you provided: Kari Kuutti, Samuli Saukkonen, Timo Jokela, Tonja Molin-Juustila, Netta Iivari, and Giulio Jacucci. Through the MC2 research project I transitioned from SoberIT to Helsinki Institute for Information Technology, where I was fortunate to be part of the Digital Content Communities research group and, while it operated, the Self-Made Media group. This research environment was a change from the research projects with a pronounced industry context at SoberIT to consumer products and increased my awareness of legal and economic concerns in relation to digital service development. Thank you MC2 for introducing me to gaming, community, and content, especially the people at HIIT: Marko Turpeinen, Fernando Herrera, Kai Kuikkaniemi, Matti Rantanen, Timo Saari, Antti Salovaara, Risto Sarvas, Janne Vuorenmaa, Herkko Hietanen, Vili Lehdonvirta, Sauli Tiitta, and Esko Kurvinen. Thank you SMM for the always enjoyable research dialogue: Risto Sarvas, Asko Lehmuskallio, Vilma Lehtinen, Jaana Näsänen, Sami Vihavainen, and Kai Huotari. I

ii

Preface

appreciated the informal and illuminative discussions in the collaborative research environment also with Giulio Jacucci, Antti Oulasvirta, Olli Pitkänen, Perttu Virtanen, Airi Lampinen, Lassi Liikkanen, Kristiina Karvonen, and Matti Nelimarkka. A shared interest in user-centred design and user-driven innovation led me to the National Consumer Research Centre, where I learned why user-centredness is much more than a bilateral developer–user affair. Consumer and innovation policy, intermediaries, and activists were made relevant to me, both theoretically and in practice, through the research projects that for me mostly circulated around energy efficiency and changing patterns of energy use. Thank you Tanja Kotro, Eva Heiskanen, Petteri Repo, Päivi Timonen, Mikko Rask, Mika Saastamoinen, Kaarina Hyvönen, Minna Lammi, and Mika Pantzar, as well as the ones who helped me with the interview transcriptions: Eija Niiranen, Tuula Salo, and Arja Luoto—I also enjoyed our relaxed floorball games. The users and innovation research group INUSE has been invaluable support during the past two years. Thank you Sampsa Hyysalo, Pia Helminen, Stephanie Freeman, Jouni Juntunen, Samuli Mäkinen, Louna Hakkarainen, Sebastian Greger, and Tiina Kymäläinen for giving me time and enlightening me about lead-users and user innovation studies, I look forward to continue working with you. In the final stages my thesis improved a lot thanks to the excellent comments from my pre-examiners, Robin Williams and Dag Svanæs—thank you. I also appreciate the thoughtful feedback on draft versions that I received from Pia Helminen, Jouni Juntunen, Asko Lehmuskallio, Vilma Lehtinen, Petri Mannonen, Samuli Mäkinen, Sirpa Riihiaho, Antti Salovaara, Risto Sarvas, and Kalle Toiskallio. Proofreading would have been a mess without the fast operations from anonymous but competent editors at Papercheck.com. Besides the already mentioned people, significant for my research funding have also been Kari Kuutti, Hannakaisa Isomäki, and Martti Mäntylä. Thank you for the support. This research has also been made possible thanks to the Department of Computer Science and Engineering at Aalto University, the doctoral program in User-Centered Information Technology, and several research projects funded by TEKES, Academy of Finland, EU, NORDUnet, and Emil Aaltonen Foundation. I wish to express gratitude to all the interesting people I have met in the academic projects that I have been fortunate to be a part of. Despite the acronyms you know who you are: VIKSU, UISB, WISE, MC2 , Smint, Pamphlet, Create Acceptance, Changing Behaviour, UDOI Booster, Kulta, Prima, UIP, UIC. I wish to thank especially those not yet mentioned whom I have collaborated closely or co-authored other work with: Pirkko Jokela, Eija Suikola, Guy André Boy, Jouni Meriluoto, Marjaana Siivola, Sami Karjalainen, Mika Röykkee, Virpi Roto, Sonja Kangas, Marcus Bylund, Peter Seipel, Magnus Boman, Kari-Jouko Räihä, Saila Ovaska, Markku Turunen, Andrea Botero, Raimo Lovio, Aleksi Neuvonen, Samuli Rinne, Edina Vadovics, Simon Robinson, Ynke Feenstra, Gian Marco Campagnolo, Hajar Mozaffar, and Neil Pollock. You have been invaluable to me for sharing your expertise, for making me see my own perspective, and in developing my work. I would also like to thank my former colleagues at Polycon, the company where I worked as a software engineer and user interface designer during my studies. The experience from real-life software projects has been invaluable in grounding my research.

Preface

iii

I wish to thank all my friends for being there and giving me something else to think about, especially the Wednesday band—Tomas, Dan, Andreas, Jan—I always return home with a big smile and humming to the latest songs we played. I am grateful for the support from my parents, your new spouses, and everybody in my extended family, also on Ninni’s side—in particular for asking me what I am doing enough times until I figured out a way of explaining it. Thank you my brothers for being there, and recently also as IT professionals in our joint venture. You are also in my head while I am writing. Finally, my deepest gratitude goes to my beloved Ninni. I appreciate your love, support, and always considerate wisdom on so many things, including our work discussions at home, your expertise on pedagogy, history, and sociology—but most importantly for balancing life with other things than work. Helsingfors, February 2013, Mikael Johnson

iv

Preface

Contents

Preface

i

List of Publications

vii

Author Contributions

ix

1

Introduction 1.1 How Does User Involvement Change with Social Media? . . . . 1.2 Research Questions and Methodology . . . . . . . . . . . . . . . 1.3 Data and Case Description . . . . . . . . . . . . . . . . . . . . .

2

User Involvement and Social Media 2.1 Approaches to User Involvement . . . . . . . . . . . . . . 2.2 Influential Turns in User-Centred Design . . . . . . . . . . 2.2.1 Design Based on Usability Evaluation . . . . . . . 2.2.2 Turn to the Social and Contextual . . . . . . . . . 2.2.3 From Evaluation to Business Process . . . . . . . . 2.2.4 A Focus on User Experience . . . . . . . . . . . . . 2.2.5 Centering Design on Value(s) and Activity . . . . . 2.3 A Parallel Debate: Participatory Design . . . . . . . . . . . 2.3.1 Activist Origins . . . . . . . . . . . . . . . . . . . . 2.3.2 What Counts as Participatory? . . . . . . . . . . . 2.3.3 New Territories for Participation . . . . . . . . . . 2.4 User Innovation Research . . . . . . . . . . . . . . . . . . . 2.4.1 The Lead-User Methodology . . . . . . . . . . . . 2.5 Social Media: A Distinct Design Context . . . . . . . . . . 2.5.1 Social Media and Industry Lifecycles . . . . . . . . 2.5.2 Functionality for Group Communication . . . . . . 2.5.3 Users, Volunteers, and Communities . . . . . . . . 2.5.4 Persistent Challenges Accentuated by Social Media

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

11 11 14 15 16 16 17 18 20 20 20 21 22 23 24 27 29 34 36

Researching Habbo: Case Study Design 3.1 Users as Co-constructed . . . . . . . . . . . . . . 3.2 Borrowing from Science and Technology Studies 3.3 Benefits of an Explorative Case Study . . . . . . . 3.4 Mapping Research Questions and Activities . . . 3.5 Research Process Details . . . . . . . . . . . . . . 3.6 Data Analysis . . . . . . . . . . . . . . . . . . . . 3.7 Notes on Quality and Ethics . . . . . . . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

41 43 46 48 50 51 53 58

3

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

1 4 8 9

Contents

v

4

Key Themes of Article Contributions 4.1 Emergence and Co-Constitution of a Social Virtual World 4.1.1 User Practices in Habbo . . . . . . . . . . . . . . . 4.1.2 From Concept Design to Beta . . . . . . . . . . . . 4.1.3 Patterns for Expansion . . . . . . . . . . . . . . . . 4.1.4 User Intermediaries: Fansites and Volunteers . . . 4.1.5 Service Evolution . . . . . . . . . . . . . . . . . . . 4.1.6 Summary . . . . . . . . . . . . . . . . . . . . . . . 4.2 Challenging the Categorisation of Users . . . . . . . . . . 4.2.1 Survey Data: Habbo Visitor Profile . . . . . . . . . 4.2.2 Learning from Habbo Fansites . . . . . . . . . . . 4.2.3 Interviews with Habbo Users . . . . . . . . . . . . 4.2.4 Interviews with Developers . . . . . . . . . . . . . 4.2.5 Habbo Developers and the ‘Average User’ . . . . . 4.2.6 Summary . . . . . . . . . . . . . . . . . . . . . . . 4.3 Beyond the Scope of One Project in User Engagement . . . 4.3.1 Insight into Early Development Practices . . . . . 4.3.2 Expansion and Maturing Development Processes . 4.3.3 Service and Community Complexity Management 4.3.4 Competition and Globalisation . . . . . . . . . . . 4.3.5 Summary . . . . . . . . . . . . . . . . . . . . . . . 4.4 Pathways of User-Created Content and Contributions . . . 4.4.1 Content Creation Capacities . . . . . . . . . . . . . 4.4.2 Tweaking and Subverting . . . . . . . . . . . . . . 4.4.3 Developer Reactions to User Contributions . . . . 4.4.4 Fansites in Community Management . . . . . . . . 4.4.5 Summary . . . . . . . . . . . . . . . . . . . . . . .

5

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

61 61 62 64 66 69 71 74 76 77 79 81 82 83 86 90 91 93 95 97 99 103 104 106 110 112 114

Conclusions 117 5.1 Summary of Contributions . . . . . . . . . . . . . . . . . . . . . 118 5.2 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124

Appendix A: Stakeholder Analysis

131

References

141

Publications

vi

I

Fansites as Sources for User Research: Case Habbo Hotel

165

II

Who Are the Habbo Hotel Users–and What Are They Doing There?

187

III

Unscrambling the ’Average User’ of Habbo Hotel

209

IV

Virtual Consumerism: Case Habbo Hotel

239

V

On the Dark Side: Gothic Play and Performance in a Virtual World

263

VI

User Involvement, Social Media, and Service Evolution: The Case of Habbo

287

VII The Virtuality of Virtual Worlds, or What We Can Learn from Playacting Horse Girls and Marginalized Developers

299

Contents


This thesis is based on the following publications, which are referred to in the text by their Roman numerals:

I Mikael Johnson and Kalle Toiskallio (2005). "Fansites as Sources for User Research: Case Habbo Hotel". IRIS’28 Conference. The 28th Information Systems Research Seminar in Scandinavia. Kristiansand, Norway, August 2005. II Mikael Johnson and Kalle Toiskallio (2007). "Who are the Habbo Hotel Users—and What are they Doing There?" Nordic Consumer Policy Research Conference 2007. Helsinki, Finland, October 2007. III Mikael Johnson (2007). "Unscrambling the ’Average User’ of Habbo Hotel". Human Technology 3 (2), 127–153. IV Vili Lehdonvirta, Terhi-Anna Wilska, and Mikael Johnson (2009). "Virtual Consumerism: Case Habbo Hotel". Information, Communication and Society 12 (7), 1059–1079. V Mikael Johnson and Tanja Sihvonen (2009). "On the Dark Side: Gothic Play and Performance in a Virtual World". Journal of Virtual Worlds Research 1 (3). VI Mikael Johnson (2010). "User Involvement, Social Media, and Service Evolution: The Case of Habbo". 43rd Hawaii International Conference on System Sciences. Kauai, Hawaii, January 2010, 1–10, (Nominated for Best Paper Award). VII Mikael Johnson, Sampsa Hyysalo, and Sakari Tamminen (2010). "The Virtuality of Virtual Worlds, or What We Can Learn from Playacting Horse Girls and Marginalized Developers". Symbolic Interaction 33 (4), 603–633.


vii


Articles I and II are written in collaboration with Kalle Toiskallio, with whom the author collaborated on a daily basis in 2003–2005. Article I is based on the author’s idea, study design, data collection, analysis, and writing. In article II, Toiskallio contributed to the study design, participated in data collection, and commented on the analysis and the article text, written by the author. Articles III and VI: The author is solely responsible for everything. Article IV: The author provided supporting research data and contributed to the analysis and with comments on the text, mainly written by Lehdonvirta. Articles V and VII: The author was responsible for the study design, data collection, related research on virtual worlds, figures, writing of case descriptions, and examples. The idea and analyses emerged in joint collaboration with the co-authors, who also contributed with data interpretations and theory from the behavioural and social sciences.


ix

1 Introduction

When technological breakthroughs happen and new categories of products and services become established in society, it becomes urgent to assess their impact on people’s everyday lives. Equally important is the assessment of how different people influence the development of the new product or service. Contrary to common belief, technology is not shaped by scientists and developers alone, but users and various intermediaries (e.g., distributors, dealers, sponsors, advertisers, etc.) play a significant role. Before we delve into the roles of users, the notion of the ‘user’ deserves a little attention. While ‘user’ is a common and integral term in the professional language of engineering, architecture, and software development, it is also a complex idea: it does not always refer to particular human beings in flesh and blood, but also to organisations and, for instance, to someone who might use a system in the future, but has not yet been identified.1 In addition, user is tightly coupled with software concepts like user interface, user profile, and user access rights.2 Because of the deep roots of ‘user’ in design practices, the term is not easily substituted with participant, social actor, human, etc. (see section 3.1)—a change which at first glance might cast individual people in more appropriate social science light. Social media are often said to be user-centred. By now it is clear that social media can be distinguished from traditional broadcast media, especially considering the greater proportion of the content that the users produce, the increased immediacy, and the smaller degree of professional editing involved in the social media setting. It is, however, less clear whether the existence of user-created content means that social media is user-centred or designed in a user-centred way? It is not obvious how these two (user-created content in social media and the value of usercentredness in product development) relate, because anything that takes place after market launch is not very explicitly considered in traditional thoughts and processes on user-centred design; see for instance this definition by the association for usability professionals: User-centered design (UCD) is an approach to design that grounds the process in information about the people who will use the product. UCD processes focus on users through the planning, design and development of a product.3 1 Stewart

and Hyysalo, 2008. 1993. 3 Usability Professionals’ Association, n.d. 2 Grudin,

1 Introduction

1

where, for example, actual use, operation, maintenance, re-development, redesign, are missing in the definition. The role of users is very well outlined in three prominent approaches to user involvement: user-centred design, participatory design, and lead-user innovation. In user-centred design users are assumed to contribute through the means of user research, user requirements definition, context of use models, use case and scenario modelling, persona descriptions, and evaluation with users (section 2.2). In participatory design users are assumed to participate in design workshops and other co-design activities (section 2.3). From the perspective of lead-user innovation, people ahead of the market, so called lead-users, presumably create their own technical solutions and solve pressing problems, which can then be adopted and packaged as a product by companies (section 2.4). Are the above roles of users valid in the context of social media? Social media is here understood as a particular design context, which is distinct from others based on the combination of (1) a specific kind of software business, (2) functionality for group communication, as well as (3) voluntary active user communities and peer production (section 2.5). Anecdotal evidence from social media startups suggest that many developers did not start with typical user-centred design methods, but rather by developing the service for themselves.4 This would suggest that many social media startups have followed one particular lead-user pathway. However, many prominent social media companies have hired user experience designers and user researchers to learn from the users of their services. These weak signals intrigue us and lead to the question, what exactly is the role of users and user involvement methods in the design of social media? User involvement can be thought of in two ways: either how to manage a particular interaction situation with one or more users, or the planning process when one decides how to approach users. We know a lot about particular methods to learn about users—interviews, observation, surveys, focus groups, and so on— and a fair deal about which factors drive the use of a method in research settings. However, we know very little about the factors that drive the selection of methods use in the long run, in a series of projects in product or service development organisations. In usability and user-centred design so far, a number of factors that influence method use have been proposed. One standard of method selection criteria5 lists a number of factors structured by software lifecycle, project, user, task, and product characteristics, as well as available skills. For instance, does the designer have access to users or are they too remote—geographically or organisationally? What ergonomics/human factors skills does the design team have? How much time and money is available? And is it in the beginning, middle, or end of a project? These contextual factors that shape the selection of user involvement methods all make sense, except for one. It looks like the characteristics of software development for social media reframe the role of project phases (in the software lifecycle: acquisition and supply; development: requirements analysis, architectural design, qualification testing; maintenance-operation)6 as a driving factor. In addition, I 4 Holzapfel,

2008. 2002, Usability methods supporting human-centred design, ISO/TR 16982. 6 ISO, 2002, There are many small variations of the software lifecycle process description and it is 5 ISO,

2

1 Introduction

will propose that ongoing software development in a social media context has additional advantages, like new sources for data about users, so far neglected by the previously mentioned standards. Why should one care about these shaping factors, and when does one need to pay attention to them? Guidelines and method resources are important in capturing and communicating design knowledge grounded in practitioner experience and research. Since there is no lack of methods, but rather an oversupply—some have even called it a ‘methodology jungle’7 —the ways that these shaping factors structure method resources are important for successful method selection. The risk is that poorly structured method resources lead to suboptimal method choices, resulting in reduced impact of user involvement. The context factors are relevant in the early stages when a project takes form, when one assesses what kind of project it might be, its potential benefits, and importance for the organisation. Also, the factors become key to the detailed planning of the project, when one needs to sort out the nitty-gritty details of engagement with users. Finally, the factors are significant in the post-mortem analyses of the project, when one assesses what went right and what could have been done better. It is the effective and learning organisation that needs to care about these matters. The emergence of social media has had a significant effect on software business and related products and services (e.g., product characteristics in the abovementioned standard). Software started out in the 1950s as something expensive that only big companies or the government could afford. The mainframes of the 1960s turned the business around as much as the personal computer did in the 1970s and 1980s. The Internet changed the distribution of software in the 1990s. Engineering efforts of miniaturisation, increased speed of interconnected parts, and mobile and widespread networks allowed for the home computer and mobile devices in the 2000s to shape and cater to new customer groups. Recently, social media changed the software business, or at least parts of it. Some parts have always remained as they were before, while other niches have gained a stronger foothold in the reigning ways of making business in the software landscape. The question of the role of the users in product development has been asked before in many different fields with somewhat varied answers related to different technologies. Business books on social media often point out the role of collaboration, scalability, open source, new functionality, and a viable business model.8 However, they are very unclear about the specifics of user involvement and user experience. Open questions includes such fundamentals as how customer dialogue and user feedback evolves over time and in changing contexts. difficult to find a canonical one, for instance ISO/IEC 12207 defined 43 system and software processes, of which ISO 16982 selected the mentioned ones to cross-tabulate usability methods with. 7 Avison and Fitzgerald, 1988; Iivari, Hirschheim and Klein, 2001. 8 Anderson, 2006; Osterwalder and Pigneur, 2010; Tapscott and Williams, 2006.

1 Introduction

3

• • •

1.1 How Does User Involvement Change with Social Media? Before going into the main traditions of user involvement, I will introduce the notion of user categorisation. The first step of user-centred design and similar approaches is to identify and categorise potential future users. This is typically done to complement market segmentation and product differentiation. Market analysts pursue market segmentation to forecast marketplace acceptance of products and services, and often base their analyses on demographics and consumption patterns.9 Such analyses give a basis for interaction design, but designers often supplement them with analyses of current and future use practices, which is the foundation for scenarios, personas, user stories, storyboards, use cases, and so on. There is no commonly agreed on umbrella concept for these use-based analyses, but here I group them together as user categorisation. I draw on the concept of categorisation instead of classification to acknowledge that the categories we create to understand the world can be conflicting and overlapping, whereas classification ‘involves the orderly and systematic assignment of each entity to one and only one class within a system of mutually exclusive and nonoverlapping classes’.10 The idea behind the techniques of user categorisation is to open up the concept of the user to nuances so that technology is not designed for an average user, which in the end might not suit anyone in particular. On the other hand, too many user categories cannot drive the design, because that would make designers torn in too many directions. It is considered good design practice to design for a few selected users: ‘a given product will have a an informative suite of about five or six personas of which we will focus our design on one or two’.11 Sometimes developers discuss various user categories as their future target groups or audiences, other times12 different designations of users become a question of representing various established user groups in design.13

User-Centred Design, Usability, User Experience There has been significant theory and methods development during 1990s and 2000s in the fields of user-centred design, usability engineering, and user experience. Early approaches in the 1980s were based on usability evaluation, then user-centred design turned to the social and contextual, business processes, user experience, and recently to value- and activity-centred design. Many methods 9 Kotler,

1997. 2004, p. 522. 11 Gaffney, 2006, Personas and outrageous software—an interview with Alan Cooper. 12 User categorisations are but a small segment of the much broader concept of user representations, which is one of the most cited concepts in science and technology studies that focus on the design– use relationship. The concept of user representations enable a multitude of analyses, e.g., of users’ ability to influence a design, of the resources for creating user representations (market research, user panels, visions, myths, experts, product demand, etc.), of the quality of a user representation (is it representative?), and of the various material forms of representation (reports, sketches, pictures, video clips, formal notations, etc.). See Akrich, 1995; Hyysalo, 2004; Williams, Stewart and Slack, 2005. 13 Akrich, 1995; Carroll, 1995. 10 Jacob,

4

1 Introduction

books have emerged, several each year, and the approaches have reached consensus enough to become standardised by the International Standards Organisation (ISO), which has an authoritative role in engineering, especially related to certification, material, and process standards. Despite a growing body of literature, conferences, and workshops, not much is known about the actual practice of user-centred design in design and development organisations. The user-centred design literature is based on normative methods books and lessons learned, but there is very little method-validation research. As Rogers noted, it appears that most user theories and methods suffer from the toothbrush syndrome: ‘fine for you to use but no one else is very interested in using it’.14 There are more than 10 widely cited national and international surveys on user-centred design and user experience practices in companies, but in most cases they only scratch the surface.15 There are several researchers and consultants performing usability maturity assessments, but while the results are sometimes published, most research discusses how to conduct the assessment, leaving out the details of the design practice for business confidentiality reasons.16 With a few exceptions,17 case studies on user involvement are very dispersed and their combined contribution to the field remains unclear, despite significant potential. Over the years, consensus has emerged that it is a good idea to involve users, learn about their needs, and spend time early in the project to understand user requirements. While there are a few common and fairly generic user-centred practices—user studies, user feedback and testing with users—the actual details remain debated. Is it enough to meet real users over a cup of coffee or are formal research methods necessary? For instance, contested assumptions are that • a user-centred design project always starts with a field study,18 • and a usability evaluation is necessary for product success.19 Other fundamental, but less-researched, assumptions about user-centred design, usability, and user experience are whether • user involvement follows the same design phases from project to project,20 • only observed and well-documented user actions or speech can be used as design input,21 14 Rogers,

2004. Nielsen, Schrøder and Stage, 2006; Bygstad, Ghinea and Brevik, 2008; Dillon, Sweeney and Maguire, 1993; Følstad, Jørgensen and Krogstie, 2004; Gulliksen, Boivie and Göransson, 2006; Gulliksen, Boivie, Persson, Hektor and Herulf, 2004; Hussain, Slany and Holzinger, 2009; Nørgaard and Hornbæk, 2006; Venturi and Troost, 2004; Vredenburg, Mao, Smith and Carey, 2002, Of these, most cited are the ones by Vredenburg et al. and Gulliksen et al., according to Google Scholar in early 2012. 16 Jokela, 2004. 17 Bargas-Avila and Hornbæk, 2011; Righi and James, 2007. 18 Norman, 2006. 19 Greenberg and Buxton, 2008. 20 User-centred methods guidelines and advice are commonly structured according to project phase, see ISO, 2002; Maguire, 2001b; UsabilityNet, 2003. 21 Many user-centred design approaches have a bias on formal methods and design rationale, e.g., Beyer and Holtzblatt, 1998, whereas the role of informal engagement remain unclear. 15 Andreasen,

1.1 How Does User Involvement Change with Social Media?

5

• users can be abstracted into a manageable number (below 10) of typical users or user groups,22 • and particular methods are needed and can be developed for user-centred design purposes.23

Participatory Design Participatory design has broadened its scope since its early trade union activist origins of new technology and work democratisation of the 1970s and 1980s. The biannual Participatory Design Conference joins together both designers and people intrested in user involvement without a design or product development background. The contributions from this research field do not form any one-sizefits-all approach, but rather offers sensibility for tackling politics in design and a smorgasboard of design workshop methods and techniques. Research in participatory design is often case study based, but a few method frameworks have been developed (e.g., MUST24 ) Social media brings to forth a number of issues that have both challenged and developed the traditional methods of participatory design. First and foremost, there is the question of scalability of the approach as the number of participating users increase. Put in more detail, in a social media context, can we still assume that • the form of user participation is equal to being active in face-to-face meetings and workshops with developers, or observing and engaging in their activities, • it is (timewise) possible for a design team to involve all relevant users and listen to their opinions, • detailed knowledge about and localisation to organisational context of use (work processes) is key to product success, as opposed to standard product packages, and • the target organisation and users are known and stable within the participatory design project timeframe?

User Innovation Research & Lead-User Methodology User innovation research has managed what many user-centred and participatory design advocates have not: to attract the attention of business people and policy-makers. In contrast with the dominant belief in economics, user innovation research has revealed that a significant part of advances in technology stem from developments and modifications by users, not from R&D departments, researchers, or designers.25 22 The Personas method suggested by Cooper, 1999, retains that 5–6 personas are enough. In a broader sense, the role of user categorisation in design is unclear, especially in new contexts like social media. 23 This harks back to debates about appropriate scope of user-centred design efforts: Is it a particular method, a process, or also a way of organising practice and attitudes? Compare, for instance, the ISO 13407 standard approach with Gulliksen’s broader scope, see Gulliksen, Göransson, Boivie, Blomkvist, Persson and Cajander, 2003; ISO, 1999. 24 Bødker, Kensing and Simonsen, 2004. 25 Hippel, 1988; Hippel, 2005; Hippel, Thomke and Sonnack, 1999; Lüthje, Herstatt and Hippel, 2005.

6

1 Introduction

However, the lead-user methodology focuses on a fairly narrow group of users, which leaves out other user groups and the role of users after market launch of a product or service. Research on social media gives reasons to believe that other users, in addition to a few active lead-users, also play significant roles in later innovation stages as well.26 Dominant assumptions appear to be that • the majority of the users is not significant to innovation, only lead-users are, • the product concept stays the same for the whole market lifecycle, and • laggards need more or less the same thing as lead-users and early adopters.

Challenges for Research To sum up, despite a great deal of user involvement knowledge in several fields, there are a few problematic issues that pose challenges for research: User categorisation. The outlined approaches—user-centred design, participatory design, and lead-user methodology—have different strategies for selecting the sample (chapter 2), but what becomes of these strategies in the context of social media? Software services are now used by millions of people, not hundreds or thousands like in the 1990s, and not much is known about how this has influenced the ways that developers consider users. Is the individual still an adequate level of abstraction for user categorisations, or are groups of users or subcommunities more suitable? How do the new social media opportunities for data about users shape user categorisation practices? How can and should the organisation talk about the users, and which user groups can steer the design best? Which user voices will be heard, which will be contested, and which will fade out to noise? Dynamics beyond the scope of one project. A dominant way of structuring the use of methods to learn about users and co-design with users has been to use project phases more or less tied to the so-called software lifecycle (acquisition and supply; development: requirements analysis, architectural design, qualification testing; maintenance-operation). Very little is known about user involvement practices beyond the scope of one project. How does user involvement evolve over time? Can one rely on informal engagement with users or are formal user involvement methods necessary? Which aspects shape user involvement method selection? How do the roles that the users perform change between projects? Pathways of user-created content and contributions. User-created content stirred up many debates when Blogger, Flickr, YouTube, MySpace, Digg, FeedBurner, Second Life, and similar services became societal phenomena. The service developers had not only created text-, photo-, video- and audio-sharing systems, but also publication systems. What one user did and contributed to the service became available to other users and contributed to their experience of the service. The fact that users are creative and contribute to social media is not news to anyone anymore, but what pathways do these user contributions take? If the users take an active role, how do their opinions and content shape the product, feature, and service to be? Can a user-created content perspective help answer the questions of what roles users play in social media and how they help make software work in practice? 26 Benkler,

2006; Bruns, 2008; Gillmor, 2004; Jenkins, 2006; Shirky, 2008.

1.1 How Does User Involvement Change with Social Media?

7

• • •

1.2 Research Questions and Methodology I started out with an interest in the process of constructing the user (section 3.1), which implies that one is not born a user, but becomes a user of a particular product or service. This perspective implies a distinction between the users as imagined by developers and the users who actually use a system, including an analysis of their interrelations. Given the research challenges presented in the previous section and my research interest, the overarching theme of this thesis became the role of users and user involvement methods in social media. Put more stringently, the research questions are: 1. How do users’ actions in and around a social media service shape its design after market launch? 2. How do social media developers’ user involvement practices evolve over time? 3. How does user categorisation change with social media? Whereas the second and third questions are fairly straightforward, based on the above research challenges, the formulation of the first question requires a brief comment on how it relates to pathways of user-created content and other contributions. The first question is a variant of ‘how does use shape design’? In many other design contexts only projected future use or limited use during evaluation influences a product, not actual use. This question intends to capture the intended and unintended effects to design of this actual use (including usercreated content), which is fairly invisible in other design contexts with longer times between product updates, where users get to influence a product mostly before market launch, during concept design and testing, if at all. The formulation is intentionally broadened from use to ‘actions in and around a service’ in order to include actions by users that bend the limit of what can be understood as use, such as actions that are design oriented or complement the service offering (peer production). Because of the unclear connection between literature and development practice, I chose to study real-life software development through a case-based qualitative inquiry. I also made a few other unorthodox choices about methodology. The user– developer relation is an understudied topic, since the literature is unfortunately usually divided into either research into design practices or research into the use of products and services.27 In contrast, I chose to study both developers and users, as well as intermediaries. Also, I chose to make a long-term commitment to a case in order to study several iterations of design and use. Institutionally this work is performed under the rubric of usability research— a core part of human-computer interaction (HCI)—at a department of computer science and engineering, but I also lean on a theoretical framework from the crossdisciplinary field of science and technology studies (STS). From this field I borrow the notion of users as co-constructed, in contrast to mainstream notions of users 27 Hyysalo, 2004; Hyysalo, 2010; Pollock and Williams, 2009; Voss, Hartswood, Procter, Rouncefield, Slack and Büscher, 2009; Williams, Stewart and Slack, 2005.

8

1 Introduction

as psychological subjects (typical in human-computer interaction), social actors (information systems), participants (from participatory design), and lead-users (user innovation research). Like many reflective HCI researchers, my ambition here is to surface the often unstated assumptions and values embedded in the research field itself and in the advice produced to inform design practice. Much research in STS concerns values in design, and here I draw on discussions about classification, stakeholder analyses, and a sensitivity towards materiality. The Biography of Artefacts framework (section 3.2) has guided me as I have followed particular Habbo features from their visionary beginnings through development, user feedback, and subsequent redevelopment. • • •

1.3 Data and Case Description This study focuses on the user involvement practices of social media developers, exemplified by the developers at Sulake Corporation Oyj, who operate a social virtual world for teenagers called Habbo Hotel. The case study was started in 2003, as Sulake participated in a research project led by the Helsinki Institute for Information Technology. A longitudinal research approach made it possible to analyse user involvement in different stages of the service evolution, as both the company and user communities have grown. Initially the startup company began operations small, with hundreds of users, grew from microsize through small and medium organisational forms to an international corporation of about 300 employees with a user base of 10–15 million unique users per month. While this study offers unique insight into the interaction between a social media company and its users, this case is representative beyond this target group and games to social media in general for a number of reasons. For instance, the case is typical with respect to software business characteristics, functionality for group communication, active user communities, and developers’ non-traditional means of learning about users (following online use, discussions, and web analytics). From an innovation lifecycle perspective, this case is about the fermentation era, where technological variation is broad, and dominant designs have yet to appear. Habbo is a virtual environment where children and teenagers meet, socialise, and play many types of games. It was first launched in August 2000 in Finland as Hotelli Kultakala (‘Hotel Goldfish’), and it was based on the developers’ two earlier online services. At the time of writing, Habbo has local payment systems in more than 30 countries, and 5 million players visit Habbo each month.28 Instead of an entrance or a monthly fee, the profit model is based on micropayments in the hotel. Virtual furniture, mini-games, and membership in the Habbo club are bought with Habbo currencies. These currencies can either be earned in Habbo or purchased (depending on the country) with pre-paid cards, bank transactions, credit cards, or special text messages that add a specified amount of money to the customer’s mobile phone bill. 28 Sulake,

2012.

1.3 Data and Case Description

9

The social interaction in Habbo is diverse. In the design of Habbo, clear winning conditions and gameplay rules have been avoided, and instead, players are encouraged to create their own objectives beyond chatting, room decoration, and meeting friends. The provided environment for these activities is a hotel consisting of public and private rooms, where the virtual hotel visitors, called habbos, chat, buy virtual furniture, decorate rooms, and arrange social and game events. Most of the teenage players log on after school, and according to the developer company, on average they spend around 40–45 minutes per day in the hotel or on its related discussion forums.29 The data was gathered both from developers and users through a multi-method approach with varying intensity over eight years (more details in chapter 3). The research started in the fall of 2003 with pilot interviews and participant observation in Habbo user communities. During 2004 I analysed texts written by Habbo users on websites, blogs, and in discussion forums, as well as visitor profiles through a survey that reached 10,000 users. What the Habbo users do in Habbo is reported in articles II and IV, and summarized in section 4.1. The non-professional websites and discussion forums produced by Habbo users, so-called Habbo fansites, have been an important data source for understanding the consumption of Habbo. Since Habbo was launched, active users have kept track of the furniture, events, changes, trends, fashion, and other debates in Habbo. I have followed and logged this usercreated documentation of Habbo since the beginning of this research. The fansite analysis is reported in article I and elaborated on in sections 4.1.4, 4.2.2, and 4.4.4. In 2005 I did 10 theme interviews with Habbo developers and three focus group interviews with altogether 12 Habbo users. In 2006 I participated in the development of customer feedback methods at Sulake. Since 2007 I have regularly tried out new features in Habbo and kept up-to-date through additional interviews with Sulake developers. In addition, Sulake representatives participated in seminars organised by the MC2 research project in 2003–2006, workshops on virtual economy, and several project meetings. Analyses of the developer processes are reported in articles III and VI, and further elaborated in sections 4.2 and 4.3. The data analysis proceeded in multiple waves over the years. The survey provided quantitative information of the use of Habbo. The fansites explored different Habbo consumption styles, popular activities, and hotel history (article IV). The topics of the user interviews were their participation histories, changing motivations, and meanings given to membership and reference groups in Habbo (article VII). Taken together, these bodies of data provide us with an excellent view of the varying forms of interchange and dialogue between the varying users and developers of this virtual world, and allows in-depth access into how Habbo users have shaped the service as well as to how the developers’ user engagement and research strategies have evolved over the years.

29 Sulake,

10

2012.

1 Introduction

2 User Involvement and Social Media

This chapter gives an overview of the most prominent approaches to user involvement: user-centred design, participatory design, and user innovation research. The section on user-centred design includes discussions on usability and user experience. The review focuses on the assumptions about the role of users made in the communication of these design approaches. Even though these three approaches to user involvement are often lumped together, they have markedly distinct assumptions about the role of users. The last section opens up the concept of social media and some of its potential implications to user involvement. • • •

2.1 Approaches to User Involvement User involvement is commonly presented as a discussion about methods to engage, involve, and learn about users. Many methods and models for user involvement became popular in the 1990s, when a dominant way of framing methods as connected to project lifecycle stages was established. Whereas some argued that user research activities should precede technical design,1 a number of authors argued for integration with technical design processes.2 In this latter view, methods were often connected with project lifecycle stages, a framing that was also adopted by the HUSAT Research Institute and the EC UsabilityNet project (Figure 2.1), which evolved into two ISO standards,3 as well as in work by the usability professionals’ association.4 Previously Muller and Kuhn (Figure 2.2) created a methods compilation, which ended up in one handbook of human-computer interaction.5 These method compilations have in common that they structure methods according to a specific time or phase in the development cycle, incidentally the horizontal axis in both figures. These more and less normative guidelines can also be compared with how designers really think about user involvement methods. Goodman-Deane et al. chose 57 methods for understanding and involving users, described each on a card, and 1 Hackos

and Redish, 1998. 1999; Mayhew, 1999; Nielsen, 1993; Vredenburg, Isensee and Righi, 2002. 3 ISO, 2000; ISO, 2002; Maguire, 2001b. 4 Ross, Nowicki, Solomon, Yarbrough and Schwendeman, 2000. 5 Muller and Kuhn, 1993. 2 Cooper,


11

Figure 2.1: Methods for Human-Centred Design by Maguire, 2001b.

Figure 2.2: Taxonomy of PD Practices by Muller and Kuhn, 1993.

12


asked 21 designers to sort the methods into groups using any criteria they liked.6 After a cluster analysis, six clusters of methods were identified (Figure 2.3). Here, the methods became structured according to designers’ tasks: examining the market and analysis, understanding users with or without direct contact, and prototyping as well as concept design.

Figure 2.3: How developers think about methods by Goodman-Deane, John Clarkson, Langdon and Clarke, 2008. Whereas the first and third figures are tied to developer activities, the second figure also includes a dimension of participation that ranges from design activities to users’ worlds. This participation dimension is related to how independent users are from developers. In the introduction to the book Appropriating Technology, Eglash claims that user involvement approaches differ in how they foster dependence or independence (from developers) (Figure 2.4).7 Eglash argues that it is not fruitful to romanticise independence, as some times more dependence is needed to facilitate institutionalisation, and other times more independence is needed to free up new possibilities. When is what approach viable? Leonard-Barton argues that it depends on the maturity of the market and how new the technology is to the world (Figure 2.5).8 6 Goodman-Deane,

John Clarkson, Langdon and Clarke, 2008. 2004. 8 Leonard-Barton, 1995. 7 Eglash,

2.1 Approaches to User Involvement

13

Figure 2.4: The spectrum of dependence-independence for appropriated technologies by Eglash, 2004.

Figure 2.5: Methods for companies to learn from the market by Leonard-Barton, 1995.

With this brief background on user involvement methods, let us examine three of the most prominent long-term schools of thought related to developer-user relations more closely. They feature in many of the methods compilations and have different enough characteristics to illustrate the range of user involvement methods. • • •

2.2 Influential Turns in User-Centred Design User-centred design has become a more mainstream topic than before, figuring not only in technical debates, but also in policy and media. However, there is a risk that the concept of user-centred design is becoming diluted as it has travelled far from its origins. User-centred design was coined in the mid–1980s and became part of specific industry practices, before the widespread use of the Internet, computers, and mobile devices. It has been updated and diversified over the years, but 14


to support a reflection on what user-centredness might mean today in the context of social media, the following sections give an overview of the different turns usercentredness has taken. The following discussion of turns in user-centred design leans on similar discussions in human-computer interaction.9 This discussion explicates assumptions behind the various turns in the user-developer relationship and the design context.

2.2.1 Design Based on Usability Evaluation User-centred design emerged as a concept in the mid–1980s.10 At that time it was closely tied to ‘design for usability’ and its three key principles, which were: (1) focus on the user early in the design process, (2) measure the quality of the system from a user point of view, and (3) multiple design iterations so that evaluation results have time to influence the end result.11 Usability evaluations help designers to ‘assess our designs and test our systems to ensure that they actually behave as we expect and meet the requirements of the user’.12 Various usability evaluation methods emerged for different stages in product development to predict or measure how effective, efficient, and satisfied users would be with the system in use. Common methods were laboratory-based user observations, controlled user studies, and inspection techniques.13 Even though usability evaluation is still one of the cornerstones of user-centred design, it has received criticism over the years. Early studies that compared the effectiveness of different usability evaluation methods have been criticised,14 which has resulted in more rigorous method evaluations. In 1998 Jacobsen, Hertzum, and John, and then Molich et al., revealed in a famous series of comparative evaluations that different usability labs provided different results for the same evaluation assignment.15 Hornbæk found several distinct problems with usability measures in 2006 through a survey of current practice.16 Greenberg and Buxton noted in 2008 that usability evaluation has gotten such a strong foothold in current practice that ‘educational institutes, academic review processes, and institutions with usability groups advocate usability evaluation as a critical part of every design process’. They warned against following the principles blindly, as sometimes thoughtless usability evaluation can be ineffective or even harmful to radical innovations.17 Barkhuus and Rode criticised current evaluation practice for having low diversity among evaluation subjects and a bias towards male university students.18 9 Baecker, 2008; Grudin, 2005; Myers, Hollan, Cruz, Bryson, Bulterman, Catarci, Citrin, Glinert, Grudin and Ioannidis, 1996; Rogers, 2004; Rogers, 2009. 10 Norman and Draper, 1986. 11 Gould and Lewis, 1985. 12 Dix, Finlay, Abowd and Deale, 1993. 13 Dix, Finlay, Abowd and Deale, 1993; Nielsen, 1993; Rubin, 1994. 14 Gray and Salzman, 1998. 15 Jacobsen, Hertzum and John, 1998; Molich, Ede, Kaasgaard and Karyukin, 2004; Molich, Thomsen, Karyukina, Schmidt, Ede, Oel and Arcuri, 1999. 16 Hornbæk, 2006. 17 Greenberg and Buxton, 2008. 18 Barkhuus and Rode, 2007.

2.2 Influential Turns in User-Centred Design

15

2.2.2 Turn to the Social and Contextual In the late 1980s it became apparent that the cognitive foundations of humancomputer interaction did not scale up well to multi-user systems. Some researchers and designers adopted broader frameworks such as distributed cognition,19 activity theory,20 and ecological psychology.21 Others rejected the cognitive foundations and ‘turned to the social’.22 Sociologists and anthropologists brought situated action and ethnography to the table.23 Debates on how these new approaches would help designers and potential implications to design emerged.24 Central to many of the initiatives were the questions of ecological validity and a desire to better take into account situational and contextual aspects.25 A widely used design approach called contextual design emerged in the early 1990s.26 It became more refined and better packaged27 and also tuned for agile development.28 The approach was developed in contrast to usability engineering and artefact examination, and emphasised interview methods conducted in the context of the user’s work, co-designing with the user, building an understanding of work in context, and summarising conclusions throughout the research.29 Although broadly influential, the social and contextual approaches have received their share of criticism. Rogers found out through a survey of designers that many of the suggested approaches require too much effort or are too difficult to use within the timeframe of normal design projects.30 Clemmensen, on the other hand, reported that many Danish designers are happy to use social theory in their work.31 Stolterman argues that many theoretical contributions are too abstract or theoretical, or do not lead to the desired results when used in practice, and that many theoretical moves have failed to make an impact, because of an insufficient understanding of design practice.32

2.2.3 From Evaluation to Business Process One point of the key principles of user-centred design is to show that usability evaluation in the end of the design process is not enough. The evaluation results require iterations of the design process in order to make an impact. The design principles advocate a focus on the users from ‘day one’ in the project, in order to let user needs drive the design. These ideas were formalised in the late 1990s into 19 Hutchins,

1995. 1989; Engestrom and Middleton, 1996; Kuutti, 1996; Nardi, 1996. 21 Gaver, 1991; Norman, 1988. 22 Bannon, 1992; Button, 1993; Dittrich, Floyd and Klischewski, 2002; Thomas, 1995. 23 Blomberg, Giacomi, Mosher, Wall, Schuler and Namioka, 1993; Nardi, 1997; Suchman, 1987. 24 Barab, Thomas, Dodge, Squire and Newell, 2004; Cooper, Hine, Rachel and Woolgar, 1995; Dourish, 2006; Harper, 2000; Nyce and Löwgren, 1995; Shapiro, 1994. 25 Kaptelinin, Nardi and Macaulay, 1999. 26 Wixon, Holtzblatt and Knox, 1990. 27 Beyer and Holtzblatt, 1998. 28 Holtzblatt, Wendell and Wood, 2005. 29 Wixon, Holtzblatt and Knox, 1990. 30 Rogers, 2004. 31 Clemmensen, 2005. 32 Stolterman, 2008. 20 Bødker,

16


engineering standards (ISO 13407: Human-centred design processes for interactive systems, and ISO 18529: Human-centred design lifecycle process descriptions)33 so that companies could evaluate their processes according to objective standards. Philip and Rourke argue that the ISO 13407 was ‘instrumental in recognising UCD as being equivalent to other key business processes’.34 With the two international standards, user-centred design became a tool for project managers. Instead of focusing on particular methods for evaluation or field studies, such as, for instance, UsabilityNet,35 the standards put the focus on business processes and organisational capabilities. It was left to the project team to decide which particular evaluation method to use when, and Maguire provided an overview of the options.36 In addition, a scale was created to measure how mature an organisation was concerning usability. The scale ranges from ‘usability unrecognised’ to implemented, integrated, and institutionalised.37 According to Nielsen, it can take as long as 20 years to become a completely user-driven corporation.38 Even though there are straightforward guides to institutionalise usability,39 it is unclear how well the maturity model based usability recommendations fit development practice. Some claim maturity models in general are better suited for medium-sized or large companies, and less suited for small teams with agile development.40 These voices try to establish maturity models tuned to agile processes, while others claim that maturity models and agile models can coexist and benefit each other.41 The integration of usability and user experience (see the next subsection) with agile development is debated, with both concerned voices and success stories.42 Recent research has also criticised these and other models for having adopted too simplified notions of projects, development context, and methods.43

2.2.4 A Focus on User Experience Mobile phones, computers in the home, and digital games reoriented user-centred design. Work was no longer the sole context or location of computer use, which 33 ISO,

1999; ISO, 2000, these standards treat ‘human-centred’ and ‘user-centred’ as synonyms. and Rourke, 2006. 35 UsabilityNet, 2003, suggests the following methods for the requirements phase: surveys, interviews, contextual inquiry, user observation, context of use analysis, focus groups, brainstorming, evaluating existing systems, card sorting, affinity diagramming, scenarios of use, task analysis, and requirements meeting. 36 Maguire, 2001b. 37 Bevan, 2001; Earthy, Jones and Bevan, 2001; ISO, 1999; ISO, 2000. 38 Nielsen, 2006a; Nielsen, 2006b. 39 Schaffer, 2004. 40 Agile development is an umbrella concept for lightweight software development methods that emerged as a reaction to heavily regulated and micromanaged software development methods. Agile software development values individuals and interactions over processes and tools; working software over comprehensive documentation; customer collaboration over contract negotiation; and responding to change over following a plan. Beck, Beedle, Bennekum, Cockburn, Cunningham, Fowler, Grenning, Highsmith, Hunt, Jeffries, Kern, Marick, Martin, Mellor, Schwaber, Sutherland and Thomas, 2001. 41 Glazer, Dalton, Anderson, Konrad and Shrum, 2008. 42 Hussain, Slany and Holzinger, 2009; Larusdottir, Bjarnadottir and Gulliksen, 2010; Straub, Patel, Bublitz and Broch, 2009; Venturi, Troost and Jokela, 2006. 43 Svanæs and Gulliksen, 2008; Woolrych, Hornbæk, Frøkjær and Cockton, 2011. 34 Philip


17

meant that effectiveness and efficiency (two of the three measurement criteria for usability) as well as ‘organisational requirements’ in the context of use analysis44 seemed difficult to apply in the new mobile and leisure contexts for design. Approaches based on usability and user-centred design started to appear outdated, and concepts such as interaction design 4546 and design for user experience 4748 became more popular. ‘Having fun’ become a design objective.49 Both concepts, interaction design and user experience, are often positioned as something broader than usability and user-centredness. Sharp et al. state that successful designers need skills from ‘psychology, human-computer interaction, web design, computer science, information systems, marketing, entertainment, sociology and business’.50 However, the new methods associated with user experience, such as persona descriptions,51 goal-directed design,52 and the methods list in Kuniavsky’s practical book,53 are also communicated as user-centred design methods, or appear fairly similar to previously presented design methods, which makes some authors treat user experience as one part of a broad user-centred design umbrella.54 User experience methods have been criticised for focusing solely on user studies, user feedback, and user testing, thereby forgetting design and performance.55 Constantine and Lockwood advocate usage-centered design instead, which would take model-driven exploration, comprehensive task modelling, and model-driven abstract prototyping seriously.56 Battarbee criticised user experience methods for focusing too much on the individual’s experience, thereby forgetting how user experience is constructed in social interaction.57

2.2.5 Centering Design on Value(s) and Activity The ideas related to being user-centred are sometimes taken out of context, which has led some to interpret user-centredness as ‘users know best’. A contributing factor to this opinion is that the innovative encounters between business, techno44 Maguire,

2001a. definition of interaction design is ‘designing interactive products to support the way people communicate and interact in their everyday and working lives’. See Sharp, Rogers and Preece, 2007. 46 Cooper, 2003; Preece, Rogers and Sharp, 2002. 47 One definition of user experience is ‘a person’s perceptions and responses that result from the use or anticipated use of a product, system or service’. ISO, 2010. Law et al. provide up-to-date scoping of the concept and note that it is dynamic, context-dependent, and subjective. See Law, Roto, Hassenzahl, Vermeeren and Kort, 2009. 48 Buxton, 2007; Forlizzi and Battarbee, 2004; Hassenzahl and Tractinsky, 2006; Kuniavsky, 2003. 49 Blythe, 2003; Jordan, 2000. 50 Sharp, Rogers and Preece, 2007. 51 Cooper, 1999; Pruitt and Adlin, 2006. 52 Cooper, 2003. 53 User experience methods explained by Kuniavsky: interviewing, profiling, contextual inquiry, task analysis, card sorting, focus groups, usability tests, surveys, diaries, advisory boards, beta testing, telescoping, log files, customer support, competitive research, published information, and consultants. See Kuniavsky, 2003. 54 Keinonen, 2010. 55 Constantine, 2004. 56 Constantine and Lockwood, 1999. 57 Battarbee, 2003. 45 One

18


logy, and user knowledge have been written ‘between the lines’. Many of the books and thoughts about user-centredness were written for developers in business contexts. Being a developer in a technology company implies a strong understanding of the possibilities and limits of the company’s technological platforms and business model. For these reasons, developer subjectivity along with technological and business values were taken for granted, perhaps under-emphasised, and the combination of knowledge about users with technological and business dependencies were left as an exercise for the reader. Two recent approaches, activity-centred and value-centred design, criticise these and other points in user-centred design. Gilmore et al. raise the question of values, value, and worth in relation to both design and human-computer interaction.58 Friedman coined the notion of value-sensitive design to push self-reflection among designers on what values become embodied in any design result, whether intentionally or unintentionally.59 Bias and Mayhew have long discussed business values and return on investment in relation to usability.60 Cockton attempts to integrate discussions on these different values for end-users and businesses through his approach of worth-centred design.61 Norman re-examined the common user-centred dogma of ‘know your user’ and argued that too much design focus upon individual people might improve things for them, but at the cost of making it worse for others.62 Instead of focusing on individual taste and preferences, he argued for a focus on the activity in which people participate. Constantine also argues for an activity-centred design approach, but talks about designing for use,63 or usage-centredness,64 which could be based on activity theory.65 However, a discussion on the Interaction Design Association’s blog shows that seasoned user-centred designers find nothing new in suggestions promoting activity-centricism, but the change in language might make the design approach easier to grasp for people from other fields.66 These debates show that the positioning of user-centredness, as opposed to system-centricity or developer-centredness, is no longer obvious, if it has ever been so. For communication purposes it has perhaps been a mishap that a focus on users, tasks, and environment (which is what the previously mentioned ISO standards, the early texts about user-centredness, and the context of use models are all about) is called ‘user-centred’. For instance, one seminal book on user-centred design devotes a whole part in the book to user activities.67 The discussions on values, values, and worth reflect an aspiration to frame user-centredness more explicitly as part of a larger picture, namely business and society. 58 Gilmore,

Cockton, Churchill, Kujala, Henderson and Hammontree, 2008. 1996; Friedman, 1997. 60 Bias and Mayhew, 1994. 61 Cockton, 2006. 62 Norman, 2005. 63 Constantine, 2006b. 64 Constantine and Lockwood, 1999. 65 Constantine, 2006a. 66 IXDA, 2008. 67 Norman and Draper, 1986. 59 Friedman,


19

• • •

2.3 A Parallel Debate: Participatory Design 2.3.1 Activist Origins While user-centred design is often presented as a win-win situation,68 the roots of participatory design are based on the perspective of value conflict. The approach emerged in Scandinavia and Germany in the 1970s as part of a cooperation between academics and people from trade unions. As new technology was entering offices and shop floors, management saw this as an opportunity for increased profitability, while workers and trade unions wanted to highlight values such as quality of work and products, democracy at work, and local development. The participatory initiatives included means for resolving these value conflicts, contracts, and organising issues (e.g., engaging workers in the design through local clubs), and an ambition to design the future job first and technology second. Commonly used techniques of today, such as future workshops, prototyping, use models, mock-ups, and organisation simulation tool kits, were pioneered in the 1970s and 1980s.69 Whereas participatory design can make a difference where traditional systems development fail, it too has received its share of criticism. Some claim the early approaches are too coupled with Scandinavian culture, and some argue they are too timeconsuming or focus too much on consensus seeking. Others argue that the initiatives lack scalability and focus too much early stages of projects, forgetting the actual design. Also, framing the design process as conflict and positioning in favour of workers can be detrimental to design contexts, where conflict resolution mechanisms are not as well in place as the ‘democratisation of work’ design contexts of Scandinavia in the 1970s.

2.3.2 What Counts as Participatory? Kensing and Blomberg find three main issues that have dominated participatory design literature: ‘(1) the politics of design, (2) the nature of participation, and (3) methods, tools and techniques for carrying out design projects’.70 The discussions around politics of design have varied significantly over the years, as the conditions for industrial democracy changed and the union power decreased. Some researchers are concerned that participatory design focuses solely on methods, tools, and techniques for individual projects, with too few projects engaged at the company level71 and the national legal and political level.72 This has blurred the boundaries between user-centred and participatory design, and Gulliksen et al. found overlapping themes:73 68 If users share their motivations and aspirations with developers, who come to understand user needs better, users get a better product and developers get the design right the first time, which reduces risk and increases profit. 69 Bjerknes, Ehn, Kyng and Nygaard, 1987; Floyd, Mehl, Resin, Schmidt and Wolf, 1989; Greenbaum and Kyng, 1991; Schuler and Namioka, 1993. 70 Kensing and Blomberg, 1998. 71 Gärtner, 1998. 72 Bjerknes and Bratteteig, 1995. 73 Gulliksen, Lantz and Boivie, 1999.

20


• when and how to involve the user in a design and development process; • practical experience of prototyping and video recording in the analysis, design, and evaluation process; • organisational obstacles to user-centred design; • the role of the facilitator in the development process; and • communication problems that occur when people with varied skills and expertise communicate with each other. It appears that user-centredness does not necessarily imply a great deal of user participation.74 Participatory design, however, has fairly strict criteria of participation:75 1. access to relevant information; 2. the possibility of taking an independent position on the problems; 3. participation in decision making; 4. the availability of appropriate participatory development methods; and 5. room for alternative technical and/or organisational arrangements. Even though participatory design is not an integrated framework for design, some analyses of participation have become more widespread than others. Bossen et al. highlight the following aspects of participation: kinds of people (with respect to role in development, e.g., end-user, manager, vendor, or other stakeholder), type (direct-indirect), degree (sources of information vs. codesigner), duration (procurement, initial design phase, throughout project), and arena of participation (project, organisation, national).76 Today participatory design is being developed in multiple directions. On the one hand, there is a renewed interest in developing yet more methods for hands-on engagement of users and business actors, intended for the concept design phase. On the other hand, several research groups are moving towards research designs with extended timeframes and a continuation of design activities across multiple development cycles.77 This includes, compared to established frameworks such as MUST, more complex collaboration between research organisations and private companies in customised public-private-partnerships.78 79

2.3.3 New Territories for Participation Hagen and Robertson discuss challenges and opportunities for participation in the context of social technologies, which they define as ‘tools and practices that constitute our increased capacity for personal communication, production, publication, distribution and sharing’. Example social technologies are Facebook, Ning, Flickr, YouTube, and Wordpress. They raise the following important topics: complex 74 Gulliksen,

Lantz and Boivie, 1999. and Besselaar, 1993. 76 Bossen, Dindler and Iversen, 2010; Gärtner and Wagner, 1996. 77 Voss, Hartswood, Procter, Rouncefield, Slack and Büscher, 2009. 78 Bødker, Kensing and Simonsen, 2004. 79 Interview with Finn Kensing 1 Nov 2011. 75 Clement

2.3 A Parallel Debate: Participatory Design

21

and variable contexts of use, emergent design, designer role, and the intertwined nature of design and use.80 Complex and variable contexts of use. In contrast with previous participatory design settings from the 1980s and 1990s, the number of both people involved and use situations has increased drastically. This challenges approaches that try to simulate or model future use in advance during the planning process. Isbister and Höök aptly state that there are too many new variables of use.81 Heterogeneity of actors has been a long-standing research theme in, for instance, technology studies, but has received new attention, as these social technologies are more large-scale than previous groupware systems. The anonymity and geographic distribution of users also present new challenges.82 Design is emergent. In contrast with traditional models of software engineering, where development ends with a maintenance phase, participatory design has emphasised that design is completed in use.83 This emergent design in and through use is very visible in service development for social media and related concepts such as perpetual beta. In the hands of users, services such as Facebook, Twitter, and Flickr have transformed from what developers originally intended them for. This emergent property of design has previously been discussed in participatory design under the themes of appropriation, customisation, personalisation, and tailoring.84 Designer role. One of participatory design’s key topics is that new technology and work practices need to be planned in tandem. In the context of social technologies, this translates into planning not only the technical platform, but also the participation of the future community of users. Researchers have noted that communities can take up concerns that have been under-addressed by designers.85 Design and use. Social technologies highlight the intertwined nature of design and use. The short development cycles and long duration of evolving projects enable use to feed into design in unforeseen ways. We can no longer study design and use as separate concerns.86 These challenges and opportunities are revisited in section 2.5 on social media, and the methodological consequences for this thesis are outlined in chapter 3. • • •

2.4 User Innovation Research User innovation research is one of the dominant literatures on user involvement. In contrast with the human factors and ergonomics origins of user-centred design and the political origins of participatory design, research on user innovations has 80 Hagen

and Robertson, 2010. and Höök, 2009. 82 Clement, Costantino, Kurtz and Tissenbaum, 2008; Ehn, 2008. 83 Henderson and Kyng, 1991; Voss, Hartswood, Procter, Rouncefield, Slack and Büscher, 2009. 84 Balka and Wagner, 2006; Mørch and Mehandjiev, 1999; Nardi, 1993. 85 Botero and Saad-Sulonen, 2008; DiSalvo, Maki and Martin, 2007; Merkel, Xiao, Farooq, Ganoe, Lee, Carroll and Rosson, 2004. 86 Williams, Stewart and Slack, 2005. 81 Isbister

22


a background in economics. The main question that this research strand asks is where innovations come from, with special focus on the role of users. Contrary to the dominant belief in economics, research has revealed that a significant part of advances in technology stem from developments and modifications by users,87 not professional researchers or designers.88 The impact of these research findings has been boosted by the visibility of opensource development projects (such as Linux) and peer content creation in Web 2.0 applications (such as Wikipedia). Suddenly user innovation has been widely seen to ‘democratize’ innovation activities, which were previously restricted for technical, scientific, and economic elites.89 Broadening the base of innovation has been argued to increase economic prosperity, bring empowerment and enjoyment to citizens, and cut dependency on the offers of large incumbent companies.90 User innovations are also considered key in regard to a variety of products available to consumers (e.g., web cameras, web mail) and in offering cheaper and free alternative products (e.g., Mozilla Firefox in web browsing, many surgical instruments). For companies, tapping into user innovation is argued to present opportunities to cut research and development costs and a source of adjoining products and platforms.

2.4.1 The Lead-User Methodology User innovation research often talks about a particular kind of user, lead-users, who have two defining characteristics: (1) they are at the ‘leading edge of the market with respect to an important market trend’ and (2) expect to gain ‘relatively high benefits from obtaining a solution to their needs’.91 A lead-user methodology has been developed for companies to complement traditional market research techniques for innovation. The basic idea is that an organisation can invite such lead-users to help and jointly develop products. It is recognised that significant knowledge can be learned from lead-users in analogue markets, markets facing similar problems. The methodology covers different ways of identifying leading market trends, relevant lead-users, and ways of engaging lead-users—for instance, through concept design workshops. The lead-user methodology has mostly focused on identifying and engaging lead-users in developing new products, but a part of the user innovation research has focused on communities around established products—for instance, sports and gaming communities.92 This research strand has also examined how users contribute to company innovation through innovation platforms and so-called user innovation toolkits.93 87 von Hippel defines users as ‘firms or individual consumers that expect to benefit from using a product or a service’, in contrast with manufacturers that expect to benefit from selling a product or a service. 88 Hippel, 1988; Hippel, 2005; Hippel, Thomke and Sonnack, 1999; Lüthje, Herstatt and Hippel, 2005. 89 Benkler, 2006; Hippel, 2005. 90 Benkler, 2006. 91 Franke, Hippel and Schreier, 2006; Hippel, 1988. 92 Baldwin, Hienerth and Hippel, 2006; Hienerth, 2006; Lüthje, 2003. 93 Hippel, 2005; Hippel, 2001; Jeppesen and Frederiksen, 2006; Jeppesen and Molin, 2003.

2.4 User Innovation Research

23

Compared to participatory design, the lead-user approach does not take a stance in favour of democracy or emancipation; rather, it is oriented towards commercial goals. In line with participatory design, the lead-user approach also suggests toolkits for user innovation and custom design. Instead of trying to completely understand all user needs, von Hippel and Katz suggest a strategy of modularisation and flexibility so that users can put the parts together themselves.94 The phrase ‘democratising innovation’, which is the title of von Hippel’s well-known book and therefore associated with this approach, has been criticised for applying a very narrow perspective on democracy, due to the focus on lead-users and user inventors only and no participation criteria.95 • • •

2.5 Social Media: A Distinct Design Context Social media has entered many debates on technology and society. Here, besides briefly giving some examples of how social media is changing business and society, I will approach social media as a particular design context, which is distinct from others based on three socio-technical criteria: type of software business, functionality, and users and use. The emergence of social media has changed the interaction between computers and society again. Previously computerisation movements96 changed warfare, industries, government administration, health care, banking—just about all sectors of society—through active entrepreneurs and innovators harnessing evolving computer-based technologies to be adopted and adapted by both public and private organisations. First, specialised and mainframe computers, then mini and micro computers, computer networks, and related software were taken into use by organisations for different reasons (productivity, democratisation, collaboration).97 This time it is a combination of useful and usable computer-based technologies for consumers,98 services for groups of people,99 business model innovations,100 and active content-sharing users101 that is changing society. Social media has changed the media industry, as new collaborative forms of producing media content and distribution channels have emerged. Gillmor details how webpages, blogs, feeds, and photos from mobile phones change media coverage of disasters.102 Benkler notes that new services tend to keep the group in mind,103 as opposed to earlier computerisation movements. Jenkins draws a 94 Hippel

and Katz, 2002. Ehn and Hillgren, 2010; Hyysalo, 2010. 96 Computerisation movement is a concept by Kling and Iacono that avoids both technological and social determinism, as it considers three components that interact with and shape each other: technological frames, public discourse, and organisational practice and use. See Elliott and Kraemer, 2008. 97 Elliott and Kraemer, 2008; Rosenberg, 1997. 98 Bruns, 2008; Jenkins, 2006. 99 Benkler, 2006. 100 Osterwalder and Pigneur, 2010. 101 Gillmor, 2004; Shirky, 2008. 102 Gillmor, 2004. 103 Benkler, 2006. 95 Björgvinsson,

24


picture of the new media landscape, where media convergence, participatory culture, and collective intelligence shift the boundaries between media producers and consumers.104 Bruns argues that users are able to move smoothly across a participation continuum, stretching from active content creation through various levels of engagement with existing content, and on to mere use of content.105 User-created content became a popular concept in 2005 because of these changes, and has been defined by OECD106 as content fulfilling three criteria: (1) content made publicly available over the Internet, (2) which reflects a certain amount of creative effort, and (3) which is created outside of professional routines and practices.107 A societal change is not only about changes in one industry, but about changes in people’s everyday lives. Shirky gives numerous examples of how individuals with tools for sharing and cooperation join together and get things done without formal organisations.108 Online social networks, such as Friendster, MySpace, and Facebook have transformed the patterns of meeting new people and hanging out with friends online.109 Camera-equipped mobile phones and photo-sharing sites have changed the picture of domestic photography.110 Tapscott and Williams argue that the way Internet changed specific business is spreading towards other sectors, along with the principles of collaboration, openness, sharing, integrity, and interdependence.111 No longer is this only a question of mass collaboration in the by-now-famous Linux, Wikipedia, Google, Youtube, InnoCentive, Flickr, Second Life, MySpace, and the Human Genome Project, but rather a phenomenon across many sectors (Table 2.1). Despite much debate about social media, a common agreed-upon definition has yet to emerge. Kaplan and Haenlein have put forward a suggestion, but one that relies on Web 2.0, which is unfortunately not well defined in the article, nor in general. They consider Web 2.0 as a term that was first used in 2004 referring to a platform based on blogs, wikis, collaborative projects, Adobe Flash, RSS, and AJAX. However, Scholz argues convincingly that definitions of Web 2.0 are vague at best, that Web 2.0 does not form a coherent technological platform, and that the claimed novelty of the associated technologies is false. For instance, the first blogs and wikis emerged in 1994–1995, CSS and RSS in 1998–1999, and social networking sites like Classmates.com in 1995 and SixDegrees.com in 1997. Despite an influential article by Tim O’Reilly’s, an annual conference, and 114,000,000 Google hits, Web 2.0 as a clear set of technologies remains too elusive and fuzzy as a ground for this research.112 104 Jenkins,

2006. 2008. 106 OECD, 2007. 107 This definition is illustrative of the ideas behind the user-created content phenomenon. However, the definition appears a bit shaky when studied in more detail. Take, for instance, (1) Facebook: it could not function without user-created content, but it is not publicly available; (2) Google searches: one form of user-created content that hardly can count as creative effort; and (3) bloggers who make money: a visible portion of bloggers are professional, but in the new profession of blogging. 108 Shirky, 2008. 109 Piskorski, Eisenmann, Chen and Feinstein, 2011. 110 Sarvas and Frohlich, 2011. 111 Tapscott and Williams, 2010. 112 Federated Media Publishing, O’Reilly Media Inc. And UBM Techweb, 2011; Kaplan and Haenlein, 2010; O’Reilly, 2005; Scholz, 2008. 105 Bruns,

2.5 Social Media: A Distinct Design Context

25

Table 2.1: Some advanced online service examples from MacroWikinomics by Tapscott and Williams, 2010.

Domain

MacroWikinomics Service Examples

financial services innovation

VenCorps.com, Prosper.com, Zopa.com, Qifang.cn Local-Motors.com, Ponoko.com, inno-360.com, XPrize.org Earth.Google.com, EyeOnEarth.eu, Carbonrally.com, EarthLab.com, WorldWithoutOil.org, Carma.org, GoodGuide.com, GreenXchange.cc ’open source grid’, theRavinaProject.org, 1bog.org ZipCar.com, BetterPlace.com, GoLoco.org, ZimRide.com, PickupPal.com, Carticipate.com AcademicEarth.org, OCWConsortium.org, RateMyProfessors.com GalaxyZoo.org, EarthSystemGrid.org, PLoS.org, NeptuneCanada.ca/o2, OpenWetWare.org PatientsLikeMe.com, WeAre.Us, FluWikie.com, MDJunction.com, OrganizedWisdom.com HuffingtonPost.com, Craigslist.org, Monster.com, eBay.com Last.fm, We7.com, Rhapsody.com, Spotify.com, Pandora.com, OurStage.com, RiffWorld.com, hitRECord.org, Nettwerk.com Netflix.com, IndieGoGo.com, Hulu.com FixMyStreet.com, mySociety.org, PeerToPatent.org, Scorecard.org, Data.gov, GovLoop.com Witness.org, CorpWatch.org, Crocodyl.org Kiva.org, Sparked.com, NetSquared.org, ReliefWeb.int IndyMedia.org, Twitter.com, Facebook.com, WikiLeaks.org, ’Blogosphere’

climate change

green energy transportation university education science health care newspaper music

tv and film public value citizen regulator global problems fighting for justice

Allen provides an insightful example of how Web 2.0 can be conceptualised, which I can lean on with respect to social media. He departs from the assumption that Web 2.0 is about technologies alone and states that it is about ‘ideas, behaviours, technologies and ideals all at the same time’. He considers Web 2.0 as a conceptual frame with four main elements: technology, economy, users, and philosophy. Technology-wise, he discusses functionality such as data mash-ups and automated collation of data about a user, which ‘enable other users to contact and interact with that first user’. With respect to economy, Allen describes a multisided business model, where the service is free for users, who become an audience for advertisers, who pay the service operator for displaying advertisements targeted to specific users. Concerning users, Allen discusses a ‘media consumer who is more engaged, active and a participant . . . in the creating, maintaining, and expanding the “content”’. Philosophically, Allen states that Web 2.0 is a political statement of a kind of libertarian capitalism, positioning users as ‘relatively equal and equally engaged participants’. None of these four elements is entirely new,

26


but their combination was once relevant (during the dot.com bubble), and the versioning number (2.0) signifies that renewed relevance.113 In analogy with Allen, I outline a few elements—not limited to technologies—for understanding social media as a design context. The point here is to be specific with respect to user–designer relations and contrast social media with other previous design contexts, such as groupware or personal computing, from which some of the principles of user involvement originate. For the purposes of this thesis I outline a set of socio–technical criteria for a social media design context, which constitutes a sensible frame of generalisation for this research (see also section 3.3). These socio–technical criteria for a social media design context are set with respect to type of software business, functionality, and users and use: • Type of software business: Relatively low cost of construction, modification, and distribution, which typically results in considerable development after market launch, as well as unconventional revenue models (see subsection 2.5.1). • Functionality for group communication: Features that support open-ended messages and other user-created content, a collection of groups, and awareness about other users (see subsection 2.5.2). • Users and use: Active user communities (dialogue among users, peer interaction) and peer production, as well as a high degree of voluntary use (that is, not a workplace technology that users are obliged to use) (see subsection 2.5.3). Intentionally excluded criteria are at least openness (of data, sourcecode, or hardware), and software development pace and iterations (the software can develop at whatever pace, the important thing is the typically low initial costs of construction).

2.5.1 Social Media and Industry Lifecycles Online services 10 years ago were different from social media today, as dominant players and designs have been established. Theories of innovation suggest that niche innovations emerge and challenge the regimes in cycles.114 Tushman and Anderson describe how technological breakthroughs ‘initiate an era of intense technical variation and selection, culminating in a single dominant design’.115 While they consider this innovation process to be random, other authors disagree with this evolutionary proposition, stating that the innovation process is neither random nor planned, but something in between.116 Niches can be pushed back by the dominant regime, and only occasionally new types of techno-economic innovations redefine the industry structures in a given sector. When this occurs there is typically an early exploratory phase when both 113 Allen,

2008.

114 Geels and Schot, 2007; Klepper, 1997; Perez, 2010; Tushman and Anderson, 1986; Van de Ven, Polley,

Garud and Venkataraman, 1999. 115 Tushman and Anderson, 1986. 116 Geels and Schot, 2007; Perez, 2010; Van de Ven, Polley, Garud and Venkataraman, 1999.


27

technological solutions and business models change rapidly and feature high variation. The uncertainty and variation gradually diminishes, giving way to a few dominant classes of producers and business models.117 My case is situated in a particular timescale of innovation; I have followed it from niche to mainstream. This location in the innovation lifecycle is important to note regarding generalisations, as there are different needs for user involvement and user research in the early phases of innovation versus later on, when the market is more mature.118 For instance, in a mature market benchmark products and services are easier to find compared to a new market. As described above, social media challenges many different sectors of society. As a consequence, different dominant designs are challenged, depending on which sector one takes as a starting point. For newspapers, the discussion would include technological frames and organisational practices around technologies such as the printing press, offset printing, typesetting, desktop publishing, digital devices, and digital distribution for instance. For software business, the dominant designs or business models stem from the developments of mainframes, mini, micro, and networked computers with related modes of software distribution (e.g., tape, floppy disks, CDs, DVDs, over the Internet). Here, the focus is on software business. Social media shakes established software business models. In the mid–1990s software business was fairly stable around three dominant business models: software contracting, corporate software products, and mass-market software products. Software contractors relied on specialisation, accurate cost estimation, and project management to survive in bidding processes on million-dollar software. Corporate software products became classic capital goods, with high initial development costs and big sales volume, but because of a high degree of mission criticalness and consequential long-term customer relationships, a business model based on quality assurance, pre- and after-sale support emerged. Mass-market software companies relied on ease of use, mass marketing, and big sales volumes.119 On this broad level, one can say that social media software business is a new mix of the above—with varying degrees of specialisation (niche services), big sales volume, after sale support, and ease of use—with the addition of user-created content. In addition, four other trends in software business influence social media service design. First, the instant and low-cost distribution of incremental feature changes to users, commonly associated with software-as-a-service.120 Second, more mature software frameworks that allow for low cost of construction and modification of design. Third, a reliance on advertising within the service itself— as some social media services have gathered large numbers of users, targeted ads have become viable business.121 Fourth, unlike software licenses (and optional support fees) that users pay for, some services sell content or virtual items with virtual currencies.122 117 Geels

and Schot, 2007; Van de Ven, Polley, Garud and Venkataraman, 1999. 1995. 119 Campbell-Kelly, 2003. 120 Campbell-Kelly, 2009; Campbell-Kelly and Garcia-Swartz, 2008. 121 Akera, 2008; Osterwalder and Pigneur, 2010. 122 Haigh, 2008; Lehdonvirta, 2009. 118 Leonard-Barton,

28


2.5.2 Functionality for Group Communication Even though there are no established functionality sets that social media services must conform to, a high-level sketch of common functionality from a user perspective can be outlined. Many social media services are based on a simple idea: users sending messages to other users in an information space. More specifically, here I scope the interest of this thesis to services that support (1) open-ended messages and other user-created content, (2) collections of groups, and (3) a high degree of awareness about other users and groups. The latter two dimensions are important for users’ understandings of other users and community dynamics that shape the developer-user relationship. These characteristics of group communication functionality enable us to explicate how some social media services are similar and different compared to groupware. Groupware,123 also referred to as collaborative software or collaboration technology, became established in the 1980s.124 Despite apparent similarities (to be outlined below), the relationships between social media and groupware have yet to be systematically explored. Traditional taxonomies of collaboration technologies— such as systems for communication, information sharing, cooperation, coordination, and social encounter125 —do not seem to fit social media services, as this new breed of services combine functionality from different types of collaboration technologies. Ongoing research in this direction has also been communicated through concepts such as collaborative virtual environments,126 social software,127 social computing,128 and social technologies,129 but these discussions do not address the aspects of functionality explored here. Open-ended messages and other user-created content refers to one of the two typical approaches to groupware. Some groupware systems were based on strict workflows, while others were based on a context or open space that allowed groups to self-organise. The workflow approach implied applying automation to group processes—for instance, letting the computer system handle ‘paper-pushing’ between people. This approach was based on detailed descriptions of work processes and predefined messages, while the open space approach let the users bring their work processes little by little to the system, which meant that it was more open to changes to the forms of communication and work processes, thus allowing for open-ended user-created content. Key functionality were open-ended mes123 Johnson-Lenz and Johnson-Lenz, 1991, ‘intentional group processes plus software to support them’. 124 Andriessen,

2003; Baecker, 1993; Coleman, 1997; Johansen, 1988. 2003, p. 10. 126 Benford, Bowers, Fahlén, Greenhalgh and Snowdon, 1995; Brown and Bell, 2004, ‘multi-user virtual reality systems which explicitly support co-operative work’. 127 Allen, 2004; Shirky, 2003, ‘software that supports group interaction’. 128 Erickson, 2011, ‘social computing refers to systems that support the gathering, processing and dissemination of information that is distributed across social collectives’. 129 Hagen and Robertson, 2010, ‘Tools and practices that constitute our increased capacity for personal communication, production, publication, distribution and sharing’. 125 Andriessen,


29

sages130 and flexibility in organising group space.131 While groupware systems, such as discussion lists and shared media spaces or workspaces, typically supported interaction between members of one group, current social media services are built for collections of groups. Even though some groupware systems, such as Usenet newsgroups or Lotus Notes, were built for many groups, the inter-group communication was typically based only on setting up a space that was shared among groups. In contrast, some social networking services excel in providing awareness mechanisms, like feeds, syndication services, and other aggregations (e.g., toplists, popular content) of what takes place in all the groups together. This discussion on awareness draws on Gutwin and Greenberg’s concept of workspace awareness for real-time groupware.132 They outline a set of 10 awareness elements that answer basic questions about who, what, and where in groupware systems (Table 2.2 below). These elements are central for supporting group processes in groupware. The questions are tuned for real-time collaboration in one workspace, but the elements could as well be applied to more asynchronous settings and inter-group communication between group spaces. Table 2.2: Workspace Awareness for Real-Time Groupware by Gutwin and Greenberg, 2002.

Category

Element

Specific question

Who

Presence Identity Authorship Action Intention Artifact Location Gaze View Reach

Is anyone in the workspace? Who is participating? Who is that? Who is doing that? What are they doing? What goal is this action part of? What objects are they working on? Where are they working? Where are they looking? Where can they see? Where can they reach?

What

Where

Another key functionality for systems that support large collections of groups is decentralised group space creation. This pattern, where users can create spaces for themselves and their peers, as well as control the access to those spaces, emerged from multi-user dungeons (MUD), virtual worlds, large community platforms (e.g., 130 It is the details and combination of various synchronous and asynchronous message forms that makes a service unique. Herring, 2007, lists 10 medium factors that can be used for analysis of the medium that conveys such messages. . For instance, a Twitter message (a tweet) is asynchronous, transmitted 1-way, persistent, has a message buffer of 140 characters, text-only, anonymous, not private, filterable, and quotable. This medium factor list is open-ended, intended for computer-mediated discourse analysis, and presented along with a list of eight situation factors, related to participants and their activities. Even buttons, such as the Facebook Like button, can be interpreted as such a message. It is asynchronous, transmitted 1-way, persistent, closed content (this user likes that) instead of a message buffer, a button that results in a text, not anonymous, nor private, ignorable, not quotable. However, here the intention is not to dig into small differences between similar services, but to point out broader similarities among services. 131 Andriessen, 2003; Baecker, 1993; Coleman, 1997; Johansen, 1988. 132 Gutwin and Greenberg, 2002.

30


Yahoo Groups, Google Groups), and social networking sites. It is created out of necessity, as the services would not grow flexibly enough, if the developers had to moderate group space creation. When there are many interconnected spaces, mechanisms to move between them and find out what is going on in other places emerge. The nuances in getting these details right are key to virtual worlds building,133 but for the purpose of this outline it is enough to point out that the information space in social media services is made up of group-sized spaces, with paths, hubs, and meeting points between them. Table 2.3 below sums up the functionality that supports (1) open-ended usercreated content, (2) collections of groups, and (3) a high degree of awareness about other users and groups. This list of functionality is shared among many social media platforms for a large audience, such as Facebook, MySpace, and Second Life. This is not to say that this case study does not generalise to other cases without this functionality; it still might do so on the basis of the two other angles presented in sections 2.5.1 and 2.5.3. Table 2.4 lists basic functionality that would have made Table 2.3 too complex: basics about asynchronous collaboration, interconnected spaces (the consequence of supporting collections of groups), and advanced features that relate to moderation and advertising. Table 2.3: Shared group communication functionalities of many social media platforms. Open-ended messages and other user-created content Open-ended messages

A variety of synchronous and asynchronous message forms between users.

Flexibility in organising group space

Choosing from predefined space elements (e.g., various content boxes, space plans) or creating new ones.

World building (optional)

In some services users can participate in more fundamental world building, in others not.

Collections of groups Interconnected spaces

To support a collection a groups, each group needs its own space (see Table 2.4).

Decentralised group space creation

Users can create actual shared spaces (e.g., groups, rooms, forums, shared blogs ) based on configurable plans (e.g., templates) by designers.

Users control access to their spaces

Designers control access to the whole service, but access control is shared with users to the places created and shared by users.

High Degree of Workspace Awareness Awareness elements

133 Bartle,

Interface elements and mechanisms that support questions like who is here, where are my friends, and what are they doing? (see Table 2.2)

2004.


31

Table 2.4: Basic functionalities of many social media platforms, added detail for Table 2.3. Asynchronous collaboration basics Persistent information space

No rebooting, no game over, no reset.

Unique and persistent user identities

Usernames, photo/avatar, profile pages persist between logins.

Micro-movement

In the current location/place (’walking’, scrolling, clicking).

Interconnected spaces User locations

Users are somewhere, not everywhere, in the social media service.

Paths and transportation between spaces

Links, doors, and buttons that take the user from one space to another.

Hubs and meeting points with feeds

Places or lists with multiple destinations, e.g., lists of groups, friends, popular locations, popular content.

Advanced features Moderation systems

Functionality to report and remove content.

Advertising and other in-world media

Ads and other ’media within media’ often require their own publication and moderation system.

Virtual currencies

Some social media services supply service-specific currencies that can be used to purchase goods and services within the service

Fundamental world building

Configurations of user profiles/avatars, virtual assets, group spaces, API:s for apps.

Such a functionality-based angle to social media would focus the attention on a blogging platform, not an individual blog, or on a CMS platform, not a particular website based on that CMS platform, and so on. This functionality table also moves this social media discussion further away from the particularities of Web 2.0 that relate to content aggregation and mashups. It escapes artificial boundaries between social networking services and virtual worlds. Instead it is sensitive to the features that makes scalability possible (decentralised group space creation), links, and those features that make users’ control over content possible. The main two dimensions of this functionality view can be made into a two-bytwo figure, where the social media services most relevant to this thesis can be found in the upper-right quadrant. The shared workspace and virtual worlds are considered to have to most advanced support for realtime workspace awareness, whereas, for instance, Facebook is not as far on that dimension for two reasons. First, real-time workspace awareness is more advanced in most groupware systems, and second, Facebook provides very personalised views to the service, which re-

32


duces the shared views to the service space, which is one key aspect of workspace awareness.

Figure 2.6: Dimensions of relevant social media functionality: awareness and support for a collection of groups.

It is noteworthy that many of the particular characteristics that people recognise in current social media can be traced back several decades: • User lists and profiles: Standard in unix servers since the late 1970s (commands like ‘who’ and ‘finger’ and the .project and .plan files). • Online advertisements: In the mid–1980s, when text-only services were the norm, Prodigy was a forerunner in online advertisements, as the service included a graphics package that was preloaded on the user’s computer so that ads could be presented in a rich graphical environment.134 • Micro-payments: Ted Nelson coined the notion of micropayments for his global hypertext network called Xanadu, in which text authors would retain copyright and get royalties—in other words, be credited with tiny payments for their input as other people read and remixed their texts.135 While Nelson’s system never became reality, AOL got a technically different but businesswise similar idea to work: online transactions just became additional fees put on the customer’s monthly statement.136 134 Akera,

2008. 1982. 136 Haigh, 2008. 135 Nelson,


33

• Liberal and countercultural norms: Turner showed how liberal and countercultural norms were embedded in the design of the famous WELL bulletinboard system, which was based on 1960s American counterculture.137 • Shared spaces for subgroups of users: Research in Groupware and Computer Supported Collaborative Work covered topics such as media spaces, shared workspaces, and collaborative virtual environments.138 Even though much of the social media functionality has its roots in systems from previous decades, those systems were different in scale, degree of automation, and societal diffusion. What is new with social media is how it influences and involves a greater variety of devices and information, a greater degree of automation and algorithms, and a greater number of people and everyday practices. This is manifested in new software business models, the incorporation of group communication functionality within the boundaries of one software application, as well as the increased dependency on users, volunteers, and communities.

2.5.3 Users, Volunteers, and Communities Active users have long been part of the development and use of information and communication technologies. Ceruzzi described how electronics hobbyists and their support networks were central to the diffusion of the first personal computers.139 The early Internet was shaped by pioneers who liked tinkering with technologies in the spirit of the hacker ethic.140 Mods and modders create valuable digital game modifications.141 America Online (AOL) was one of the first to involve a large number of volunteers to monitor e-mail and discussion forums; by 1999 some 15,000 volunteers helped AOL.142 Users are creative and contribute to social media in many ways. So far peer production has stirred up many discussions regarding copyright and the value of user contributions in business and the public sector.143 The aim here though, is to place user-created content and online communities as part of an ongoing interaction between users and developers. Few users think of developers as an audience of their content, since their intended audience is often friends, family, and colleagues. However, the content that users create sometimes also shapes specific features of a product or a service, not only the business model. To study user-created content as part of a user-developer relationship means that it needs to be compared to content created by other stakeholders in development. Designers and users are generally treated as distinct, but interacting groups. Sometimes this is the case, sometimes not. There are case studies in which designers act as users,144 and users act as designers.145 Lindsay’s study on the TRS–80 137 Rheingold,

1993; Turner, 2005. 2003; Baecker, 1993; Coleman, 1997; Johansen, 1988. 139 Ceruzzi, 1999. 140 Himanen, 2001; Levy, 1984. 141 Postigo, 2007; Sihvonen, 2009. 142 Postigo, 2003. 143 Benkler, 2006; Tapscott and Williams, 2010. 144 Beyer and Holtzblatt, 1995. 145 Fleischmann, 2006; Lindsay, 2003. 138 Andriessen,

34


microcomputer shows that users can take over many of the functions of producing and maintaining a technology that developers normally cater for. Twenty years after the original company abandoned the product, users have filled in as designers, producers, marketers, distributors, and technical support.146 This phenomenon can be theorised as role hybridisation. Role hybridisation is about the individual’s ability to shift from one knowledge domain to another, allowing for simultaneous membership in two otherwise distinct social worlds. The traditional power relationship between designers and users often renders designers as omnipotent and users as victims, or at least dependent on designers’ interest in listening to users’ articulation of their wants and needs (section 3.1). The participatory design movement aims to alter this power balance. Studies have shown that developers configure users,147 but also that users configure developers.148 The point being, it is detrimental to both research and practice with focus on designuse relationships to assume (1) an inside or outside distinction between designers and users, or (2) a constant power dichotomy in place.149 Role hybridisation is, however, not enough, since it is not only a question of users and designers. Suchman encouraged researchers and developers to question the designer/user dichotomy.150 The argument is that if we look close enough, we will notice many intermediary actors who make a technical system possible: We need to begin by problematizing the terms ‘designer’ and ‘user’ and reconstructing relevant social relations that cross the boundaries between them. Attempts to avoid this conclusion lead to various sorts of surrogates, proxies, stand-in’s for ‘the user,’ designed to allow the creation of usable technologies in the absence of these other relations.151 Previous research has identified many cases of intermediary actors between users and developers and developed useful nuances among the users (section 3.1). Some users channel information between other users and technical support staff, while others report that experienced users teach less experienced ones.152 Stewart and Hyysalo (2008) argue that user-side intermediaries are crucial to innovation success and show how intermediaries affect the shape of new information and communication technologies. Based on the social learning in technological innovation (SLTI) framework, they open up the designer/user dichotomy by mapping various intermediary positions and different types of innovation contexts (e.g., linear development, off the shelf products, user-centred design, technology experiments and co-design, and innofusion/domestication).153 These discussions have also been framed as design for community, design by a community, or a question of online community. There are two rather different views on users and online communities. Some authors describe online communit146 Lindsay,

2003. 1991. 148 Mackay, Carne, Beynon-Davies and Tudhope, 2000. 149 Fleischmann, 2006; Freeman, 2011. 150 Suchman, 1994b; Suchman, 2002. 151 Suchman, 2002, p. 94. 152 Blomberg, 1987; Nardi and Miller, 1991. 153 Stewart and Hyysalo, 2008; Williams, Stewart and Slack, 2005. 147 Woolgar,


35

ies as an opportunity for people to meet around a shared interest, break free from geographical barriers to social interaction, and as a gift economy.154 Other authors point out that some online communities are commercially engineered and that companies make profit on the hours that volunteers spend online, which renders the first mentioned view as too romanticised and ideal.155 Both views do justice to the studied communities, but they also show how online community is not a homogeneous phenomenon, and like discussions on users, one cannot assume a typical online community power situation. For instance, recent research by Freeman on open-source communities noted a shift from hacker ethic and bazaar governance to a more professionally and strategically regulated community. Freeman also observed how the ambiguous concept of community can become a powerful strategic tool for orienting towards multiple real and imaginary audiences.156 The intention here is to steer clear from normative discussions—e.g., what is a ‘real’ community or what should be the legal status of community contributions— but rather to study the actual roles of users in social media service development, keeping an open mind to both individual and collective actions of users (see the methods sections, 3.1 and 3.2).

2.5.4 Persistent Challenges Accentuated by Social Media Besides generic guides for Internet-related services,157 not many studies have been published specifically about social media in human-computer interaction regarding design processes and user involvement, user experience, and user-centred design.158 Hagen and Robertson pinpoint the challenges to complex and variable contexts of use, emergent design, designer role, and the intertwined nature of design and use.159 Holzapfel asked whether a user-centred design process is suitable for social interaction design, and researched the processes behind 16 highly successful social media services.160 Based on brief interviews and secondary sources, he found that most of the services were not designed using a UCD process, given that such a process includes user research, personas, scenarios, prototyping, and testing (Table 2.5 below). Based on these findings, Holzapfel put forward a number of hypotheses (e.g., ‘Self-centred design is a valid approach’) and contingency factors that he argues decide whether UCD processes are suitable to social media or not. Although cursory, his brief case studies provide an interesting parallel for this study. Holzapfel’s study can be criticised for not covering details about the design process nor the user community, and lacking a change perspective. Services start out as one thing 154 Baym,

1995; Rheingold, 1993; Smith and Kollock, 1999. 2011; Postigo, 2003. 156 Freeman, 2011. 157 User-centred guidebooks for website design by Garrett, 2011; Krug, 2006, community-building by Kim, 2000, e-government services by OECD, 2009, and even iPhone apps by Ginsburg, 2011. Chan has coined the concept ‘social interaction design’ for the design of social media based on user experience and interaction design, see Chan, 2007, but the concept remains unclear for research purposes and is not widespread. 158 Estes, Schade and Nielsen, 2009, provide details about the user experience of messages, notifications, and alerts, but they do not report on the design process. 159 Hagen and Robertson, 2010. 160 Holzapfel, 2008. 155 Campbell,

36


Table 2.5: Summary of case study findings by Holzapfel, 2008.

Service

Was the service designed using a UCD process?

Was the evolution of the early service based on user feedback?

Was the early design determined largely by the founders’ own tastes and needs?

No No No No No No No No No No No No Unknown Unknown No No

No Unknown Yes Unknown Yes Yes Unknown Yes Unknown Yes Unknown Yes Yes Unknown Unknown Yes

Yes Yes Yes Yes Yes No Unknown Unknown Yes Yes No Yes No No Yes No

Basecamp Blogger Craigslist Delicious DeviantArt EBay Facebook Flickr FriendFeed Gaia Online Last.fm Livejournal MySpace Skype Twitter Wikipedia

but develop over time as features, user communities, and development conditions change. This thesis explores three inherent user involvement challenges that designers encounter in their practice: user categorisation, beyond one project, and pathways of user-created content (see below). These challenges are at the same time persistent, as every product and service design needs to address them, but also accentuated as social media has become mainstream.

User Categorisation in Complex and Variable Contexts of Use In a perfect world, everyone could contribute to any technology he or she wanted and all relevant social groups would be heard. However, not everyone has the know-how, access, or motivation to participate in technology production. Nor do the producers know in advance whom to contact or have the resources to contact everyone. Also, because people are too heterogenous, the business reality for most companies operating in broad markets is that they cannot serve all potential customers in that market.161 Instead companies turn to techniques of market segmentation, targeting, and positioning, as well as product differentiation, product mix, and product line management.162 This results in a inherent tensions for user involvement. On the one hand, designers are asked to engage and empathise with real people, but on the other hand, successful business requires selling the product or services to more than one person. This means an inevitable abstraction process, 161 Hippel, 162 Kotler,

2005, Ch 3. 1997.


37

away from the specific settings of unique individual users. Besides the mentioned market and product decisions, user groups and categories are ways of dealing with this abstraction.163 User categorisation is sometimes done explicitly, sometimes implicitly. The most formalised user categories can be found in interactive systems that rely on built-in user models, which are used to provide personalisation of news feeds or product recommendations, for instance.164 Still explicitly written down, but perhaps not as formalised, user categorisations are recommended by user-centred and similar design approaches that encourage the multidisciplinary design team to seek and represent stereotypical users and their practices with, for instance, scenarios,165 use cases,166 and persona descriptions.167 Empathic design is all about steering clear of (1) assuming that ‘everyone else is just like us’, and (2) ‘people are so different from ourselves that we think of them as “them”’.168 Regarding implicit user categorisation, Akrich argued that the user is always present in the design process, not in a physical sense, but that it is impossible to design anything without an idea of how the end result is going to be used. When designers use themselves as an example user, Akrich called that I-methodology.169 The previous sections in this chapter gave an overview on three user involvement approaches, which all have different strategies for one aspect of user categorisation: selecting a sample of users to work with. In the lead-user methodology, users are selected with respect to relevant market trends; in participatory design, with respect to job role and organisational power; in user-centred design, with respect to cognitive and physical attributes or tasks, for instance. What becomes of these strategies in the context of social media? Social media accentuates the dilemma of user categorisation in software product and service development. The number of potential users involved is greater in social media applications compared to many prior software development markets. With the new communication possibilities of social media, a broader part of everyday life than before potentially becomes computer-mediated. With this increased computer-mediation of everyday life it becomes more complex to get an overview of user practices, which means more difficult user categorisations.

Dynamics beyond the Scope of One Project A dominant way of structuring the use of methods to learn about users and codesign with users has been to use project phases more or less tied to the so-called software lifecycle (acquisition and supply; development: requirements analysis, architectural design, and qualification testing; and maintenance-operation). Very little is known about user involvement practices beyond the scope of one project. It is assumed that in the beginning of a project, the design scope is open and there 163 Pollock,

Williams and D’Adderio, 2007, call this ‘generification’. 1990. 165 Carroll, 1995. 166 Cockburn, 2001. 167 Cooper, 1999; Pruitt and Adlin, 2006. 168 Fulton Suri, 2003, p. 52. 169 Akrich, 1995. 164 Allen,

38


is little knowledge about users. At the end of the project, it is assumed that the design scope is closed and that designers know a lot about users. The aim here is to contest and develop the notion of the project as a basis for user involvement guidelines and advice. While the stand-alone project is probably the right way of abstracting experiences for software contractors and usability consultants who work with many different one-off projects, the situation is different for those who operate services for a longer duration of time. One can speculate that, over time, knowledge about users aggregates in an organisation, which would reduce the need for succedent user studies. On the other hand, with social media service success follows a large number of users, which suggests increased complexity and heterogeneity, which would then again increase the need for user studies. For these reasons, the following questions become relevant: How does user involvement evolve over time? Can one rely on informal engagement with users or are formal user involvement methods necessary? Which aspects shapes user involvement method selection?

Pathways of User-Created Content and Contributions The characteristics of social media enable a study of user-created content and its development pathways. Social media is different compared to many other software products and previous studies of user-developer relations. The time period between releases is often counted in weeks or months, compared to years, and many social media services rely on user-created content. This means that user feedback and what the users do have a greater potential to influence a social media service or product compared to other developer–user settings. It also means that it becomes possible to study several iterations of design and use within the timespan of a research project. As a consequence it becomes possible to map out not only what content is created by whom, but different developer reactions to user-created content, development pathways of that content, and boundary shifts taking place between developers and users in who creates what content. In social media applications, the pathways of user-created content becomes a key part of the characterisation of the user-developer relationship.


39

3 Researching Habbo: Case Study Design

This research is driven by two controversies outlined in chapter 1 and chapter 2. The first one is about the vast masses of unusable and complicated software that make our everyday experience of computers less than optimal. A central claim of many who want to improve the situation is that developers lack knowledge about users and use practices. A consequence is the debate on whether and how the innovation process can be democratised. Several remedies have been proposed, prominently mostly within user-centred design, participatory design, and user innovation research. What these approaches have in common is a belief that developers should involve users, the sooner the better. Some argue that a design project must not start coding on day one, but that one should do some (or a lot of) research up front first. On the other hand, current popular software methods in the industry favour so-called agile processes, where coding starts immediately and research comes later, as needed. The question is when and how one should do research on use practices for design. The second controversy is about the nature of social media. Sometimes social media is said to be user-driven, and other times it is said to be developer-driven. Both claims have strong supporting evidence, but there is no evident underlying explanation for how these claims can coexist or be resolved. This matter is of importance to social media development companies and related education. Should companies embrace a developer-driven culture, like Facebook has done,1 or open up the process for user participation, or even go for open design? How can one get answers to these normative questions? The first step is to convert these questions into researchable questions, and the second step is to make methodological choices. To have a basis for shedding light on the above controversies, we need further understanding of the following questions examined in this thesis: • How do users’ actions in and around a social media service shape its design? • How do social media developers’ user involvement practices evolve over time? • How does user categorisation change with social media? I arrived at these questions starting from an interest in the construction of the user (section 3.1). I set out in the field to find out how designers and users encounter 1 Lee,

2011.


41

each other, but as the research progressed I decided to focus on the changes in the constructions of the users and the consequences of these constructions (e.g., changes in user involvement practices and feature development with particular user constructions in mind). The literature review in chapter 2 provided several insights about user involvement practices and the role of users. The following observations of the existing literature are noteworthy and have methodological consequences for this thesis work. (1) There is a bias towards presenting research results in a normative form with unclear connections to actual development practice in the research about usercentred design, user experience, participatory design, and user innovation. There are a lot of guidelines, best practice process guides, teaching books, and ISO standards for development, but it remains unclear to what degree and how these are used in development practice. • Choice: I chose to research real-life software design, and not ponder on the research questions too analytically without grounded research data. (2) Industry surveys of development practice often only scratch the surface. The questions are often formulated as ‘do you use such and such method?’, without finding out the reasons for applying a method, and the level of context-sensitivity is the company, not a particular project. The risk is that particular contextual factors or other practices that perform the same function as the method go unnoticed. • Choice: I chose to start with a case-based qualitative inquiry into software processes, because I wanted to get a deeper understanding of organisational processes and developers’ sources for knowledge about users than is possible with a survey approach. (3) There is a tendency towards studying either use (e.g., research on actual user experience and usability evaluation) or design practice (e.g., methods and processes for controlling IT development or designing for a particular value), not both, nor their interdependence. • Choice: I chose to study both users and developers, as well as intermediaries. This way I could observe simultaneous changes in both use and development processes and structures. (4) There is a tendency towards snapshot descriptions in reports of what happened in a pilot project, but not much is known about user involvement practices in the long run. It is unclear which user involvement practices change over time and in changing contexts. • Choice: I chose to make a long-term commitment to a case. Then it became possible to study several iterations of design and use. These methodological choices are not mainstream in research on software processes, user-centred design, or information and communication technologies in general. However, I draw on recent emerging research that (1) is grounded in real-life case studies on user-centred design,2 (2) is interested in what developers 2 Righi

42

and James, 2007.


do and how they understand users and related methods,3 (3) critically examines the so-called user/developer dichotomy and actual power relations,4 (4) is sensitive to the technologies involved, both regarding affordances and business, as well as changes in them.5 These choices led to an explorative case study, a genre that has a long and respectful history in HCI and related fields in opening up new research terrain. In line with this tradition, this thesis also seeks means from social and behavioural sciences to address so-far unaddressed aspects of HCI, in this case informed by science and technology studies in general and the Biography of Artefacts Framework6 in particular (see section 3.2). • • •

3.1 Users as Co-constructed This section clarifies my understanding of the concept of the user. Designers and users are commonly treated as trivial roles in research on design processes, but the question ‘Who are the users?’ is far from trivial. Depending on who is asking, when and for what purpose, there is a wide range of alternative designations for the users. The ambiguousness of the user has been noted and problematised by a few researchers, and there are several different approaches to studying userelated processes of design and innovation in information and communication technologies.7 Recent user-centred design literature defines the user in two major ways. First, users are defined in terms of their knowledge of computers or of a particular computer program: as novice, intermittent, or expert users.8 Second, a distinction is often made between those who actually operate the computer, the primary users, and the secondary users, who are indirectly affected by the computer system. In addition, Courage and Baxter recommend considering anti-users, those who would not buy or use the product. Hackos and Redish warn against confusing users with buyers, against interacting with surrogate users only, and recommend studying users as members of communities.9 However, these definitions and also the ISO standard definition of user as a ‘person who interacts with the product’10 underspecify the phenomenon. It is left open for researchers and practitioners to contextually agree on whether the user is understood as particular human beings in particular settings, or target groups imagined by the developers. To remedy these 3 Goodman-Deane, John Clarkson, Langdon and Clarke, 2008; Hertzum, Clemmensen, Hornbæk, Kumar, Shi and Yammiyavar, 2011. 4 Eriksén, 2002; Suchman, 1994b. 5 Hyysalo, 2010; Pollock and Williams, 2009; Sarvas and Frohlich, 2011; Williams, Stewart and Slack, 2005. 6 Hyysalo, 2010; Pollock and Williams, 2009. 7 Hyysalo, 2010; Oudshoorn and Pinch, 2003; Stewart and Williams, 2005; Westrup, 1997. 8 Shneiderman, 1998. 9 Courage and Baxter, 2005; Hackos and Redish, 1998. 10 ISO, 1998, product is considered a synonym of service and system in these standards for designing computer-based interactive systems.

3.1 Users as Co-constructed

43

underspecified definitions, I outline five approaches to users: as psychological subjects, social actors, as participants, as lead-users, and as co-constructed. Users as psychological subjects. In studies where a particular theme in psychology is investigated in a human-computer interaction setting, users become psychological subjects. Typical examples are design as applied perception,11 motor behavior models,12 information processing,13 cognitive function analysis,14 interruptions,15 and mental models.16 In these studies, claims of generalisation are typically made regarding most human beings, without distinctions based on social structures and processes. Users as social actors. When identities, lifestyles, and social relationships in multiple social contexts become relevant for design, users are treated as social actors. Lamb and Kling have sought to improve on those user concepts that remain limited to individualistic or cognitive dimensions by reconceptualising the user as a social actor.17 Their view of a social actor is based on four dimensions: affiliations, environments, interactions, and identities. Users as participants. The definitions of users can also be seen as reflecting design philosophies, as has been the case particularly in the computer-supported cooperative work (CSCW) field.18 For instance, Mike Hales discusses different conceptions of users based on different design styles: users as clients (the ‘specify and deliver’ style), users as actor–constructors (the ‘enable and empower’ style), and users as codesigners (the ‘reflect and reinterpret’ style).19 These distinct design styles designate the different ways of users participating in the design. From this perspective, the term user is not about identification but participation in the technology production. The user is seen as a very established category referring to a particular way of participating in technology production. It is not a very favourable position, because traditionally the innovation is seen to flow from the developers via the other stakeholders to the users, and not vice versa. Users as lead-users. In contrast with many innovation theories, research on user innovations have revealed that a significant part of advances in technology stem from developments and modifications by users.20 However, this understanding of ‘user’ is different from the other approaches because what so called lead-users do is defined in relation to specific market trends. Lead-users are said to be ahead of a particular market, where producers and commodities have yet to emerge. However, that means that the fate of the so-called lead-users is undetermined. They can choose to remain users, become entrepreneur-developers, or share their innovation with a company or in so-called user communities.21 The lead-user perspective highlights the ability of users to innovate, but differs from the participatory per11 Ware,

2003.

12 MacKenzie,

2003. 2003. 14 Boy, 1998. 15 Oulasvirta and Saariluoma, 2006. 16 Norman, 1988; Payne, 2003. 17 Lamb and Kling, 2003. 18 Mackay, Carne, Beynon-Davies and Tudhope, 2000. 19 Hales, 1994, p. 155. 20 Hippel, 1988; Hippel, 2005. 21 Heiskanen, Hyysalo, Kotro and Repo, 2010. 13 John,

44


spective both in the assumed power relations and desirable ways for companies to engage with these users. Users as co-constructed. Since the early 1990s, some researchers have shaped an approach that studies the ‘configuring of the user’. The argument is that qualitative research should not take a category such as the user as given, but instead acknowledge the considerable work that has gone into its constitution.22 This implies a distinction between the users as imagined by the developers and the users who actually use the system, including an analysis of their interrelations.23 Woolgar coined the notion of configuring for the process of ‘defining the identity of putative users, and setting the constraints upon their likely future actions’.24 His work is an important theoretical move for studying how users are imagined in computer systems development.25 Williams, Stewart, and Slack warn against the tendency of the early work26 to demonise designers as omnipotent manipulators of users, which they see as a consequence of studying snapshots of design or use processes. They argue that technologies should be seen and studied as sums of many projects, configurations of previous technical frameworks, and never complete. All actors involved at multiple locations need to be considered, as well as their interrelations, while remembering that information about the users is typically incomplete and uncertain.27 To sum up, there is no commonly agreed answer to the question ‘Who is the user?’. I have outlined five different approaches to the user, finding that especially the reconceptualisation of the user as a social actor is appropriate if the focus is either on design or use alone. When users are seen as participators in computer systems development, the point is that not all stakeholders can participate on an equal basis. When users, designers, technology, organisations, and so on are seen as co-constructed, the point is that the distribution of agency, power, and actors is an empirical question. These perspectives share an implicit idea about the ‘user’— that is, only in entering into relationships with technology or its development (either by using or being imagined to use) do people (or imagined people) become users. The notion of the user thus does crucially important work in making groups of people relevant to technology design and pointing to how the ‘user’ or ‘usership’ changes in the course of the evolution of a technology.28 The five perspectives highlight the fact that we need to make the relevant aspects of the user explicit, both theoretically and empirically. Only this way can we understand, for instance, how different aspects of the ‘user’ relate to each other, how they vary in different contexts and through time. In this thesis work I have adopted the users-as-co-constructed approach. In the sections to follow, especially 4.1, 4.2, and 4.4, I do not start from an a priori definition of users, but analyse how different relevant actors use Habbo. section 4.1 outlines how the founders and original Habbo developers at Sulake started out as 22 Westrup,

1997. 2004. 24 Woolgar, 1991, p. 59. 25 Mackay, Carne, Beynon-Davies and Tudhope, 2000. 26 Akrich, 1992; Akrich, 1995; Bijker, Hughes and Pinch, 1987; Woolgar, 1991. 27 Williams, Stewart and Slack, 2005. 28 Hyysalo, 2010; Stewart and Hyysalo, 2008. 23 Hyysalo,

3.1 Users as Co-constructed

45

users themselves, and some active Habbo users handled certain functions that later became the responsibility of the producer organisation: moderation and discussion forums. section 4.2 outlines various practices of categorising users. section 4.4 takes a content creation perspective to Habbo actors, and defines users through their emergent content creation capacities. • • •

3.2 Borrowing from Science and Technology Studies The standpoint outlined in the previous section, considering both users and developers as co-constructed and configured, was borrowed from science and technology studies. In this I join a long row of reflective human-computer interaction researchers who have strived to broaden the field in various ways (as described in section 2.2)—for instance, CSCW and the ‘turn to the social’,29 values in design,30 reflective HCI practice,31 and the ‘X considered harmful’ debates.32 The ambitions have been to surface the often-unstated assumptions and values embedded in human-computer interaction, both assumptions in the research field itself and in the advice and guidance produced to inform design practice. In many cases these reflections have been rooted in the multidisciplinary field of science and technology studies (STS), a blend of social sciences, history, and philosophy, that studies the relationship between society and technology—or put in other words, the interaction between technology and the social.33 Many classical STS papers concern values in design, but although eye-opening, some unresolved issues make the lessons difficult to apply in HCI and design practice. It has been established that artefacts carry/embody a political agenda, which is clearly visible in reproductive,34 military,35 and energy technologies.36 For designers, architects, and engineers this is not news, since they are used to juggling many kinds of values (business value, cost-efficiency, aesthetics, ecological concerns, and build quality, to name a few) and their always messy operationalisations (functional and non-functional requirements, user needs, design goals, constraints, etc.). However, STS studies point to emergent values of technologies that may not be on the design agenda, but rather emerge as unintended consequences. HCI and STS share the same concern that not much is known about the evolution of design agendas of particular projects, since many of them remain business secrets until years later. The lack of knowledge about how values and ‘understandings of the users’ flow through a design organisation and get embodied into technologies 29 Bannon,

1992; Button, 1993; Dittrich, Floyd and Klischewski, 2002; Thomas, 1995. 1997; Gilmore, Cockton, Churchill, Kujala, Henderson and Hammontree, 2008. 31 Dittrich, Floyd and Klischewski, 2002; Dourish, 2006; Sengers, Boehner, David and Kaye, 2005; Sengers, McCarthy and Dourish, 2006; Suchman, 2002. 32 Crabtree, Rodden, Tolmie and Button, 2009; Greenberg and Buxton, 2008; Kaptelinin, 1992; Norman, 2005. 33 Hackett, Amsterdamska, Lynch and Wajcman, 2008. 34 Oudshoorn, 1999. 35 Fallows, 1999; MacKenzie, 1999. 36 Hoogma, Kemp, Schot and Truffer, 2002; Winner, 1982. 30 Friedman,

46


makes it difficult to draw conclusions about steering design and technologies in any direction. Another avenue in which HCI debates about users, user groups, and user involvement can benefit is the STS discussions on classification. One mechanism that is thought to influence the embodiment of values in design is the process of classification and categorisation during design. Researchers have discussed how the categories we use—to talk about activities, experiences, objects, and people (i.e., how we name and classify things)—are critical to understanding.37 Software design relies heavily on abstraction and classification, not only regarding information and database design, but also for understanding use practices. Because it is not generally considered feasible to work with all potential future users, designers strive to include a representative sample of the user population in design. Already this basic step of deciding which users to involve is a matter of classification. Different design methodologies have different ideas for constructing the sample, which is an interesting research topic in itself. Here, though, the focus of section 4.2 is on the categorisation of Habbo users, which is particularly interesting since the sheer number of potential users for social media applications makes sampling difficult. STS has many contributions to technology stakeholder identification and analysis, which is one of key practices of user involvement and certain kinds HCI research. In the 1980s, a framework for the social construction of technology (SCOT) was proposed to understand different meanings ascribed to technology by various relevant social groups.38 The SCOT framework also applied the notion of closure,39 which refers to when the meanings of an artefact stabilise among the relevant social groups. SCOT has been critiqued and further developed, and is not used precisely in its original form today.40 However, SCOT introduced an interesting notation for problems, artefacts, solutions, and relevant social groups that I applied in subsection 4.1.2 regarding Habbo’s first prototype. For the second prototype, some of the artefacts changed, as did many of the relevant social groups.41 However, over the few next years, Habbo became much more complex, which is why I decided to follow the abstractions that the developers made themselves to make sense and keep track of Habbo (see Table 4.1). This table is more sensitive to changes in materialities of Habbo, along with some of the critique towards SCOT, especially from the proponents of Actor-Network Theory.42 While Figure 4.3 concentrates on the patterns of standardisation between different Habbo user communities, that abstraction already conflates many, many user groups. Appendix 5.2 and its summary in subsection 4.1.1 open up the diversity of user practices in Habbo, with the aid of yet another STS analysis framework by Clarke.43 Different 37 Bowker

and Star, 1999; Suchman, 1994a. and Bijker, 1987. 39 Latour and Woolgar, 1979. 40 Humphreys, 2005. 41 The concept of relevant social groups has been criticised, because it is not easy to find criteria for when and why any particular social group is relevant, see Winner, 1993, what makes up a relevant social group, see Hyysalo, 2006, and how relevant social groups are embedded in other social structures, see Klein and Kleinman, 2002; e.g., when did the Mobiles band—crucial for the first Habbo prototype, but less significant for later versions, not to mention current Habbo—stop being relevant, or should it still count as relevant? 42 Latour, 1992. 43 Clarke, 2005. 38 Pinch

3.2 Borrowing from Science and Technology Studies

47

frameworks were explored because of the varying complexity of Habbo at different times. More recent STS theoretical frameworks overcome prior HCI and STS constraints. HCI has long been concerned with design processes before market launch, and few studies analyse so-called post-deployment usability.44 SCOT has been criticised for focusing too much on how different social groups think about an artefact, while neglecting that it must be explained what they are able to do with it.45 Whereas section 4.1 tackles both, the latter point is detailed in subsection 4.4.1. The Biography of Artefacts Framework is one STS approach that overcomes many of the prior constraints.46 It goes beyond conventional focuses on design or adoption processes only, by following ‘the actual packages themselves as they evolve and mature, progress along their lifecycle, and move across sectoral and organisational boundaries’.47 This is similar to what I have done in my case study, as I have followed particular Habbo features from their visionary beginnings through development, user feedback, and subsequent redevelopment. For instance, in 2004–2005, when planning interviews I focused on the following features: The room Navigator, Habbo Pets, Habbo Club, and Habbo Rollers. However, other interview topics were also the development pathways of Habbo from a user perspective, Sulake internal processes, Habbo user communities, and the fansite landscape. • • •

3.3 Benefits of an Explorative Case Study Section 2.5 explored the relationships between software business and user involvement practices. Software business focusing on social media makes a different business context than previous software development contexts. Around 2000, any company starting a social media business had few solid models to lean on. Amazon had already proved that recommender systems worked for online bookstores, the first virtual worlds and massively multiplayer online roleplaying games (Everquest and Ultima Online) had emerged, but otherwise the IT bubble was about to burst. There were, however, a few technological and societal changes that enabled new software business involving ‘cool graphics online’: computers in the home, broadband access, and maturing animation frameworks due to multimedia CD-ROMs, as well as maturing web browser technologies. An explorative case study is an excellent tool for understanding the ferment era of new technology business. Research on technological change in various industries suggests a pattern of cyclical technological change, where the pendulum swings between technological discontinuities and dominant designs.48 One or many technological breakthroughs initiate an era of ferment, which Anderson and Tushman describe as ‘an era of intense technical variation and selection, culminating in a single dominant design’. As the technical variation is broad, with 44 Chilana,

Ko, Wobbrock, Grossman and Fitzmaurice, 2011. 1993. 46 Hyysalo, 2010; Pollock and Williams, 2009. 47 Pollock and Williams, 2009, p. 80. 48 Geels and Schot, 2007; Klepper, 1997; Perez, 2010; Tushman and Anderson, 1986; Van de Ven, Polley, Garud and Venkataraman, 1999. 45 Winner,

48


similar uncertainty of business model and organisation, an explorative case study becomes a good match with explorative business. A theory-testing approach could be feasible on well established research topics; however, the link between user involvement practices and business is an underresearched topic. Also theory-testing is difficult to apply or requires a massive research effort on fast-changing research objects such as those in a ferment era. What does this case represent—that is, what is this case about? Software development in the 2000s? No. There are several, somewhat different respects in which this case is representative: It is a popular service used by millions of users in many countries. It lives up to the widespread characteristics of popular social media services in general. One aspect of the ‘social’ in social media. The company’s social media service is typical in the sense that users have a high degree of awareness of what other users are doing in the service (compared to other kinds of products and services where users have limited access to other users’ use, see subsection 2.5.2). Functions that facilitate open-ended user-created content. The social media platform features the necessary functionalities to support user-created content: persistent IDs so that ownership can be maintained between login sessions, as well as enough spaces that users can control and fill with content (see subsection 2.5.2). Despite comparably minimal content design features,49 Habbo is open-ended and not tightly scripted, which means that users can ‘put the pieces together’ in their own ways. This implies that we can study the results of this creative freedom of the users, and consider whether and how this form of user action shapes the service. Additional means to learn about users. This is about certain aspects in the developer-user relationship in a social media context, more precisely what this context brings to the available means for developers to learn about users. Sulake developers, like other operators of popular online services, can (1) log in to their service like users and check what is going on, (2) rely on the user communities to raise important topics in their online debates, and (3) follow real-time statistics of use, e.g. web analytics, virtual asset sales, user-to-user transactions, and other things that are logged. The developers’ adoption of these additional means can be contrasted with their use of established ways of learning about users, such as interviewing, observing, and questionnaires. Accumulation of data about users. Instead of following one project, this case covered several iterations of design and use, with related user research. Previously done user studies have a big influence both on design and succedent user studies. This allows an analysis of how project phases shape the use of user involvement methods. From self-centred design to design for others. Sulake–Habbo is similar to many other digital startups that have started by developing a service for themselves and their friends, and then later noted that their service has more general value and a broader audience. This shift in development target group, from self and 49 Many other virtual worlds, including Second Life, have features that support more extensive world building, the design of objects, and avatar clothes by users. These features have not yet been implemented in Habbo, perhaps due to the risk of losing simplicity in user interface and the comparably low target group age.

3.3 Benefits of an Explorative Case Study

49

close peers to a broad market, is something that has been below the radar in much theorising in the fields of user experience, user-centred design, and participatory design. Research on user innovations has noted this shift, and it has become noted as one potential development path, from lead-user to entrepreneur, but detailed case studies are scarce. This research has followed in detail what all is entailed both in the company and its user relations in the course of moving through that path. Software as a service and a ‘perpetual beta’ software development process. From a software business perspective, this case is typical to software-asa-service accessible over the Internet. The distribution of incremental feature changes to the users can be made low-cost and fairly instant (see sections 2.5 and 2.5.1). This is different from download-and-install software, shrink-wrap software, and embedded software, for instance. With regard to the design process, the case is representative for small teams who do their thing in a few weeks, where manymonths-long projects are the exception, not to mention years. This implies that there was an opportunity to follow several small feature development micro-cases both from developer and user perspectives. Innovation in the fermentation era. From an innovation lifecycle perspective, this case is about the fermentation era, where technological variation is broad and dominant designs have yet to appear. The results are of value to other startup companies and also to bigger organisations with smaller teams working in ‘startup mode’. Since software is not yet perfect, new ferment eras are bound to happen again and again (see sections 2.5 and 2.5.1). • • •

3.4 Mapping Research Questions and Activities The research started with an activity and continued with another, which both supported all three research questions, namely: • Identify important stakeholders, developers, users, and those in between, e.g., intermediaries. A pilot interview in 2003 with a user revealed the importance of the hotel/community manager, fansite authors, and the volunteer program. Different actors were approached through the following means: interviews, surveys, and online material. • Learn about change in the service. The early history of Sulake headquarters and its digital services was reconstructed based on the first interviews in 2003–2005, but later interviews showed how the service, stakeholders, and practice evolved. These activities are summarised in section 4.1. To answer the first research question—How do users’ actions in and around a social media service shape its design after market launch?—I engaged in the following research activities: • Find out what are the users’ actions. This started with pilot interviews, continued with a survey regarding the visitor profile, and a study of online user

50


forums (details below, section 3.5). This work is reported in articles I, II, IV, V, and summarised in subsection 4.1.1. • Collect examples of users’ active and inactive roles in feature development. In the interviews, developers told stories about users shaping the service. The developers’ point of view was covered there, whereas the user forums and the user interviews told the story from the point of views of the users. This work is reported in section 4.4. To answer the second research question—How do social media developers’ user involvement practices evolve over time?—I engaged in the following research activities: • Learn about online sources. The pilot interview also revealed the following sources for knowledge about Habbo: an archive of the developer-produced online fan magazine (www.kultakalankuvalehti.com), the archive of Sulake’s weekly letter to the community, as well as popular user-created fansites of the time. These are summarised in article I and sections 4.1.4 and 4.2.2. • Learn about change in user involvement methods use. Online sources and various interviews gave a very good picture of the situation in and before 2005. Engagement in development of user feedback mechanisms for Sulake in 2005–2006 deepened that insight. Follow-up interviews in 2007, 2009, and 2010 made it possible to learn what had changed. This is summarised in section 4.3. To answer the third research question—How does user categorisation change with social media?—I engaged in the following research activities: • Learn about Habbo users and consider ways of categorisation. To have something to compare with and be able to learn from the details of the developer practices, I conducted my own user categorisation project with the idea of retracing the steps of the developers. • Find out how designers know about users. This was uncovered by interviews with game developers and succedent follow-up interviews (details below in section 3.5). A key question of the interviews was ‘how do you get user feedback?’. Both of these activities are reported in Article III and summarised in section 4.2. • • •

3.5 Research Process Details In the present study on Habbo Hotel, several bodies of data were gathered over the years, and are presented below chronologically so as to do justice to the gradually deepening access to both developer and user communities that could be negotiated during the study. 1. The project started in 2003 with participant observation in Habbo user communities, pilot interviews, community manager interviews (n = 4), and an explorative survey (June 2004) on the visitor profiles (n = 10,000), as Sulake’s

3.5 Research Process Details

51

first global youth survey was published two years later. I also participated in Habbo, explored the features and their affordances, and analysed default values and users’ degrees of freedom. The visits to Habbo occurred, on average, about twice per month throughout the first three-year span of research, and every second month from 2007 onward (for more on user groups, see articles II and V). 2. Fansites became an important source of knowledge about the user communities. In 2004, 173 Finnish Habbo fansites were identified, and 23 of these that were created for a large Habbo audience were analysed in detail (article I). The size of a fan site varied between five and 50 web articles, and the most active sites had forums with thousands of posts. Whereas the survey findings provided background statistics, the fansites and forum discussions provided insight into active user groups and popular activities and on Hotel history. Since its very beginning, Habbo users have maintained several sites devoted to what has changed in the hotel and how, including kinds and looks of furniture, characters, in-game games, and design flaws. These members’ own documentation has been carefully followed and stored50 throughout the eight years of the research. 3. Throughout the research project, I collaborated with other Finnish Habbo researchers doing thesis work on a Habbo topic. One Sulake employee wrote her humanistic master’s thesis on communication and action in Habbo.51 In addition, two youth workers wrote bachelor’s theses based on their youth work in Habbo,52 leading to an evaluation study by Merikivi.53 The collaboration has provided important secondary sources of Habbo data. 4. In 2005 I conducted thematic two-three-hour interviews with 10 Habbo developers, or about two-thirds of the Habbo game development organisation at the time. Six of the developers interviewed (graphical designers and both client and server developers) had been with the organisation since the beginning, five years earlier, while four developers had about one year of Habbo experience. In conjunction with these interviews, historical materials were collected (e.g., screenshots, access to previous versions, press releases, advertisements, etc.) about the development of Habbo and its predecessors to help construct its development in the years 1999–2003. 5. Having examined the user groups, adjoined sites, and developer practices, the study turned to in-depth interviews with particular users. The users volunteered to participate in interviews in the survey mentioned in step 1. The interviews took the form of two-three-hour individual, pair, and group interviews with 11–16 year-old users (n = 6) and users aged 30 and older (n = 6) to focus on their participation histories, their motivations, the 50 Early on I relied on a software that automatically downloads sites from the Internet called SiteSucker. Later on, instead of downloading whole websites, I limited the capturing to those online texts that I read by using another software in parallel with my browsing, EagleFiler. Regardless of which software I used, text search of the downloaded material was available. EagleFiler had the added benefit of allowing for easy tagging of the downloaded material. 51 Pietiläinen, 2004. 52 Koskinen, 2006; Sihvola, 2005. 53 Merikivi, 2007.

52


meanings they give to Habbo, groups they participate in, and so forth. The interviewees were invited to the office of the researchers, where they could show the researchers what they normally did in Habbo. 6. Habbo has also been a topic for students in various usability research courses at Helsinki University of Technology, with which the first and second authors are affiliated. Two student efforts are noteworthy: a usability test with new 10–12 year-old users (n = 8) in 2004 and a software engineering effort in 2004–2005 to develop a fansite starter kit for active Habbo users. I mentored the second group and conducted a pair interview in 2007 about the playability testing and usability practices at Sulake. 7. After these activities, I had the opportunity to take part in an intervention study with Sulake. For release 9 of Habbo, in 2006, a set of user feedback methods was explored with different stakeholders inside Sulake. This included confidential datasets: database statistics and surveys from two countries on the use of a new feature. In addition to these research activities, Sulake representatives participated in project partner seminars held every six months, two workshops on virtual economy, and many project meetings arranged by the research project. These meetings made informal discussions and a continuous dialogue with Sulake possible. Table 3.1 below sums up these data sources. • • •

3.6 Data Analysis Taken together, these bodies of data provide us with an excellent view of the varying forms of interchange and dialogue between the varying users and developers of this virtual world. The data analysis has proceeded in multiple waves over the years and on multiple levels. Four broad themes can be distinguished: • Habbo Service and Sulake Headquarters History (section 4.1) • Users’ online actions in Habbo (sections 4.1 and 4.2) • Habbo designers’ ways of knowing about users (sections 4.2 and 4.3) • Users’ ways of shaping Habbo (sections 4.4 and 4.1).

Habbo Service and Sulake Headquarters History In contrast with many studies of innovation processes, I had the added benefit of active users documenting the product/service history online. Triangulation and source criticism became key analysis methods. For instance, with the survey data I could debunk user stereotypes found in online texts (section 4.2). Source criticism was key to reading online texts, as, for instance, fansite articles were shaped by both the article author’s and the fansite’s agendas. Multiple histories about Habbo can be told. My account is shaped by the access I had to Sulake headquarters in Helsinki, Finland, the insights I got from users active at a particular time period in Habbo, and the online data available to anyone interested in Habbo with a working web browser. Had I approached Habbo a few

3.6 Data Analysis

53

Table 3.1: Data sources in this thesis. Stakeholders

Methods

Years

Users 11–16 year-olds 30+ year-olds Habbo Finland 2 Habbo countries Online Data Habbo screenshots Habbo videos

Finnish fansites user hackers misc fansites

4 interviews, 2–3h, 6 interviewees, of which 2 volunteers and 1 fansite author 1 group interview, 2–3h, 6 interviewees visitor profile survey, N = 10,000, 6% RR database statistics, and translated survey

2003–2005

∼50 from online participation 14h screencaptures, user interviews 1,5h screencaptures, online participation ∼100 videos on youtube.com online analysis of 23 websites 15 ’crew’ and 2 ’phishing’ websites ∼50 webpages and 3400 discussion posts

2003–2010 2004–2005 2003–2008 2004–2010 2004 2004 2003–2010

2005 2004 2005–2006

Developers (Sulake) founders game developers community managers community director user insight mgr business R&D usability expert documents Online Data fanzine archives weekly letter intro & help pages news & press releases

2 theme interviews, 1–2h; one founder participated in project meetings 8 additional theme interviews, 1–2h 1 interview with devel. director, 1h survey, N = 4, 50% RR 1 pair interview, 1–2h 3 theme interviews, 1–2h 3 theme interviews, 1–2h 1 pair interview w. user insight mgr, 1h 1 pair interview w. usability expert, 2h 1 pair interview w. R&D rep., 2h 5 slidesets, 1 worksheet, 2 reports, 2 books

2005 2003–2006 2005 2005 2004 2005 2004–2005 2005, 2010 2005 2007 2007 2005

21 issues 85 letters 33 web articles 60 web articles

2001–2003 2004–2006 2000–2010

Third parties youth workers students media articles Habbo researchers

2 meetings, 2–3 hours 2 Habbo assignment reports, mentoring 122 articles in main Finnish newspaper data comparison with 3 colleagues

2005–2006 2004 2001–2010 2004–2010

Total interviews interviewees surveys images video website archives

54

24 19 3 ∼10000 images stored, 10% processed 20h ∼600 MB (see Online Data above)


2003–2010 2003–2010 2004–2006 2003–2010 2003–2010 2000–2010

years later and, for instance, from the perspective of a particular Sulake country office, this reconstructed Habbo history would certainly be different in terms of the significance of events, development phases, and possibilities to analyse interrelations on various timescales. As part of entering the world of Habbo I learned about numerous discussions on Habbo authenticity—in other words, opinions about who was the first in what respect, what was important, where are Habbo’s origins, etc. The following became the most important ones: • I first learned about Habbo history as a user entering the Habbo world in the fall of 2003, learning from other users directly and through online texts in the user communities and the fansite magazine produced for users by Habbo developers in 2001–2003. • Through interviews with developers I reconstructed key development projects, design ideas, and what was important when. These were compared with an internal document of the key developments between 2000 and 2003. • Two Sulake-produced books confirmed the ‘grand history’ based on the interviews. In 2004 Sulake published a book called Just Another Day—The Story of Habbo Hotel, with a compilation of the public rooms and furniture collections in Habbo. In 2006 a book called Habbo Käsikirja (Finnish for ‘handbook’) was published, which worked as an introduction for new users and a reference guide. These books compared well with my own lists and screenshots of public rooms and furniture, and included some of the stories of developer–user interaction episodes that I encountered in my own interviews. Informal comparisons of these gave me insight into both interviewee biases and the degree of polished rhetoric found in the published books. • From 2004 on, the user-created fansite Habborator.org has kept track of changes between different Habbo releases. This ‘self-documentation’ has been a convenient foundation for comparisons of findings from interviews and observations. To assess the quality of this self-documentation, I compared it with my inside knowledge of the release details in 2004–2006. I found the self-documentation to be a very representative picture of the officially announced changes and have since had no reason to mistrust this source.

Users’ Online Actions in Habbo A significant part of the research has been devoted to making sense of what Habbo users do in Habbo. The analysis started out with the idea of constructing user groups to answer the simple questions of ‘who are the Habbo users and what do they do there?’, which are common to user research in HCI. Over the years, it turned out that definitive answers to those questions are hard to give. I realized that the question is underspecified and needs more input in terms of ‘why do you want to know’. However, these detailed analyses of Habbo users’ online actions from various perspectives made it possible to ask and understand the answers to more sensible questions, such as what categories Habbo developers use and why. The different data sources have been analysed from the following angles.

3.6 Data Analysis

55

• An initial list of popular activities and user groups was drafted by drawing on fansite articles, established room categories, interviews with users, and Habbo documents by developers (article I), in (emic) Habbo terms. • Demographics, activities outside Habbo, fansite interest, and Habbo background were compared with online activities through a survey, with the techniques of frequency distribution, cross-tabulation, and cluster analysis (article II). • Common Habbo fansite elements, size, and lifecycle were outlined based on content analysis in article I. • Membership categories54 were analysed and grouped into predefined visual categories (created by Habbo developers), emergent visual categories, and non-visual categories (article II). • Habbo group activities were distinguished in terms of open–closed, fleeting– durable, hierarchical–democratic, and visible–underground55 (article II). • Summary of important aspects in Habbo. When writing up article II, we realised that there are no self-evident ways to group all the users for all design situations. There is simply too much diversity in user practices. Instead of ending up with a fixed set of user groups, the article suggested a number of dimensions important for Habbo users, from which succedent user-grouping activities might start. • Based on a particular framework from symbolic interactionism56 we dissected the empirical material in 2007 with reference to nine different actor types: stakeholders of interest, social relations and power structures, discourse formations, norms-regulations-rules, history-tradition-persistence, interactional enablers, communicational forms, business models, and nonhuman actants. (See Appendix 5.2 and subsection 4.1.1). • The interviewed Habbo users were analysed with regard to their Habbo careers (i.e., membership trajectories), which entailed how their motivations changed over time. Article VII outlines these in five stages: first encounters, early position, establishing presence, continued involvement, and waning participation.

Designers’ Ways of Knowing about Users Here I am analysing the answers to my questions to the developers regarding their understanding of the Habbo communities, getting user feedback, and finding out what users expect. Instead of going into details regarding utterances or mental models, I became interested in how these practices changed over time and the change factors. Whereas the first research activities supported insight into Habbo user practices, the interviews with Habbo founders, developers, community managers, and others provided multiple insider perspectives to Habbo. An initial list of ways of getting 54 Baker,

2004. 2004; Strauss, 1978. 56 Clarke, 2005. 55 Bartle,

56


user feedback was drafted after the theme interviews with Habbo developers in 2005. Differences among developers in the forms of interaction with users were noted and interpreted in connection with work role (e.g., graphics designer, user interface programmer, server coder) and years of experience at Sulake. In 2005 it was evident that user involvement practices had changed a lot since market launch, but it was not until follow-up interviews in 2007, 2009, and 2010 that I could assess the stability of the practices over time. To explain and make sense of the changes, different phases of Habbo’s service evolution were outlined (article VI and section 4.3). In addition to coding the interviews in themes, the descriptions of developer responses and their changes were systematically compared across different interviewees. This triangulation57 provided a relatively coherent picture of the methods used in different times even as there were variations to which methods different professionals used. In 2006 and 2008 Sulake published two market surveys on Habbo activities, which provided additional sources of comparison. For instance, there was considerable overlap between my cluster analysis (subsection 4.2.1) and the lifestyles in the 2006 survey. These reports concerned Habbo demographics and pop culture in different countries, and compared the Finnish hotel with the other hotels.

Users’ Ways of Shaping Habbo In this part, the analysis has been fairly detailed to catch the nuances and avoid too broad generalisations about users’ role in social media. Fairly uncommon ways of getting insight into users’ ways of shaping Habbo are presented: a detailed analysis of user-created content, how developers use it as design inspiration, and how users sometimes perform functions normally associated with developers. One subanalysis of what users do in Habbo is the question of what content users create. Content is here understood as any user action that the Habbo system gets some kind of record about. This varies between moving around in Habbo, chatting, decorating rooms, organising events, and so on. Of particular interest was to understand how user-created content diffuses in Habbo, which is why an analysis focusing on content visibility (ranging from private to a user to public in a Habbo hotel) and the temporal dimension (only temporary content vs. more lasting content) was made (subsection 4.4.1). In addition to what content users create, one can ask how well it fits the intentions of developers. Articles I and II already showed that users are creative and bring in their own themes and backgrounds to Habbo. In articles V and VII we analysed in detail two examples of tweaking and subverting Habbo: gothic play and playacting horse girls (subsection 4.4.2). A slowly paced dialogue between developers and users developed through the Habbo material. During the research I collected examples of how users developed an idea starting from something the developers made. Sometimes developers picked up these ideas suggested by users and implemented them as features in the hotel. Different pathways of user-created content are outlined in subsection 4.4.3. 57 Denzin,

1970.

3.6 Data Analysis

57

Some users are more active than other users, and sometimes certain users take over functions usually assigned to developers. Section 4.4.4 discusses these boundary shifts between developers and users by examining the role of fansites in community management. • • •

3.7 Notes on Quality and Ethics Due to the diversity in the forms of qualitative studies, widely accepted quality standards and criteria have yet to appear.58 Instead, many researchers have moved towards explaining their strategies for managing quality and how they reduce the risk of researcher bias.59 In the following paragraphs I will elaborate on the quality strategies adopted here, e.g., triangulation60 and communicative validation,61 as well as related ethical concerns. In terms of data and method triangulation, I benefited from ‘naturally occurring data’ (fansites, Sulake documents, books, and reports) and complemented it with a mixed methods approach with multiple data sources (interviews, surveys, and participant observation), as presented in the previous sections (3.5 and 3.6). Investigator and theory triangulation took place as I collaborated with Toiskallio (study design, data collection, and early analysis in 2003–2005), the Privacy-in-theMaking project (2006–2010 see Appendix 5.2) and pursued different theoretical perspectives (e.g. consumption, subcultures, and symbolic interactionism) with article co-authors. These triangulations allowed me to sustain the search for appropriate social science research methodologies to capture the changing forms of user involvement as service and user relations evolved, as well as to pick up particular emerging issues. Purposive sampling was used as emergent insights from one data source guided the sampling of the next data source (section 3.5). For instance, before doing the game development interviews in 2005, I had prepared by following the user comments on feature changes in 2003 and 2004, with the idea to get the inside story in the developer interviews. So I did, and this inside perspective continued strong until mid–2006, when the MC2 research project ended and access to developers was renegotiated. The following multiple purposive sampling strategies were employed: (1) Game development interviews: one of Sulake’s co-founders nominated 10 developers 58 Flick,

2009; Seale, Gobo, Gubrium and Silverman, 2004. 2007. 60 The concept of triangulation was first presented by Denzin in 1970 as a way of increasing validity, by combining theories and methods in the study of the same phenomena. Through various criticisms the concept has been developed and there is consensus that one need not assume that different methods represent the ‘same’ object. Triangulation does not necessarily increase validity nor objectivity, but a combination of theories and methods develops a fuller picture and a more in-depth understanding than using a single method. The resulting prolonged engagement with the research site(s) also provide researchers a chance to reflect on their own biases. Denzin, 1970; Flick, 2009. 61 Communicative validation refers to conversation about the findings with interview subjects, the general public, and the scientific community of scholars. Such dialogue with relevant stakeholders can reduce the risk of researcher bias and provides an opportunity to assess the credibility and plausibility of interpretations. Communicative validation is not unproblematic, as both confirming and critical comments can be a result of disinterest or power issues. Flick, 2009; Kvale, 1995; Kvale, 2007. 59 Flick,

58


for researcher interviews to be representative with respect to different job roles and age in the organisation. (2) Other Sulake people were suggested to us based on both our task to develop user feedback methods and our research interests: community moderation, user research, playability, and usability evaluation. (3) User interviews: the pilot interview was based on convenience; user moderators as typical representatives of volunteer moderators; and the rest of the interviewed users were selected to represent typical age groups in Habbo. (4) User-created content: typical examples. (5) Fansites: biggest Finnish fansites aimed at a large audience, official and other big international fansites, and Habbo hacker fansites. The data was analysed on multiple levels: careers of individual users and their reference groups, processes in the producer organisation, service and development headquarters history, careers of particular fansites, pathways of particular features, and user contributions (section 3.6). The analysis also spanned several iterations of design and use. Between 2003 and the end of 2010, there was more than 60 releases62 with several feature changes per release, which gave plenty of opportunity to follow specific feature iterations. Communicative validation was practiced in the process of article writing and with interviewees during interviews to increase the credibility and plausibility of the findings. Insights from fansite discussions and prior interviews were discussed with interviewees in subsequent interviews. Article drafts were sent to Sulake and were developed based on the received comments before publishing. Dialogue with the research community was important, as research findings were presented and developed in both research group meetings and at conferences.63 Internet inquiry raised additional considerations to quality and ethics, especially concerning field site boundaries and privacy.64 Some of the early internet studies drew strict boundaries between what takes place online and offline to support the argument that the internet can function as a cultural context, but recent studies increasingly see the internet as an integrated part of everyday life and treat it methodologically, not only as a place, but also as a tool or way of being.65 However, as Hine notes, the field site is an emergent construction that depends on research interest and the phenomenon under study, which implies that for some research purposes a discussion forum is a valid field site in its own right, for instance.66 This thesis aims to provide an account of the user–developer relations, where the different articles put different pieces of the puzzle together: Articles I, II, and IV focus on online activities, whereas articles III and VI discuss development practices, and articles V and VII study particular user subcultures. With respect to the privacy of research subjects, interview situations and surveys are fairly established settings, whereas the privacy expectations in online settings can vary considerably. Elm suggests that online environments can be treated as a continuum on the private–public dimension, where chat rooms and web pages 62 Habborator.org

has kept track of the release details. conference papers and presentations are either drafts of the articles included of this thesis, or explorations of a particular topic: Johnson, 2006a; Johnson, 2006b; Johnson, 2006c; Johnson, 2007; Johnson, 2009; Johnson, 2010; Johnson and Tamminen, 2007; Salovaara, Johnson, Toiskallio, Tiitta and Turpeinen, 2005a; Salovaara, Johnson, Toiskallio, Tiitta and Turpeinen, 2005b. 64 Markham and Baym, 2009. 65 Baym, 2005; Hine, 2009; Livingstone, 2008. 66 Hine, 2009. 63 These

3.7 Notes on Quality and Ethics

59

that do not require registration are seen as public, while online communities and social networking sites that require registration are seen as semi-public. Online environments and chat rooms protected with access codes are seen as most private.67 However, this is a complex matter, since (1) people have different understandings of the privacy of spaces, (2) private discussions sometimes take place in public spaces, and (3) computer technologies mediate social interactions in unexpected ways.68 In this research, all interviews and surveys were reported maintaining the anonymity of the interviewees and respondents. The online survey data was treated according to Finnish national legislation on privacy protection, which includes storing the data safely as well as writing a privacy policy for research aims, sharing it online, and implementing it. Publicly available online data, that is, fansite discussions that any internet user could participate in, were treated as public and quoted with sources identified. Research data from Habbo use sessions was treated as more sensitive and only screenshots were published. Screenshot size in print was considered to make it difficult to identify avatars. Age became an important consideration in this study in three respects: interviewing minors, age difference in online encounters, and interviewing developers of the same age. Standard practice in youth research requires parental permission for engaging with minors as research subjects. For the user interviews, written parental permission was requested and the online survey was clearly marked as conducted by a research organisation, not the service operator. Parental permission is fairly difficult to acquire for normal chat situations, which is why I did not conduct formal interviews online, but rather asked users in person to show me what they do online and spent time in Habbo as a user and learned about the use experience. The discernible age difference between the author and the majority of the service users made it difficult to chat with unfamiliar users, since my polite and honest answers to inquiries about my real age often led to a waning discussion. As a consequence, I belonged to the crowd in the service that, for varying reasons, tended to participate in more scripted play situations (role-play, furniture trading, and user-created games) or hang out with their own friends, not so keen on socialising with unfamiliar users. The small age difference between the author and the interviewed developers shaped the interviews, as did other overlaps in background knowledge. For instance, when discussing video games that played an influence in concept design, references to Commodore 64 games from the 1980s were easily communicated. Other shared reference points were software business, programming practices, languages, and tools. This meant that interviewees could communicate certain aspects of their practices through stating the names of technologies (e.g., servlets, Macromedia Director, version control) and only develop the description further if they had applied the technology in a way that differed from standard software engineering practices. With these words on method I hope to have provided a sense of the degree of robustness of this study. The next chapter gives an overview of the insights and key themes of the articles that form the basis of this thesis. 67 Elm,

2009.

68 Bylund, Johnson, Lehmuskallio, Seipel and Tamminen, 2010; Lehmuskallio, Tamminen and Johnson,

2008.

60


4 Key Themes of Article Contributions

This chapter summarises the article contributions into four key themes. The first theme gives an overview of how Habbo emerged, developed, and evolved in dialogue between developers, users, and other stakeholders. The second theme relates to the research question of investigating social media developers’ user categorisations and how they change over time. Both of these themes (sections 4.1 and 4.2) include fairly broad details from the case study in order to make the reading easier for the reader not familiar with the Habbo context. The third key theme—relating to the notion of project as a basis for user involvement guidelines and advice—raises the abstraction level and moves the focus from the user communities to company practices and their evolution over time. Perhaps the strongest and main arguments of this thesis can be found here, in section 4.3. The last section again brings in more detail about the service, with the aim of giving the reader a deep understanding of the analysed user-created content and contributions. To sum up, after the service evolution section in 4.1, each succedent section concentrates on and answers one research question, as follows: • 4.1 Emergence and Co-Constitution of a Social Virtual World: Baseline regarding service evolution and change for the succedent sections. • 4.2 Challenging the Categorisation of Users: How does user categorisation change with social media? • 4.3 Beyond the Scope of One Project in User Engagement: How do social media developers’ user involvement practices evolve over time? • 4.4 Pathways of User-Created Content and Contributions: How do users’ actions in and around a social media service shape its design after market launch? • • •

4.1 Emergence and Co-Constitution of a Social Virtual World This section provides an overview of the life in Habbo, how it came about, who and what were involved in its creation, and how it has developed during the 12 years since the first prototypes. Subsequent sections deepen the themes introduced


61

here. Special attention is given to changes in the business context and in the userdeveloper relation.

4.1.1 User Practices in Habbo What the users do in Habbo is described in many of the articles (II, IV, V, and VII) from different points of view, but a brief summary deserves a place here. The means for communication in Habbo are very similar to other established social media services, but the combination of different channels and the forms of presentation are unique. Under the unique surface, the communication is based on synchronous and asynchronous communication patterns researched in the fields of CSCW and CMC, such as chat, messages, discussion forums, location-based notes, and user profiles (subsection 2.5.2). The difference compared to media spaces and shared workspaces described in the prior literature is that Habbo is not one shared location, but a collection of many virtual rooms. Habbo is not that unlike so-called multi-user dungeons that were invented in the 1970s, but with the addition of a graphical user interface in a pixelated cartoon style (Figure 4.1).

Figure 4.1: Idyllic image from an online ad for Habbo by Sulake in 2006. Even though Habbo started out as a graphical chat and grew as users created more hotel rooms, Habbo is not so far from social networking services1 either. All Habbo users have a Habbo homepage, where they can describe their interests and profile. As Habbo users make friends with other Habbo users, they can formalise this relationship and become Habbo friends in the service through a virtual handshake with a few mouse clicks. Being listed on other users’ ‘list of friends’ enables users to better communicate with each other and see each others status and whereabouts in the service. As discussed in detail in article VII, virtual worlds and other social media are not so easy to distinguish based on feature typologies,2 since features once considered typical to one kind of service have spread to others as well. This necessitates an increased sensitivity to process and contingency, how activities in virtual worlds are open-ended and not predetermined by game rules.3 1 boyd

and Ellison, 2007. and Aarseth, 2007; Messinger, Stroulia and Lyons, 2008. 3 Malaby, 2007. 2 Elverdam

62


Habbo activities take place either in the meeting places created by Sulake or in rooms and collections of rooms created by users. Users create virtual hotel rooms based on a set of floorplans and decorate the rooms with various pieces of furniture, either acquired through trading or purchased from a Habbo furniture catalogue. The activities vary from child’s play to imitating and recreating familiar cultural narratives from sports, TV, and online dating. The diversity of Habbo actors and activities became apparent early in the research process. The stakeholder identification and analysis process became an ongoing challenge instead of something completed early on. In the same way that the diversity in Habbo was a challenge for this research, it is a continuing challenge for user research at Sulake. For instance, are there any obvious categories of users to design for, given that designers cannot interact with everyone? This is the topic of section 4.2. Appendix 5.2 provides an overview of the stakeholder analysis based on the insight of the research activities. With the aid of an advanced framework based on the research tradition of symbolic interaction,4 it was possible to distinguish nine main categories of different actors (see Appendix 5.2). Stakeholders of interest were divided into individuals (e.g., current and previous Habbo users, their parents, and Habbo founders), collectives (e.g., Sulake, its owners, competitors and business aliences, organised groups in Habbo), and implicated/silent actors (e.g., siblings, significant others, friends, school classmates, and celebrities). The number of users and their demographics were already described in the case description (section 1.3). However, if one considers particular individuals relevant to the service, it is worth noting that prior Habbo users not active anymore, and current Habbo users’ parents have a significant influence on the service. The parent is often the paying customer and regulates the teenagers’ access to the service. Prior Habbo users left many trails in the system, for instance in the forms of invented Habbo games and taken avatar names. These different stakeholders talked and wrote about Habbo in markedly different ways. Habbo was presented as a hotel, a game, child’s play, a chat, a community, a dating place, a place for collecting and trading furniture, cool graphics online, a crime scene, means for advertising, and so on. These different framings of Habbo shape the expectations of how Habbo is understood and the consequences of entering or supporting Habbo. Habbo also appeared as different objects when connecting it with different relevant business models, revenue streams, and funding alternatives. User payments and advertisements were obvious in the context of social media, but the history of Habbo also revealed experimentation with business models, such as Sulake customising the game engine as subcontractors for Coca-Cola and Walt Disney, or making merchandising deals with Hollywood. Venture capital, or the lack of it, shaped the Sulake organisation during the whole period, which was very visible in the downsizing maneuvers in 2007 and 2009. Different stakeholders have different power in relation to the technologies of Habbo, rooms in Habbo, and the regional or national Habbo communities. Certain developers can change the Habbo architecture substantially while moderators and 4 Clarke,

2005.

4.1 Emergence and Co-Constitution of a Social Virtual World

63

hotel managers can influence the local hotel configuration, whereas users create content (more details in section 4.4). Early on, there was not much specialisation among the developers, but over the years specific roles emerged, such as project manager, art director, IT architect, graphic designer, user interface programmer, and server coder. Likewise, in the Habbo communities several different pathways for gaining fame emerged (see articles II, IV, VII, and subsection 4.1.4).

4.1.2 From Concept Design to Beta In the early Habbo days, the hotel was developed by a handful of game developers with core competencies in graphic design, Macromedia Shockwave Flash clients, and Java server programming. At first they developed the hotel called Kultakala (‘goldfish’ in Finnish) for themselves and their friends, but just a year after launch it became popular among teenagers. Through the internationalization of the hotel, the organisation grew, and every country got a local office with a few employees working on moderation, community management, customer relations, and marketing. More administration and business personnel have joined, and since 2005 game development involves more than a dozen game developers. Habbo was based on two previous launches, Mobiles Disco (October 1999) and a snowball game called Lumisota (February 2000). Mobiles Disco provided the basic hotel infrastructure in the pixelated ‘retro’ style: rooms, easy navigation from room to room, the chat interface with speech bubbles, and avatar customization and movement inside the rooms. For Lumisota, the developers implemented paying through cell-phone messages, which at the time was a practical solution for the Finnish market. This worked well also for the Finnish Hotel Kultakala (August 2000). Figure 4.2 below shows the stakeholders involved in the first prototype, Mobiles Disco. The figure is drawn with the notation of Pinch and Bijker,5 developed to highlight how relevant social groups interpret artefacts differently and the development pathways of problems and solutions. This figure shows how solutions similar to those that the co-founders developed for new media companies, first at To The Point Oy6 and later at Satama Interactice,7 became the basis for Mobiles Disco, which was a solution for both Sulake co-founders and their friends’ band, called Mobiles. Mobiles Disco became a proof-of-concept for ‘cool animations online’ by the Habbo founding developers. The next project, Lumisota, was made while the founding duo was working at advertising agency Taivas and had a customer, the Finnish Internet service provider and mobile phone operator Elisa/Radiolinja. 5 Pinch

and Bijker, 1984; Pinch and Bijker, 1987. The Point Oy was one of the prominent new media companies in Helsinki during the 1990s. It was founded in 1991 and attracted creative people with experience in visual arts and computer graphics, of which some were still in secondary school. The company worked with multimedia and digital communication, such as graphics and animations distributed on information kiosks, multimedia CD-ROMs, and websites. It has been characterized as the first new media ‘school’ or local society in Helsinki. Mander, 2001. 7 Satama Interactive was one of the first serious Internet consulting and design companies in Helsinki. This dot-com company was formed in 1997 by a merger of three smaller digital media agencies. It employed several hundreds of people and made a long-standing impact in the new media and digital technology service landscape in Finland. Pelkonen, 2005. 6 To

64


Figure 4.2: Social groups, artefacts, problems, and solutions in the concept design of Mobiles Disco, based on the notation from Pinch and Bijker, 1984.

Some of the stakeholders had changed—the customer for instance, but some stakeholders continued from Mobiles Disco to Lumisota: the founding duo and some users, including active user volunteers. When the next development, Hotelli Kultakala, was launched, the number and diversity of stakeholders increased. Both strategic and local advertising partnerships were formed and a number of payments mechanisms were implemented. Mobile phone features were added. Development became a more complex matter, as the spreadsheet from one of the founders (Table 4.1) shows. This table shows the important dimensions (rows in the table) of the service to keep track of between late 2000 and early 2003. Hotelli Kultakala had transformed into Habbo Hotel, a name invented for the UK launch. Some developments were country-specific (marked with country code if known and ‘-’ if unknown) and some more generic (marked with a ‘*’).


65

Table 4.1: Habbo History Squeezed, anonymised version of Sulake co-founder’s spreadsheet from 2003.

2000 Q3 Launches

2001

Q4

FI

Q1

Q2

UK

Q3

Q4

Q1

** *

* *

* *

User sales enhancements / CRM

**

UK

*

*

** *

** * UK

UK

UK FI

UK

UK

FI

Q1

FI UK

*

UK UK UK UK

FI

UK

* FI *

UK

*

UK UK

- UK

UK UK

UK

Technical Improvements

FI

-

UK CH UK, FI

Consumer Marketing

Payment Systems - SMS -billing - Youth cards - Other

Q4

UK

Strategic Partnerships

Mobile Phone Features

Q3

JP

FI

FI

Q2

2003

CH

New game features

Local Advertising Partnerships

2002

UK

FI

UK

UK

-

--

---

-

---

--

4.1.3 Patterns for Expansion With the UK hotel also followed the teen invasion8 of Habbo. It was not designed for teens at first, as illustrated by the developers making furniture sets with bar desks as integral elements and one of the new public rooms being greatly influenced by stereotypical English pubs. The large number of teen visitors also meant a large number of concerned parents. To keep all this together, community management received a lot of attention. A set of guidelines for good behavior was established, called Habbo Way, and tools for governing it were developed. The moderators got their own interface to Habbo, which meant that they did not have to be in the same room to follow a discussion, and a text filter was developed to screen for swear words. The customer service got an automated response system to reduce the amount of personal service needed. In 2002–2003 Sulake developed its business, the hotel, and the community simultaneously. Strategic partnerships were made, and brands such as Mountain Dew and Britney Spears entered Habbo, the former as a vending machine serving Mountain Dew bottles and the latter as a large poster to be hung on walls. Later these 8 In my interviews, developers described the change in target group (and their own marginalisation) as teenagers invading the service.

66


brands disappeared as Sulake tried another approach by developing Coke Studios for Coca-Cola and Walt Disney Magic Kingdom for Warner Cinema. These were customized versions of the technical game engine behind Habbo, which helped fund the development of the hotel. The hotel’s technical architecture was stabilized, and security improvements were made to package the hotel into a product that could be more easily rolled out in new countries. During 2004–2005 around 10 new hotels and local offices were established in different parts of the world. Habbo actively wanted to become the largest teen brand in the world. The English-speaking UK hotel was split into separate hotels for the US, Canada, and Australia. Also, game development recruited more developers in 2004, and new features were packaged into simultaneous releases in all hotel countries. The rapid expansion was possible because a ‘recipe’ for managing a hotel and entering a new market had been developed. Figure 4.3 below sketches the ideal configuration that Sulake strived for. The figure is developed based on my understanding of the organisation and operations around 2005 and 2006. It also shows a feature in different hotel contexts and for different stakeholders. The figure shows what is centralised (Feature Development and Marketing & Sales) as well as what is duplicated in every user community. At that time, the idea was to launch a Habbo hotel and managed user community in each new country. The country office would take care of the local technical configuration of the hotel, community management, player support, local campaigns, and advertising. To function, the hotel needs to be run on a server and payment mechanisms need to be established. Userside intermediaries (volunteers and fansites) need to be catered to, and the parents of the users need to be convinced about the safety of the service. Other intermediaries also play an important role in Habbo, e.g., from youth workers and other NGOs to the consumer agency and ombudsman in certain hotel countries. In contrast to the small company where everyone did everything, specialisation had already occurred. Developers specialised in graphics, user interface design, server functionality, or project management. Ads and campaigns were separate from community operations and user insight. The image intends to show that centrally developed features end up in many similarly structured hotels in various countries with various stakeholders and many user groups. This structure of Habbo separates it from many social media services developed later, where the technical structure is often more centralised and different language versions appear within the same service, not as separate instances of that service. The figure also shows the considerable amount of work (e.g., organisations, country-specific techno-economical infrastructures, and software packaging9 ) needed to make features flow as ‘immutable mobiles’10 between countries. 9 In the early stages of Habbo, there was different code on different servers in different countries. Through redevelopment and modularisation, software packaging became possible to standardise feature rollout to the Habbo hotels in various countries. 10 Latour, 1987.


67

Figure 4.3: Centralised feature development and typical stakeholders of Habbo user communities in 2005–2006.

68


4.1.4 User Intermediaries: Fansites and Volunteers Innovation studies have pointed out the importance of intermediaries. Bessant and Rush identified four generic roles of intermediaries: knowledge transfer, knowledge sharing, brokering, and user needs diagnostics.11 Howells (2006) and Stewart and Hyysalo (2008) have deepened insight into the role of intermediaries.12 In the history of computing, the crucial role of intermediaries such as hobbyists is well known.13 In online business, AOL can be seen as the forerunner, at one time managing more than 15,000 volunteers.14 Even though Habbo started out with a fairly low number of actors and a close relationship between developers and users, over time the number of actors has grown and more intermediaries have emerged. On the supply side, the first major step was the establishment of the UK country office, with UK specific hotel/ community management and customer support. But how did the user-side intermediaries landscape develop? Here we sum up the role of fansites and volunteers. Fansites. Every now and then groups of active Habbo users team up and write Habbo-themed websites in the form of blogs, online magazines, or discussion forums. Taken together, these so-called Habbo Fansites15 make up one of the most interesting things in Habbo for five reasons: their role (1) in Habbo communities, (2) among Sulake developers (subsection 4.2.5), (3) in user research as a data source (subsection 4.2.2), (4) in Sulake’s community management strategy (subsection 4.4.4), and (5) in highlighting how functions in the Habbo social worlds have shifted between developers and users (subsection 4.4.4). This subsection, based on article I, explores the role of the fansites in Habbo communites. Fansites emerged around all Habbo Hotels in their respective countries or language regions. They vary in size and temporality, from small sites with a few web pages that operate for a few weeks to the biggest fansites with hundreds of thousands of page views, readers in more than one country, and that operate for many years. While most fansites remain fairly underground phenomena, the more popular ones can get recognized by Sulake as being an ‘Official Habbo Fansite’. In our Habbo survey from Summer 2004, about 50% of the respondents visited some fansite at least once a week. The survey showed that the fansites were at the time broadly known among Habbo users, and many visited them as often as the hotel. Fansite builders write typically for a large audience, complementing Sulake’s official site by providing more detailed information about the hotel from an experienced visitor’s point of view. Hints, secrets, and guidelines, and stories about Habbo fashion influence the boundaries for acceptable behavior in Habbo. In addition, some fansites provide a discussion forum, either for all Habbos or for a specific group in Habbo. Thus, the fansites serve four important functions: they complement the official website, strengthen the governance policies of the developers, reproduce and reinforce social positions (like potential Habbo career paths or legitimized visitor groups), and improve the Habbo visitors’ awareness of 11 Bessant

and Rush, 1995. 2006; Stewart and Hyysalo, 2008. 13 Campbell-Kelly, 2003; Ceruzzi, 1999; Friedman and Cornford, 1989. 14 Postigo, 2003. 15 Fansite is a blend of fan and website, like fanzine is a blend of fan and magazine, a nonprofessional and nonofficial publication produced by fans of a particular cultural phenomenon. 12 Howells,


69

the fan cultures around Habbo. Based on interviews, the fansite builders are motivated with getting approval, getting friends, doing things for fun, using writing skills, and becoming respected. In addition, analysis on format and content of the fansites shows that fansite builders want to connect themselves with the whole Habbo community, be able to influence the ways other people behave in Habbo, provide a place to debate important topics of Habbo, and be able to link avatar names to real persons via fansite content (although not much appreciated by Sulake). The fansites are not only important to the visitors, but they also play a considerable role in the game development. Developers visit them regularly to follow what’s going on, reading both articles and forum discussions. Some more active fansite readers among the developers send e-mails to the other developers about interesting articles. Through the fansites developers learn both what users expect and which topics are not discussed, which is important for prioritising features for new releases. However, since not all users visit fansites, it requires careful judgement from the developers’ side, and perhaps triangulation with other data sources, to develop an understanding of how representative user comments, requests, and wishes on the fansites are. The fansites have always been key to Sulake’s community managers, who maintain close relations with the biggest fansites. In the early days, fansites performed the roles that Sulake did not have the resources for. As the company grew, Sulake incorporated some of the fansite functionality into its Habbo website, such as discussions forums. Since the addition of user forums to Sulake’s Habbo service in 2006–2007, the number of Finnish fansites have been in decline. Not all of the Habbo-themed websites created by users are positive towards Habbo and Sulake, which creates a delicate community management dilemma (more about these topics in subsection 4.4.4, Fansites in Community Management). Volunteers. For the first five years of Habbo, Sulake leaned on volunteers to moderate the online activities. A volunteer was called ‘Hobba’, whose function was to mediate conflicts, send warnings to misbehaving Habbos, kick them out of the hotel rooms, or ban them from the hotel. We interviewed two Hobbas together (theme pair interview 2 hours) and according to them one motivation for becoming a moderator is to take responsibility for the community. Since they are there anyway, they can handle some cheaters as well. Other motivations mentioned are that becoming a Hobba is one way of ‘climbing’ socially in the community, and to become famous. In a fansite discussion forum, in a thread about questions to the Hobbas,16 WildChild asked why the Hobbas became Hobbas: WildChild: Why did you moderators become moderators? Isn’t it quite tough? Garon: It’s fun of course. No moderator would be a moderator if it wasn’t fun. Ferquz: What do hobbas do in their free-time in Habbo? Nakki: They are idle and they chat. Nerola: Also for Hobbas being in Habbo is free-time... 16 Translated

70

from Finnish to English by the author. Nerokala.com, 2004


In 2005–2006 Sulake brought the moderating function in-house by employing moderators in its country offices. The volunteer program changed, and experienced Habbo users could apply to become ‘Habbo eXperts’, who did not have moderating powers anymore, but could get into a room that was full. In 2008 the volunteer program changed again, and eXperts became Habbo Guides, who volunteer to welcome new users and explain Habbo’s features. In 2009, Guide ‘Bots’ were introduced, answering basic questions about Habbo. This reflects Thomas Hughes remarks about technological systems: ‘Over time, technological systems manage increasingly to incorporate environment into the system, thereby eliminating sources of uncertainty, such as a once free market’.17

4.1.5 Service Evolution As we saw in the previous subsections, what Sulake–Habbo consists of has changed significantly over the years. Habbo started as a pet project for a few developers and their friends, grew to become a popular online world among new media people, and within a few years it became mainstream for a teenage target group. Technical, economical, and organisational bottlenecks were solved so that the service could grow and scale up to become a transnational service. Figure 4.4 below is a developer’s drawing describing this service evolution from 2005. Sometime during 2006 and 2007, the strategies changed again as Sulake entered the social networking market. Not only did Habbo get new social networking features, but Sulake also acquired IRC-Galleria, the most popular social networking service in Finland at the time. As IRC-Galleria users were a little bit older, on average around 20 years old, the strategy to concentrate on teens was opened up to a variety of age groups. The rapid expansion of 2004–2005 stopped and Sulake concentrated on BRIC-countries such as Brazil, Russia, and China. For the first time, in 2007, the organisation was forced to make downsizing maneuvers, as apparently the organisational growth was not sustainable. Figure 4.5 shows the launches and closings of Habbo hotels. In 2008–2010 Sulake regrouped internally, invented more automation behind the scenes, and introduced a new way of managing transnational hotels. All Englishspeaking Habbo hotels were merged into one hotel as of mid–2010. This can be seen as a response to a more competitive market, as witnessed by launches of several new sites for teenagers18 and the increasing popularity of Facebook. Based on this information, I grouped the service evolution into five stages. Table 4.2: Habbo Service Evolution. Concept

Beta

Expansion

Complexity

Competition

1999–2000

2001–2003

2004–2005

2006–2007

2008–2010

Concept refers to the first prototypes in 1999 and 2000: Mobiles Disco, Lumisota, and Hotelli Kultakala. At this time, the development resources were minimal, as the two founding developers created the first proto in their free time after work 17 Hughes, 18 KZero,

1987, p. 53. 2009.


71

Figure 4.4: Developer’s drawing of Habbo evolution, 19.4.2005.

72


Figure 4.5: Habbo Timeline, based on factsheets at Habbo fansite Habborator.org, 2011.

and during weekends. With the first prototypes money could be raised, the Sulake company was formed, and more developers were hired. It was at this concept stage that the inventions that inaugurated Habbo were created: clever use of browser client-server technology by Macromedia; custom-built rapid animation development tools; and the revenue model of selling virtual items and micropayments via mobile phone text messages. Habbo also got its technological style; it was made to look retro with an air of playfulness through developers’ anti-game-developmentpattern design choices. Beta refers to the time period between 2001–2003 when much of the basic functionality was completed. Internationalisation started through a UK partnership, followed by a Swiss partnership. Intermediate projects with a Brazil radio station and Coca-Cola provided more funding to continue development. These projects were also experiments with the business model—that is, selling the so-called software engine behind Habbo with rebranded graphics to Coca-Cola and later Walt Disney (as Walt Disney Magic Kingdom). In 2003, about 50 people worked at Sulake. In this beta stage, in Hughes’ terms,19 more conservative inventions were made, such as solving scalability bottlenecks. The Habbo technology and its accompanying organisational pattern was transfered from the country of origin to other Habbo countries. Expansion refers to 2004–2005 when the product was packaged so that it made a rollout possible in more than 10 new countries during one year. Before that different code was used in different countries. Earlier software development had made possible synchronized feature releases in different countries. Release management 19 Hughes,

1987.


73

and code centralisation emerged as a development strategy, in contrast to ‘different projects and different code’. In this expansion stage technology transfer was made more efficient. Several reverse salients20 had been identified, of which the web browser client’s memory was particularly limiting growth, as memory limited the variety of different pieces of furniture in the service. The client had been optimised earlier, but required continuous tuning, and later, a way of dynamically loading furniture into memory was developed. Complexity refers to 2006 and onwards, when the product was extended to a social networking service. These four phases were presented in article VI, but based on feedback from Sulake in 2009 and 2010 I felt it necessary to add a new phase: competition. It reflects the increased amount of social media services for children, and that teenagers increasingly adopted Facebook. The details regarding user involvement is outlined in subsection 4.3.4, where new data-driven development techniques replaced some earlier forms of developer engagement in user communities.

4.1.6 Summary The findings in this section provide the basis for the next sections. There are many things that can be said about the evolution of Habbo, but concerning user involvement, I will highlight three notable findings here. (1) The wide variety of user practices and stakeholders involved challenges old models of user involvement. There is no single reason for why users use Habbo, nor a single thing that the users do in Habbo. Instead Habbo appears more like a mini-Internet of its own; there is something for everyone in the target group. In the following two sections I will argue that this increased complexity makes it challenging to follow traditional processes of user involvement—for instance, deciding which users to engage in development. One emergent model of the Habbo developer-user collaboration could be termed ‘online community journalism’.21 The developers of Habbo encourage Habbo users to report and write article about important topics in Habbo. Instead of the developers having to find out what matters most to the Habbo users through expensive user research, developers can rely on users doing the work. The role of qualitative inquiry is then transformed to being a means for source critique of what the users write about Habbo. More about these observations in the next section. It appears that seeking good representations (various notations, process flow descriptions, and other transformations of field data for design) for users’ actions and activities become less important in computer-mediated services such as Habbo, as the virtual hotels of Habbo themselves already function as adequate user rep20 Reverse salients are components in the system that have fallen behind or are out of phase with the others; Hughes, 1987. 21 Thanks goes to David Frolich and the Bespoke Project for this connection. http://www. bespokeproject.org/

74


resentations. Anyone can log in and check what is taking place. (More in the next section.22 ) (2) Even though Habbo can be considered ‘new’ software business, parts of it behave like other technical developments. In various social media debates, there is a tendency of simplification and reduction, to overestimate the role of one actor, be it users, developers, a particular business model, or advertisers. It is clear that many f(actors) shaped Habbo: not developers alone, not users alone, not technology alone, not business alone. The evolution of Habbo appears to have followed similar pathways as other large-scale technical systems.23 Variations: Early on it was not clear that Habbo was the concept that was going to make it, as Habbo was but one of many concept variations produced with the same technologies. Closures and reopenings: The for many years stable concept of Habbo was reopened as social network features were added in 2006–2007 and later again when making Habbo a Facebook app. Environment incorporation: Moderation and discussion forums shifted from being something provided by users to being provided by developers. The design space was not opened completely for every little development project, but rather when radical changes were planned. In section 4.3 I will argue that the degree of design space openness, among other factors, is more relevant to user involvement guidelines and advice than whether it is the beginning, middle, or end of a project.

(3) The ‘social distance’ between Habbo developers and users has increased. The developer–user relationship was very different in the beginnings of Habbo compared to Habbo three, five, and 10 years after market launch. Early on, the developers built the service for themselves and their friends. Gradually, more and different stakeholders became relevant to Habbo and various gaps between developers and users emerged: organisational, geographical, age difference, education, and so on. At first these emerging gaps were bridged by developers’ active participation in use communities, and volunteer users participating in development, such as moderation of the service. However, later these bridging activities became more rare. From having been insiders in the user subcultures, the developers found that they did no longer know what was ‘in’ among users. They had become big brothers and sisters of the users, then the responsible adult—excluded socially and culturally, albeit not technically. These changes in the developer–user relationship can be interpreted as an increase in social distance. Here this increase in social distance refers to an increase in uncertainty and unfamiliarity of the other group’s practices, resulting from a combination of (1) increased diversity in use practices, (2) an increase in differences between developers and users, (3) decreased developer participation in use prac22 On the other hand, if over time the diversity of user practices increase, which means that logging in and checking online activities covers a smaller part of the whole, then again aggregate representations become more important. 23 Hughes, 1987.


75

tices and vice versa (decreased user participation in development practices), and (4) increased indirect contact between developers and users through both social and technical mediators.24 The first mentioned aspect, increased diversity in use practices, seems fairly inevitable for successful social media services that facilitate open-ended user-created content, as the number of users and the amounts of local contexts where the service is used increase. However, this case highlights the latter three aspects shaping social distance: difference, participation, and mediators. The adoption of the concept of social distance between developers and users contributes to the understanding of various gaps between developers and users.25 This case study shows how the social distance between developers and users can be dynamic and contingent even within the same case. It implies that the developer– user social distance is not an inevitable gap and is more dynamic than, for instance, Norman’s and Nielsen’s writings suggest.26 Some earlier use of social distance and difference, such as in discussions on ethnicity and race,27 are based on fixed boundaries between inherently different groups. In contrast, this construction of social distance is relational, contingent, and variable. In this sense it is closer to the notion of distance employed in social network analysis, which often is based on interaction frequency and network topology.28 It is a contextually contingent question whether the social distance is manifested in symmetrical or asymmetrical social relations, which in turn could mean either exploitation or empowerment of users, for instance. Social distance here is not foremost a matter of taste and class, as in Bourdieu’s theory of distinction.29 The use of differences between developers and users here is open to multiple axes of differentiation. In the following section, I will argue that this increased developer-user social distance has consequences for the categorisation of users, a key activity in user involvement. • • •

4.2 Challenging the Categorisation of Users When I started the research on Habbo, key questions on the minds of many were: Who are the Habbo Hotel users? And what is so important in this virtual world that one quarter of the Finnish 10–15 years-old population likes to spend time there regularly? At the time of writing, in 2011, when Facebook has half a billion of users and regional social media around the world come and go, no one asks precisely this question anymore with such curiosity and intensity. It is clear that social media services have become communication media among other forms of mediated communication. Asking this question has become as relevant as asking 24 Mediators in Latour’s sense: ‘Mediators transform, translate, distort, and modify the meanings of the elements they are supposed to carry.’ Latour, 2005, p. 39. 25 Grudin, 1991. 26 Nielsen, 2008; Norman, 1988. 27 Bogardus, 1925; Park, 1924. 28 Granovetter, 1973. 29 Bourdieu, 1984.

76


who the telephone users are and what is so important on the telephone lines that people talk so much on the phone. Not that asking who the telephone users are would be a silly question; it is still important to reflect on how media changes our everyday lives and which forms of communication come to reign when, where, and in which situations. Being a telephone user, say, a hundred years ago, would imply significant social status, and an exploration of the demographics of telephone users at the time would probably have made social stratification and particular social networks visible.30 However, the point I wish to make here is relevant to specific practices in user involvement, product design, and service development, namely the practice of categorising users. Having read this chapter, the reader will be able: . . . To categorise users.31 One of the first steps of user-centred design and many other forms of user involvement is to identify and categorise users. This first step is proposed because it is considered unrealistic to work with all (potential) users during product lifecycles. Therefore it is deemed necessary to define the characteristics of the user population and work with a representative sample of the user group. Different user involvement approaches32 have their own strategies for selecting the sample: • lead-user methodology: select users with respect to relevant market trends • participatory design: select users with respect to their job role and power attributed to people in similar job roles. • user-centred design: make selection based on people’s cognitive and physical attributes like technology expertise, skills, and demographics, or their tasks and other factors in the context of use model. The question is, what becomes of these strategies in the context of social media? As social media brings change in design-use relationships, developers need to know when and where to adjust their process to meet the new demands on production and collaboration with users. How can large amounts of simultaneous users be supported technically and socially, as well as represented in design? One way of approaching this problem area is to study who the social media users are and what they do there. This is the topic of article II. It gives an account of my explorations of the question of user categorisations. The article presents user categorisations based on data from participant observation, a survey, user interviews, and fansite articles.33 The idea at the time in the research project was to trace the steps of Sulake developers in understanding users, use the additional research resources available, and contribute with sociological insights.

4.2.1 Survey Data: Habbo Visitor Profile We (Johnson and Toiskallio) started out by exploring Habbo and conducting pilot interviews, ending up with a few preliminary categories: furni collectors, chatters, late-evening party people, mafioso, sheriffs (volunteer moderators), and cheaters. 30 Fischer,

1992. and Baber, 1999, p. 17. 32 More detail in chapter 2. 33 Data described in detail in section 3.5. 31 Noyes

4.2 Challenging the Categorisation of Users

77

The second attempt to find suitable user categorisations was the visitor profile survey (June 2004). Table 4.3 shows the model of the Habbo visitor that I together with my colleague Kalle Toiskallio came up with based on our preliminary interviews and participant observation. Table 4.3: Model of the Habbo Visitor. Theme

Survey questions

Traditional background Other activities Online Habbo activities

Age, gender, region School, hobbies, other games, friends Chatting, furniture, decorating rooms, creating games, making friends Favourite, visit frequency, reading & writing, forum discussions Age in Habbo, visit frequency, network connection, logon place

Habbo fansites Habbo background

These were the dimensions of the user model that the survey was based on. The survey gave us data on how these dimensions were interrelated, e.g., correlate. Based on the survey, we were able to explore a number of different ways of categorising users: • Age in Habbo: We were able to identify a learning curve of about three to four months, after which knowledge and certain opinions about Habbo stabilise. • Visit frequency: The data suggests that those who visit Habbo often are more likely to arrange events, trade furniture, and meet with Habbo-friends outside Habbo. Also, those who visit Habbo rarely are more likely not to want people in their rooms, not to have heard about gangs nor fansites, and not like spending time in their own rooms. Cluster analysis gave some correlations between demographics and activity in Habbo. Table 4.4 shows six clusters that explained 72% of the survey data. Table 4.4: Habbo User Groups Based on Cluster Analysis. Cluster

%

Oldtimers

15%

Playmakers

14%

Silent majority

15%

Gangmembers

11%

I don’t pay Older people

16% 1%

Characterising quote ‘We can log on when we want to, and have been in Habbo forever’ ‘We like to visit often and arrange events for others’ ‘We don’t want to be disturbing or in anyone’s way’ ‘I like to spend time in Habbo with my regular gang’ ‘I don’t pay for anything in Habbo’ ‘We just want to help keeping this place nice’

However, these clusters were based on all variables, which meant that the background variables (e.g., gender, age, etc.) took over. For instance, when we tried to 78


create seven clusters instead of six, the next emerging group would have been an all-girls group. To remedy this issue, we decided to analyse background variables separate from Habbo activity variables. An analysis suggested that two dimensions regarding online Habbo activities were more distinguishing than others: 1. privacy – publicity • wants privacy for oneself and/or friends vs. wants publicity, to become famous 2. organiser – participant • arrange events in one’s room vs. doesn’t like spending time in one’s own room The survey was also used to debunk categorical statements about boys and girls in Habbo, such as ‘boys collect all those furni there, and girls chat’ (pair interview 10 Jun 2004). We found that a majority of both boys and girls trade furniture in Habbo. Although we were satisfied with the survey and what we learned from doing it, the survey still left some questions open. First, we targeted the survey to answer the question ‘who are the Habbo users?’, not what motivates them to go there. Second, the survey was successful in describing a snapshot of the Habbo community, but community research tells us that the motivations for community members to participate change during their career in the community (see article VII). To answer these questions we studied Habbo fansites and interviewed Habbo users (sections 4.2.2 and 4.2.3). Third, when doing the explorative survey we did not fully understand the diversity of all the Habbo activities going on. During our analysis we learned that a Habbo ‘event’, as we called it in the survey, can be broken down into many dimensions. For instance, the event (or series of events, making it more like an activity) can be described in terms of large or small, long-term or temporary, visible or underground, open or closed. Furthermore, the group activity leading up to the event can be hierarchic or democratic, valuing uniqueness or competition, and the purpose of buying furniture and decorating the room varies from just for fun to becoming rich to imitating real-world games. The point being, in the survey, where people answered that they would create an event to get people to visit their rooms, we lacked data on what kind of event they were talking about. Our fansite studies made us wiser in the above respects, as the next section shows.

4.2.2 Learning from Habbo Fansites In 2004, when I started to analyse Habbo fansites, it was already evident that Habbo was not the only online service with fansites. Fansites had emerged in the US around the single-player game The Sims, where one could create virtual characters and manage their careers and lives. A website called The Sims Resource had a fansite index, where the top four fansites had more than 5 million hits. The fansites offered the latest information about The Sims, including news, articles, reviews, and lots of extensions and modifications to the game for downloading.34 34 Sihvonen,

2009; TSR, n.d.


79

Since fansites studies were only beginning to emerge, the research question was fairly general at this stage: Are they useful as sources for user research? We did not focus on technical issues behind the websites, but rather on two subquestions: (1) What kind of content can be found on the fansites? (2) What can we learn about user groups and popular activities? I identified 173 Finnish fansites, of which 23 fansites aimed at a large audience were selected for closer analysis. Content analysis revealed common fansite elements, which are presented in Table 4.5. Based on the fansite content, it was possible to distinguish eight different user groups and 11 popular activities. Based on this research activity, Johnson and Toiskallio (article I) drew the conclusion that fansites are useful sources for user research, as they complement other sources informing information systems design. Table 4.5: Common Habbo Fansite Elements (Article II). Fansite Elements News and rumors

Participation Links Hints, secrets, guidelines Reviews and lists Histories Fashion and celebrities Graphics Habbo fiction About Real life

Description Fansites are convenient for Habbo visitors who want to reach a large audience, a fast way of spreading information about Habbo happenings (e.g., competitions, pop idols visiting Habbo), new features, news about Sulake. The fansite audience is provided ways to comment on the fansite through discussion forums, guest books, polls, etc. The fansites link to relevant Habbo places: other fansites and the hotels in other countries. Fansites teach newcomers both basic and advanced tricks with which to impress others. Guidelines on acceptable behavior are frequent. The fansites keep track of the features and possibilities in Habbo: public spaces, different furniture items, pets, etc. Two major histories are told on the fansites: the history of Habbo and the history of that particular fansite Habbo ’journalists’ interview Habbo celebrities, avatars who have become famous in Habbo, and report on fashionable clothing and activities. Edited screenshot pictures are an integral part of many fansites; some even provide pixel graphics drawing schools. A few fansites write fictional stories about Habbo characters. Who comprises the fansites staff, number of visitors, updates, banners, etc. Habbo meetings ’in real life’, stuff not about Habbo that is important to teenagers, as well as blogs, e-cards, etc.

Based on the fansites, we clustered the hotel residents into eight groups: (1) furniture traders and collectors, (2) chatters (in public rooms), (3) gang-members and VIPs (insider groups not open to everyone), (4) supervisors with administration powers, (5) cheaters, (6) quiz-makers and players, (7) the hotel manager (a Sulake employee), and (8) celebrities. Similarly, 11 popular activities were identified: trading furniture, casinos, dating, beauty contests, competitions, dice games, team sports, formula tracks, talk shows, clubs and hotels, and orphanages.

80


More important than the exact details of these listings are two observations about Habbo that they convey: the diverse and commonplace qualities of Habbo. First, there is not one particular Habbo activity that attracts all Habbo visitors, but many different ones. Second, the activities going on in Habbo resemble games with rules and pretend play familiar from schoolyards, playgrounds, youth clubs, and so on.

4.2.3 Interviews with Habbo Users After the pilot interviews, the survey, and the fansite analysis we continued with interviews35 with Habbo users. A few dozen persons, of which 12 actually showed up at the researchers’ office, were recruited through the survey to participate in focus groups. The interview situation was arranged so that we had a computer with access to Habbo. This meant that the interviewees could show us their Habbo rooms, where they normally go in Habbo, and what they do there. During the interviews we learned more about the Habbo users’ ways of categorising other Habbo users. For instance, we learned about • different clothing styles: punk, gothic, teenie, wannabe, own style • Habbo professions: some Habbo visitors pretend to have a profession in Habbo, which involves decorating their room and behaving according to the selected profession. Some fansites have written articles on this role-playing phenomenon and discussed the following professions: journalist, nurse, TVshow host, bartender, pharmacist, actor, police, doctor, fireman, postman, veterinarian. • attitudes towards Habbo furniture: some find collecting and decorating an end in itself, some as a means for something else (props for role playing, showing status, etc.), and some were indifferent to furniture. • visitors from other hotels: those who speak another language. • time of day: daytime (children with flu at home, mothers), after school (preteens), evening-night (older, best discussions). During our interviews, we tried to find out whether the users would identify themselves with one particular user category or Habbo activity as presented above. None of the interviewees really accepted being labelled as one particular category, as they all mentioned that they started out doing certain things in Habbo and then moved on to try other things. They claimed that they got bored of playing the same game or keeping the same room activity after a few weeks. When going through all these different ways of describing the Habbo users, the diversity of the Habbo uses is striking. Answering the question of who the Habbo users are with one particular categorisation does not seem fair. Neither is a simple two-dimensional model enough. What we could do, however, was to provide a summarising list of aspects that are important in Habbo (Table 4.6 below). It is possible to distinguish different user categories based on all of the dimensions in the list. Depending on the motives behind the question, different categorisations can be made. The list can also function as a map for discussing a Habbo career 35 Interviews

described in more detail in section 3.5.


81

Table 4.6: Summary of important aspects of Habbo use (Article II). Aspect one’s own avatar(s) one’s own rooms and furniture Habbo homepage friends play

Habbo career testing boundaries and rules

Description clothing styles, character description collecting, trading, decorating, browsing the furniture catalogue one’s avatar’s homepage that is visible to anyone on the web. school, hobbies, new friends, dating, distant friends beauty contests (popularity), TV shows, games of chance, Habbo-sports, insider clubs, roleplay, playing with the spatiality of the virtual world celebrities, getting rich, popular room, in a game or gang, being a fansite author, being a Habbo guide expressing self, treating others (e.g., cheating, bullying), finding and using glitches in the hotel architecture

with a Habbo user: which aspects drew the user into Habbo, which aspects made the user come back, and so on.

4.2.4 Interviews with Developers The next step in our research project concerning user categorisation was that in 2005 I did 10 thematic interviews with about two-thirds of the people in game developement at Sulake. During 2005–2006 five shorter interviews regarding user feedback were made with representatives from development, user insight, business, and community management. In 2007 I followed up with a pair interview about usability and playability testing with users together with the usability specialist and a representative from R&D. In 2010 I carried out a follow-up interview with the user insight manager, as well as a few project negotiation meetings with the research coordinator. These interviews gave insight about changing user involvement practices in development (see more section 4.3), but here I will distill the relevant things from a user categorisation point of view. I learned that the developers talked about users differently, depending on their job role and depending on how long they had been working for the company. Server developers were farthest away from the users, whereas client developers and graphic designers were closer to the users. Some of the original developers did not talk in one specific set of user categories, but told stories about what users did in certain situations. The more recently hired developers leaned on the user categories created by marketing, since they did not have as much own personal experience to draw from as the more experienced developers. A server developer also included in-house stakeholders as users. For instance, one part of the developer’s job was to develop a back-end system for statistics and hotel configurations, called Housekeeping. For this developer, the hotel manager and others in the country-specific organisations as well as marketing were users, too. Another developer reflected on what different users may or may not have in 82


common and stated that the only thing users have in common is that they have installed the Shockwave plugin in their web browser. Taken even further, to its logical conclusion, what is common among Habbo users becomes exactly that, the experience of detailed practices in using the service: everybody logs in, everybody customises their Habbo avatar, everybody moves around, everybody types text that emerges as speech bubbles. Even though many of the developers started out as insiders in the Finnish Habbo community, with internationalisation and growing numbers of users, it became more difficult to keep track of the Habbo users. Informal engagement was no longer enough, and Sulake applied methods such as usability evaluation, market segmentation, and personas (elaborated in section 4.3).

4.2.5 Habbo Developers and the ‘Average User’ In the interviews with Habbo developers we found that they visit fansites regularly to follow what’s going on, reading both articles and forum discussions. Some of the developers feel a responsibility to check out fansites as part of their job, and send e-mails to other developers about interesting fansite content. Fansite feedback is perceived as immediate compared to other options and especially valued right after new releases, when users discuss new features. Not all developers have equal interests in the fansites, often because of their different work tasks. In design guidelines, the category ‘average user’ is mainly used in two ways. On the one hand, it is contrasted with more technically skilled developers, suggesting that developers should remember not to assume too much technical competence among most of the users.36 On the other hand, in HCI literature a common phrase is, ‘There’s no such thing as an average user’,37 which can be seen as a warning against reducing identities, practices, and tastes into too abstract user needs. Abstract user needs that are not grounded in particular settings might harmfully shape a design into something that nobody can identify with.38 However, in my empirical data, yet another aspect of the average user has emerged, which is not about the technical skill of developers or average users, nor about statistical methods to advocate a representative user.39 In my data, the average user is used in relation to other user groups, not developers. By analysing the following quote, one can better understand the complexities of categorisation practices in design for complex and heterogeneous communities: If one goes to the [guest] rooms . . . then one gets feedback from the average user. But in the forums, the users have used Habbo Hotel for a longer time and slightly grown out of it [Habbo] and they have moved on to the forums to discuss it. There are the [furniture] collectors, the older ones, and the other HC [Habbo Club] users. . . . I have the impression that the opinions are not that black and white among normal users. In the forums everything is either extremely great or then it really sucks. (Game Developer) 36 Spillers,

2006. 2004, p. 54. 38 Cooper, 2003. 39 Muller, Millen and Strohecker, 2001. 37 Budde,


83

I started the analysis by reading what was stated by the developers about the average user category, and then what this category’s relation was to the other mentioned categories. In the above text fragment, the average user category is used twice,40 first to state that the average user rarely writes in the fansite forums, and then that average users have more nuanced opinions than those expressed in the fansite forums. However, the reason for talking about the average user emerges by reading the whole paragraph. The function of the average user in this text is to contrast that category of users with other mentioned categories: furniture collectors, older users, and Habbo Club users. So, actually the average user cannot be taken literally, since the developer is not talking strictly about the average user of the whole Habbo population, but about the users who are not opinionated furniture collectors, older, or Habbo Club users. This interpretation is supported by the change to talk about ‘normal’ users. Based on his experience, this game developer feels there is a large group of users whose opinions do not get voiced in the fansite forums. Still, he feels it is important to include them in the design considerations. The problem is that since they are not so opinionated, they are hard to reach, and not much is known about them. Even though the group is probably as heterogeneous as any other user group, it is difficult to distinguish the subgroups. Therefore they are grouped together as the average user. Why the name ‘average user’? One interpretation is that it implies large masses of users; as the developers want to please as many users as they can, the ‘average users’ get more emphasis than other more marginal (although well seen and heard) user groups. The designer not only describes the user groups, but also actively constructs and configures the user groups in his speech, while reflecting on the constructed user groups that have influenced his earlier actions. He actively speaks for some users, and devalues other users, who in his perspective can and do speak for themselves. Akrich argued that successful artefacts depend on the ability of developers to generate user representations and integrate them into their design.41 She observed many different techniques for creating user representations, both explicit and implicit. The explicit techniques included market surveys and consumer testing, whereas the implicit were the I-methodology, experts, and other products. The term I-methodology exists when the designer puts him/herself in the position of the user and uses his/her own knowledge as a base for design. On the one hand, the concept of the average user could be seen as manifesting Akrich’s I-methodology: when things are uncertain, the designers go with their intuition or feelings based on their own experiences. On the other hand, their knowledge seems to be grounded in their proximity to the users: they can go and look at what’s happening in the hotel and they can read the fansites. The fansites give the developers a way of getting a feeling for what the users want by reading between the lines. The developers draw on their cumulative experience, as they were the users themselves in the early Habbo days, and they have been there since. As a consequence, here in the Habbo case, it becomes hard to tell the difference 40 Actually once, but I interpret normal as a qualitative version of the more statistical average and analyze them together. 41 Akrich, 1992; Akrich, 1995.

84


between I-methodology and feedback from use, since the designers’ reflections repeatedly refer to feedback from experience of use over time. Those who follow the fansites and the activities in the hotel appear as legitimate representatives for the users. The designers balance being representatives for the users, for the business, and for their own interests. Analytic means for grasping this issue further were developed in subsection 4.1.6 through the notion of ‘developer–user social distance’, where at one end I-methodology is distinct from actual use experience and at the other end the use experience of developers and users holds more similarities than differences. As a usability designer and researcher, I have been taught that one should never design for the average user. Doing so would probably result in a design that does not fit anyone, because the users are always a heterogeneous group. My first reaction when the game developer started discussing the average user was to shut my ears because it did not make sense. The use of the term ‘average user’ seemed to contradict HCI guidelines, such as ‘know the user’ or ‘define the user groups’. It seemed like the user opinions at hand were not followed, and the developer designed for the unknown average user. However, by using a membership categories analysis and extending the timeframe of the analysis, I realized that my initial reaction was wrong. When considering the development history and the developer’s experience of feedback from use, it made more sense to see the developer as a representative of the silent majority of the users. The developer brought in the users who did not voice themselves and denoted them by the category ‘average user’. In this way, he accomplished his goal of not having to strictly follow the immediate user feedback, which may not reflect the opinions of the majority of silent users, but he also accomplished his goal of considering the previous user feedback based on his experience. The rhetoric about the silent majority and the user is familiar from other contexts, for instance, in politics. The politicians often argue that they themselves represent the (mythical) citizen and give voice to the silent majority of voters. This might or might not be the case; it depends on the politician’s own agenda and its proximity to that of the voters. While the citizen as a concept decontextualizes individual human beings, it also makes it possible to talk about the broader concepts of the rights of citizens and human rights. It appears that the concept of the user works in a similar way, giving developers a way of discussing possibilities and restrictions for all users (e.g., in terms of user access rights, user profiles, and groups visible to the computer system). In addition, as some citizens need more empowerment than others, so do some user groups. In this case study, the ‘average user’ was not literally the average user of the Habbo population, but denoted those who needed to be voiced. Figure 4.6 presents a dynamic illustration of the user-developer dialogue through the fansites. The first panel shows some of the user groups on the fansites. In the second panel, the developer discusses the user groups that get voiced on the fansites. The next panel shows the developer creating the ‘average user’, which in the final panel dominates the other user groups. In this comic strip, the active role (agency) of both the game developer and the average user become visible. Making the average user bold and larger than the other user groups in the


85

Figure 4.6: Co-constructing the ‘average user’ and other user groups (article III).

fourth panel symbolises the work the category does. Even though the average user is fairly shapeless and unknown, it seems to direct the design. The developer wants to design for the average users rather than pay much attention to what the opinionated user groups write in the fansite forums. Creating the average user concept is a way of legitimising this. Using the average user in connection to the fansite forums (referring to the users not writing there) also leaves room for more representative user feedback from other sources. Yet another reason for leaving the average user shapeless is perhaps to allow and give room for new designs. Had the average user been completely defined, there would not be room for creativity. To sum up, the category ‘average user’ cannot be taken literally. It gets its meanings in relation to the other categories defined and mentioned. These categories are different, as they are locatable in user practices, while the average user is not. Furthermore, the categories are shaped or configured by the developer to fit his aims. At the same time, these configured categories shape the developer, as they keep the design space open. This study on the developer–user dialogue highlights the importance of the fansites as one user proxy for the Habbo developers. In this case, rather than describing developers as ‘malevolent manipulators of users’,42 the case points out their role in balancing and governing different user interests. Digging into the practices of design brings forth its complexity and shows how design and use are intertwined. Over time, the knowledge of the designers is not easily separable into ‘own’ knowledge versus feedback from use, as design and use knowledge becomes mixed.

4.2.6 Summary This section summarised my own explorations of categorising Habbo users as well as insight into the user categorisation practices among Habbo developers. Based on this data and analysis I put forward the following observations and arguments. (1) The social distance between developers and users significantly influences what counts as sensible user categories. 42 Stewart

86

and Williams, 2005.


(1a) Developers’ early informal engagement and personal experience reframes the need for user categories in social media development. If the developers start out developing the social media service for themselves and their friends—as many successful services have (subsection 2.5.3)—the need for user categories is reduced, since the developers already have a good understanding of the nuances of the intended and actual use practices. Sulake developers that were hired later on had a stronger need for user categories, as they did not have the same engagement with the user communities as the founding developers. (1b) Successful social media services face a need for stricter user categories later on in the service lifecycle. Even if a service starts out with small social distance between developers and users, later the sheer number of users of a successful service makes it difficult to follow user practices with informal methods. As the development organisation grows, new people, who do not have the same experience as the founding developers, enter and engage with development, and suddenly the ‘familiar’ gap43 between developers and users has emerged. With the developer–user gap development is ‘back to normal’, the situation assumed in many user involvement approaches, and the role of user categories as both a communication device for user knowledge and a steering device in development emerges. (2) Advice about categorising users has not fully incorporated the opportunities present in a social media design context. In a social media design context, like other online services, three new opportunities for data about users are present: user actions in the service, online user discussions, and server log data. As online user discussions are often public, they can be analysed with standard methods for text analysis from both a qualitative research and a data mining point of view. Social media service operators have an additional advantage, since they can use web analytics to analyse their server and service logs regarding all sorts of statistics of online user action and activities: site visits, transactions, and use patterns. This implies that, after the launch of a social media service, (2a) Developers can use multiple situation-specific user categorisations, instead of relying on one broad categorisation. Developers have easy access to online user action, so whenever a question of uncertainty comes to mind, a developer can just log on and check what users are doing and writing about just that topic.

43 Grudin,

1991; Nielsen, 2008; Norman, 1988.


87

(2b) Developers need not only rely on a representative sample of the user population, but can also use a census strategy. Most traditional research in psychology is based on the notion of studying a representative sample of a population. Researchers choose dimensions and criteria for what counts as representative, which guides the selection of people for their studies and the making of generalisations. This strategy was adopted in many user involvement strategies for a good reason. However, the new data sources of social media change this. On some dimensions (those automatically recorded), developers can draw on a full set of observation objects and subjects belonging to a given population (i.e., complete enumeration or census). Irrespective of whether a service is anonymous or not, users are often required to fill in their age, sex, and location in their user profile. This means that data mining sales statistics for what is trendy among a particular user demographics (e.g., Habbo users who are 14-year-old girls in Sweden or 12-year-old boys in The Netherlands) becomes easy. (2c) Since census-style data about users is available, the role of traditional qualitative and quantitative user research methods is reframed. The cost-effective approach becomes to check the census-style data first, e.g., what one can learn from web analytics and data mining the server logs, and then focus surveys and interviews on missing data or unexplainable patterns in the server data. However, as our work in categorising users shows, the analysis of censusstyle data can be significantly enriched by more sophisticated ways to differentiate users, activities, and roles present in a social media service. Knowing what to look for requires artful integration of census-style data with qualitative inquiry and a sociologically-informed analytic approach. (2d) Social media data can allow for categorisation of user groups and user pathways. Since the diversity and time-scale of the data points about users available through social media services are broader than before, developers can start thinking about new things to categorise to enlighten development. For instance, server logs and online data can provide detailed knowledge about the dynamics of social networks as well as the long-term changes in the user actions and activities. Users do not act alone in social media, which raises the question of categorisation of user groups. If the social media service provides data about social interactions (e.g., friend requests, messaging patterns, transactions between users, and duration of co-location in particular digital spaces) it makes good sense to use social network analysis and design the service for particular pairs and groups44 44 ‘Group’ is another overloaded concept that people with different backgrounds treat differently. In marketing, a group can mean any bunch of people with something in common, and in sociology distinctions between membership groups (where the one talking is a member) and reference groups (other groups) are crucial to make, for instance. Here I refer to a more social psychological notion of group, where the members have computer-mediated contact with each other in a service. I am not very particular about the degrees of a shared objective, place-boundedness, collective actions, and other common distinctions. For the purpose of distinguishing, categorising, and designing for these user

88


instead of particular individuals. The diversity of user motivations to use a service is broad, but how varied are the potential user pathways of engagement45 in the service? Users get drawn in to communities for one reason, continue to participate for another reason, become insiders for a third reason, and fade out for yet another reason. Are there regularities in long-term user participation pathways that could be used in social media service design? (3) User categories are not really characterisations of people, but characterisations of a development–use relationship. Take the analogy of a dog on the street. A casual observer would call it just that, a dog on the street, to distinguish it from humans, cars, and cats, for instance. A dog-friendly observer could distinguish whether the dog is a labrador retriever, a collie, or something else. A serious dog expert would probably note even more details about the dog, such as whether it is a hybrid breed of some kind, details about its training visible in the dog’s posture and behaviour, and so on, and use relevant dog categories. In this way, describing a dog on the street becomes not an objective description, but an encounter where the background of the observer is brought forward in the process of describing the dog. In the same way users are described in my data from different points of view. A server developer includes in-house marketing people as users, game developers talk about what the users do, and marketing at one point created a lifestyle-focused categorisation. Researchers have long argued that categories (and artefacts) are not neutral, but have in-built politics.46 The user is certainly one such category, which is exemplified in the section about the average user, where a developer configures user groups and himself, many use-development relationships in his speech. User categories are not business-neutral either. For instance, when customer loyalty, acquisition, monetization, and retention become important starting points for user research and resulting user categories, the business values are in-built in user categorisations. As soon as issues that are deemed significant for the business and development of one particular (social media) service are brought in to guide user categorisation, the categorisations are not made anymore with a starting point in the everyday life of people, but with one particular development-use relationship in mind. Users are not treated as a ‘naturally occurring’ phenomenon, but issues relevant to the development context guide the categorisation. As always, the point here is not that there would be anything wrong in Sulake’s user categorisation practices, but that there are many kinds of user categorisation practices: groups, the important thing is the particular link between users in the service that is counted as a measure of closeness between two or more users (e.g., the above-mentioned friend requests, messaging patterns, transactions between users, and duration of co-location in particular digital spaces), the frequency of interaction (and other interaction patterns) through these links, and the size of the group. 45 Often called trajectories in symbolic interactionism, and also in the Locales Framework. Fitzpatrick, 2002; Mansfield, Kaplan, Fitzpatrick, Phelps, Fitzpatrick and Taylor, 1997. 46 Bowker and Star, 1999; Suchman, 1994a; Winner, 1980.


89

• sometimes user categorisation is like particular kinds of social science, where users are studied as ‘naturally occurring’ and there is an effort for categorisations that depict users and their practices per se; • sometimes user categorisations are more subjective in many ways, specific to technologies and business models; • sometimes ‘the user’ becomes a melting pot of different stakeholder interests, where there are no obvious user categorisations; and • sometimes user categorisations become strategic, building on and complementing previous categorisations related to business and R&D needs (the topic of the next section 4.3). A user researcher needs to know which situation is relevant. • • •

4.3 Beyond the Scope of One Project in User Engagement This section opens up a second key topic in this thesis: contesting and developing the notion of the project as a basis for user involvement advice and guidelines. A dominant way of structuring the use of methods to learn about and co-design with users has been to use project phases more or less tied to the so-called software lifecycle (acquisition and supply; development: requirements analysis, architectural design, qualification testing; maintenance-operation).47 Project management was a key business model/strategy in software development in the mid–1990s; however, the emergence of social media has since changed software development processes and business models. What can we learn from this case regarding long-term user involvement? In the following subsections I start from my interest in the role of usability evaluations in development. I will analyse how it came about that the service worked well without a formal usability evaluation for four years, why it still became necessary, and why it continues to be part of development. This analysis is in contrast to many user involvement studies where usability evaluations are taken for granted. Because of its grounded nature, the analysis not only covers usability evaluations, but many other encounters between developers and users. These encounters, more or less planned ways for the developers to engage with users, and their changes over time are here the second operationalisation of the user–developer relationship—the user categorisation processes being the first. The starting point in my interviews with the developers was to focus on encounters between developers and users. I asked the developers where they got design inspiration and user feedback from, and how they knew what the users wanted. The answers to these questions included a wide variety of ways of engagement with the users. A basic understanding was gained through the developer interviews in 2005, and deepened through project meetings and feedback method work in 2006, but 47 ISO,

90

2002.


it was not until an additional follow-up interview in 2007 that I could figure out which were stable and which were changing practices of user involvement. I learned that the notion of ‘project’ changed a lot over the years: early on there was no project structure, then came separate customer projects, and after a few years the product–service combination became stabilised enough to enable release management; later on agile development with monthly releases was applied. As there was no stable notion of project to generalise my findings to, I decided to group the service and its evolution into stages. Based on changing dynamics of business, technology, and user communities I came up with the following stages: concept, beta, expansion, complexity, and competition (section 4.1). The next subsections discuss user involvement activities in these stages.

4.3.1 Insight into Early Development Practices In 2005 when I conducted my first interviews with the Habbo developers, I was surprised to learn that the first usability evaluation of Habbo was conducted in 2004, four years after market launch, and that only recently had usability evaluations become part of the software development process. This is in contrast with one main assumption present in my own professional education in user interface design, usability engineering, and user-centred design. I realised that the standard claim that usability evaluation is a critical part of every successful product or service design process was an overstatement.48 Here is an online service that had managed successfully without a formal evaluation for four years after market launch and even longer if one counts the first steps in development. What can explain this? When analysing my interviews and other data sources I found a number of compelling reasons for why usability evaluations were not necessary in the early design phases: • the low social distance between developers and users, manifested in active users and developers’ continuous informal engagement with users • the characteristics of social media services in general, which here refers to short release cycles due to immediate distribution potential of social media and the efficient mediation of user practices through social media • the excellence in user interface design and user experience, perhaps due to cultural maturation and mature user interface genre. These points are further elaborated below. Low social distance. In the early Habbo days, the developers themselves were part of the Finnish user community, which made direct and active developer–user communication possible. The developers were developing a service for themselves and their friends. Many of the users were professionals in new media, while the current user population of teenagers became the norm in the service only after a year or two. Trusted users, who had a volunteer history from the first Mobiles Disco and Lumisota services, continued as volunteer moderators in Hotelli Kultakala. 48 Greenberg

and Buxton, 2008, later made similar observations in other development contexts.

4.3 Beyond the Scope of One Project in User Engagement

91

The characteristics of social media49 enabled very short release cycles. The lead developers could get an idea, work on it, and put it out for the users still on the same day. Active users, especially volunteer moderators and other insiders, tried out the new feature, and gave immediate feedback to the developers. User activities in Habbo were mediated to the developers through both direct and indirect communication channels. In the same way that users found out what happens where in the hotel, developers could also log on to Habbo and check what is going on through the trend mechanisms.50 Sales statistics were developed in the server back-end over time as needed by development, marketing, and management. Developers got feedback through e-mail and via visiting user-created fansites. Excellence in user interface design. The Habbo user interface was modeled after cartoon-like video games in the 1980s—it was minimalistic, pixelated, and had an axonometric view. Navigating in a room and between rooms was made easy, chatting was just writing in a box, and the animated avatars could only go in four directions and did not have many special moves. Basing the design in a proven category, early video games and Commodore 64 style, while adding ‘online’ and ‘multi-user’, made an easy-to-use interface. Unique design choices and creative content production were the foundation of a highly rated user experience. In contrast to business products where functionality is more critical than user experience, consumer software often comes with easy-to-use interfaces.51 So did Habbo, in a combination of simplicity, high graphical ambition, and a pay-as-you-go business model. The interviews with the developers also revealed informal evaluation practices, such as the Habbo Ladder (see Figure 4.7). The Habbo Ladder was made to contrast Habbo with the comparably higher threshold to start playing massively multiplayer online games at the time. The idea was that during the first few minutes that anyone is willing to spend on a new web service, one should be able to login and create an avatar easily, learn the basic navigation, and have a chat with someone else. Each step on the ladder makes more reasons to return to the service, as a developer explained: Especially if one gets the first friend on the list of friends, then that is a reason to return, that you have really got to know someone with whom you might have had an interesting discussion or of whom an interesting image has been conveyed. Then further on, when you have your own room, well that is of course a real investment, even the notion that you have something own going on there, then that is already a good reason to return, especially if you have decorated the room, really purchased something. (Interview 19 Apr 2005, translated to English by the author). 49 Compared to tangible products, software does not require a separate manufacturing or assembly process where the physical parts are put together. Compared to shrink-wrap software, the distribution of social media and other online services does not require vendors, stores, and installation media (disks, CD-ROMs, DVDs). Compared to client–server softare, social media does not require installations on the client computer, but rather runs in the web browser. (More in section 2.5.) 50 Throughout the Habbo history, users and developers have created various mechanisms to spot the ‘best place to hang out’ in Habbo: room lists sorted by popularity or number of co-located avatars, weekly or monthly picks, event marketing on websites and discussion forums, automatically indexed rooms and avatars (tags), and other room categorisations. 51 Campbell-Kelly, 2003.

92


Figure 4.7: Habbo Ladder, drawn by a developer, interview 19 Apr 2005. The Habbo Ladder is described in article VI, as are the following user involvement practices significant in the early development of Habbo: • avatar activities, • developers as users, • informal evaluations, • e-mail feedback, • volunteers, • volunteer forum, • weekly newsletter and polls, • fansites, • official web fanzine, • summer meetings, • sales statistics, and • customer service.

4.3.2 Expansion and Maturing Development Processes In the previous subsection we learned that some of the functions that usability evaluation normally has in development—meeting users, seeing others use the software, getting feedback and inspiration—were already in place through other means early on in development. However, during the interviews in 2005, it was already apparent that the situation then was different from early development. Marketing was segmenting the users, usability evaluation had emerged, and beta testing was formalised into release pilots. It was still the same social media service with an easy-to-use interface, but things had changed. What changed in the development context and user involvement methods? Based on the interviews, I could locate three emerging change themes • increased social distance between developers and users – younger users 4.3 Beyond the Scope of One Project in User Engagement

93

– developer online presence problematic – bigger development organisation, more specialisation – internationalisation • technical stabilisation • more focus on the economics of the development In the beginning, most users were in their late teens and early 20s, but this changed fast. The product appealed to teenagers—first in the UK, but then also in Finland, younger users became the norm. By 2004, 75% of the users were between 11 and 14 in Finland. This implied that the developers were no longer developing for their own generation, but a younger audience, which made developer experience and in-house testing less adequate as arguments in quality discussions. Some of the original developers had reached a legendary status in the Finnish Habbo community. This made it virtually impossible for them to spend time online in the hotel, because of too many fans trying to communicate with them at once. Also other development staff members immediately got surrounded by fans unless they were incognito. This development in the community made the immediate developer–user communication more difficult. During the first four years lots had happened in the development organisation. The company had grown to about 100 employees, each month more than 1 million players visited the hotels, and turnover was 5 million euros in 2003 and almost three times that in 2004. In 2004 the company also expanded to about 10 new countries. More specialisation occurred and not everyone could be involved in the user communities. Technically, the software architecture had been refactored and rewritten, the administration was easier, and a release process with two or three releases annually had been established. There was less firefighting, more time, and resources available. No longer was there different code in different countries; Habbo had turned into a configurable service where local operators could turn features on and off on demand and new features could be rolled out to all country-specific hotels simultaneously. Business had become bigger, which meant that there was less margin for error. Also, in contrast with the development situation earlier on, where things worked out well as those in the core group could develop what they were most inspired by, now new feature development had economic measures. Potential impacts on revenues had to be estimated. With the international expansion emerged a need to know whether the user communities were similar or different in different hotel countries. An outsourced market survey in 2004 generated customer segments and their regional distribution. To prepare for the rapid international expansion that happened during 2004–2005, focus groups were conducted. The applicability of Habbo pixel-style graphics and use of colours was evaluated for the Asian market. The first usability evaluation in the fall of 2004 was targeted at checking the usability of service registration and those services in Habbo that were subject to a fee from the viewpoint of 10–14-year-olds. This evaluation was one among other quality enhancement and packaging efforts made to ensure that a high-quality

94


service was duplicated to the 10 new hotel countries in 2004–2005 and that feature distribution could take place in a more controlled manner. For these reasons it is understandable that more documented knowledge about users and formally tested quality gained importance. This was exemplified by the new process enhancements described in article VI: • market segmentation through market survey, • focus groups, • usability evaluation, • playability testing, • a CRM system, and • release pilots.

4.3.3 Service and Community Complexity Management In the previous subsection we learned that the developer–user social distance had increased from being very small at launch to being somewhat broader, but developers were in good touch with the user communities. Again, through additional interviews I found that the development situation in 2006–2007 was different from 2004. The following change themes emerged: • increased complexity of the service, • new ways of knowing about users, and • changes in development rhythm. In 2007, turnover was 43 million euros (still a negative net return, though), there were 9 million monthly users, and the company employed about 300 persons. The users came from communities in 32 countries, and Sulake had 14 local offices besides the headquarters. This made it impossible for one person, or even a group, to have detailed knowledge about what happened in all the hotels. Each local office had a good grip on the activities in its hotels, but the need for communication inhouse was greater than before. The sampling problem was bigger, and making sense of what all the different user communities and subcommunities wanted was more difficult. The service expanded to include more social networking features (article VI). Besides the friends list that had been there from the start, the user profile was changed: it was not only available when online, but every Habbo avatar got an automatic and customisable Habbo homepage. It was also possible to form Habbo groups, which meant a logo, a group homepage, and a discussion forum. In addition, users could ‘tag’ their avatars, which meant that users could attach a set of clickable one-word descriptors to their avatar. When a user clicked on a tag, the service generated a dynamic index of all the users and groups with that particular tag. Due to complexity of the service and the diversity in use, evaluation became more difficult. Through my work with Sulake’s user feedback methods in 2006, I realized that a simple question like ‘is this newly developed feature good enough?’ became more difficult to answer. Users would complain if there were lots of bugs in a new feature, but to get comparative data, challenges emerged: Which features


95

can meaningfully be compared with each other? How can one assess the meaningfulness of a particular feature for particular users and their use practices?52 Such analysis seemed not possible with single surveys, but meaningful comparisons would require triangulation with other data sources and an analysis of a series of surveys and standardised questions in order not to get too overwhelming. Sulake developed new ways of knowing about the users. The user and group homepages in combination with tags provided new mechanisms to see what is popular among users. Enabling discussion forums on Sulake’s servers, instead of discussions taking place only on fansite discussion forums, made data mining easier for developers (subsection 4.4.4). Also, as described in article VI, an online user panel was developed. Sulake recruited 200 volunteers in one country to form an online panel. The online panel was given a weekly task consisting of a set of questions regarding design sketches and an opportunity to share opinions regarding the sketches in a forum. A global youth survey created a new segmenation of Habbo lifestyles: achievers, creatives, loners, rebels, and traditionals. Through a pair interview with two usability and playability experts in Spring 2007, I learned that the usability processes had developed. After the first outsourced usability evaluation, Sulake established an in-house usability process through a pilot project. The focus of the succeedent evaluations was not the totality of the service, but they followed the game development, as new features and product extensions were developed. The reasonings for why usability was important were not only service quality, but that usability was part of contracts with other big, related organisations, and raised quality expectations from the market: And as we are such a big player on the market now, we have to make sure that the product meets the standards set by others, and we want to lead our slot.—Usability Expert I also learned about nuances relevant to development: what was old code, what was new code, and what was hard-to-change code. Development considered highpriority features those that were used often—use influenced the prioritisation of features—or were directly critical to business: login, registration, and payment. When investigating which features had not been changed I found three categories: (1) low-priority features, e.g., Peelopaalu; (2) features that were difficult to implement due to lacking support in the underlying development framework (Flash), e.g., drag’n’drop; and (3) features that relied on external standards, frameworks, and processes, e.g., credit card payment processes. To sum up, I learned about internal development rhythm: • more frequent releases, from twice a year to four times a year; • usability evaluations not a separate process, but followed feature development; • a yearly interval between the market surveys; • some continuous targeted interaction with users weekly: 52 Sofas can be compared to chairs, because you can sit on both. But a metal chair in Habbo can also be used en masse, creating an airplane setting. In order to assess the quality of a piece of Habbo furniture, detailed knowledge about its use is needed. Is it used as suggested by its designed form and function, or have users taken it into use in their own ways? Is it more or less critical to which use practices?

96


– weekly newsletter – weekly assignments in online user panel; • some features were unrelated to internal development rhythm. Internal rhythm in an organisation becomes more important as the organisation grows and specialisation increases. When more people do different things, syncing between processes becomes more challenging and clearer rhythms need to be developed. Such rhythms are an underresearched subject in studies on user involvement. If there are discussions about rhythm, they are often connected to one project following the so-called software lifecycle and its stages, or rhythms related to agile development: development practices that take place daily, weekly, and monthly. The findings here regarding development rhythm indicate that • one cannot assume a typical development process or rhythm; • there are more concurrent rhythms than one; • development rhythm changes over time. Timing user involvement is critical. Development rhythm needs to be included in studies assessing the use of user involvement methods.

4.3.4 Competition and Globalisation When presenting my overview of the user involvement methods (Table 3 in article VI) in 2009 to Sulake’s research coordinator, one reaction was that the article did a good job of describing the development of user involvement practices, but it did not describe the current situation very well. To get insight into the changed situation, I did a follow-up interview with the user insight manager in March 2010. It became evident that Sulake had implemented a new strategy to learn from users: • user experience testing, • data-based personas, • more data mining of user activities, and • automated surveys. Following current research terminology, in 2008 Sulake talked about user experience testing in addition to usability or playability testing. While the usability testing had evolved from a more stand-alone practice into a tight integration with agile software development, Sulake conducted user experience evaluations with both new and old users, internally called ‘live tests’, in one country for every major release, about once a year. As part of further developing user-centred design processes, during Spring 2009, Sulake applied the Persona method.53 Six user archetypes had been constructed from data to represent the users. The idea was that developers have an updated reference to the goals and needs of Habbo users at hand, which could inform design solutions and evaluations. In 2010 the user insight team employed four people. Besides the manager, who had been doing the research for more than five years, there was one person doing 53 Pruitt

and Adlin, 2006.


97

data mining, another expert on quantitative methods, and a third on qualitative methods. Data mining and monetization of user data had become more important topics in the organisation. The process of learning from surveys had been significantly developed with the aid of automation. Based on certain triggers—for instance, awhile after becoming a Habbo user, or after a few months, or after not being active for a while—users got a survey to answer. There were about a hundred different surveys ongoing at any one moment. The use of data mining and automated surveys had shifted the role of qualitative research and evaluations in person. Such methods were used more rarely and were more focused—for instance, when a pattern in the data from other sources could not be understood or when examining cross-cultural differences where a wide social distance between user researchers and users were assumed. The six data-based personas figured actively in feature development, communicating user research results to service developers. In contrast to previous service lifecycle stages, the representation of user needs, motivations, and aspirations were more processed, but also less direct. Instead of checking out Habbo for oneself, developers could now rely on a set of continuously updated data-driven personas. This again changed the rhythm of user involvement: • surveys: continuous • data mining: continuous • personas: continuously updated based on the above two • yearly user experience testing • (bi)annual global youth surveys: 2006, 2008, Brand Update 2009, Global Habbo Hotel Fashion Survey 2010. Table 4.7 below shows an overview of the user involvement methods used during different phases of Habbo service evolution. The position of a particular method in the table columns signifies the first use of that method. Some methods like avatar activities and sales statistics continue to be important in latter phases as well, since their introduction as an organisational practice. (More details in article VI.) Table 4.7: First occurrences of user involvement methods by service evolution phase.

98

Concept

Beta

Expansion

Complexity

Competition

1999–2000

2001–2003

2004–2005

2006–2007

2008–2010

Avatar activities, Developers as users, Informal evaluations, E-mail feedback, Volunteers

Volunteer forum, Weekly newsletters and polls, Fansites, Official Fanzine, Summer meetings, Sales statistics, Customer service

Market survey, Focus groups, Usability evaluation, Playability testing, CRM system, Release pilots

Online user panel, Global youth survey, User and group homepages, Tags

Data mining, Automated surveys, User experience testing, Personas


4.3.5 Summary The notion of the project has structured much user involvement advice. Different practices and methods are suggested early in the project, in the requirements phase, compared to other phases such as implementation, testing, and maintenance. The rationale was that in the beginning of a development project, the design scope is open and there is little knowledge about users. During the project, the design scope becomes more closed as ideas converge and developers learn more about users. In the end, the design scope is closed, new ideas must wait for the next project or change management during maintenance, and designers know a lot about users. However, this notion of project and its phases structuring and giving rhythm to user involvement methods seems ill-suited to both Habbo and social media development in general. It seems more fit for software contractors that offer standard services, not software companies that own and operate a service. In this case study I found that the notion of the project was not stable enough to allow for structuring and generalisation of user involvement advice. This has to do with the transition of the development organisation from startup to small and medium-sized firm: varying degrees of organisational specialization, developer–user social distance, technical stability, project scope, and multiple changing development rhythms. The variety in projects also relate to the dynamics of social media today: relatively low threshold for startups, fast changes in consumption and business environment (service competition, hardware, and software frameworks), and no standard size and scope for software business deals. This is unlike established software products for business, e.g., cars, airplanes, and ERPs,54 where products and companies have a long history and more standardized business routines. Following this finding about the instability of the project, I put forward the following arguments about rhythms and change in user involvement. (1) The notion of project and project phases is not beneficial to structuring user involvement advice for social media, nor for generalisation across cases. In my case study, projects were used to structure early development, but they were all very different from each other in terms of contribution to the service or needs for user involvement. Later, the concepts and practices around releases and sprints structured service development, but their character varied similarly very much. Given the dynamics of social media today, it is unlikely that a stable notion of project will emerge for service operators in the near future in this domain. It did not happen in this case, and I doubt that it will happen in other cases. Perhaps with the exception of (1) companies with a considerable amount of resources (e.g., Amazon, Google, Facebook) that have the resources to shield development against external conditions, or (2) companies with hardware devices that provide a rhythm for their software development (e.g., Apple). 54 Campbell-Kelly,

2003.


99

(2) User categorisations are cumulative, emergent, and strategic: the criteria for successful categorisations change depending on changes in the development–use relationship. In contrast to typical psychology-based approaches in HCI, the key criterion for Sulake’s user categorisations has not been fixed over time nor set to understanding users and their practices per se, but rather has reflected how valid (useful, even inspirational) the information is for design and business concerns. The development went through the following overlapping business-sensitive phases: 1. Design for socialising, making Habbo a cool hangout online. As the developers were also users themselves, the informal engagement with the user community gave the developers a good enough implicit understanding of the users. 2. Design for use practices; when the developers realised that there are too many users to keep track of, they started thinking about what is common for all users: logging in, learning to navigate in Habbo, connecting with others, etc. (see the ‘Habbo Ladder’ in subsection 4.3.1). 3. Design for a changing target group and to account for an increased social distance between developers and users. As younger teenagers took over Habbo (and the developers grew older), the developers realised that an age gap of 10 years had emerged between developers and users. The developers were no longer insiders in the user culture, which meant that market and user research strived for an understanding what is cool and ‘in’ within the target group. Means for safe playing were implemented and the parent of the user became a key stakeholder in website communication. 4. Design for cost-efficiency and double-sided business. The target group of the first usability evaluation was new users and business critical service features. In 2004–2006 marketing needed to show advertisers that ‘Habbo users are normal people, not freaks’, so at one point the focus of the persona descriptions became lifestyles, not activity inside the hotel. This reflects the double-sided business model of Habbo—income from users is one part, income from ads is another. When economics started driving development more, focus changed to customer loyalty, acquisition, monetization, and retention. 5. Design for global competition and multi-sided business. In 2009, when the Personas method was implemented in a data-driven fashion, the focus became to ensure that the six persona descriptions should reflect the growing and declining market areas as well as have an even gender and age spread. In addition, they were created to serve service optimisation and new business models on monetisation of user data. This is a contribution, compared to how lessons on user involvement are commonly communicated in HCI: user categorisations are used strategically and cumulatively. At first user practices and their (implicit) representations drove business, later user representations became means to implement business strategies. Understanding users and their everyday life as such was replaced with ‘whatever works and helps the organisation to the next step’. 100


The strategy is emergent, but within boundary conditions. It has been adopted in response to unforeseeable changes in the target group and business environment. However, with hindsight, it is difficult to imagine social media business that would not have to deal with these development–use–business components of socialising, use practices, developer–user social distance, cost-efficiency, and global competition. Sulake–Habbo was influenced by these components in a particular sequence; other services might be influenced in a different order, but the components will be there. (3) Instead of assumed conditions related to ideal project phases, advice regarding user involvement could be structured by the actual conditions in the development context. The dimensions that made a difference to user involvement in this case study and are believed to have broader relevance regarding user involvement guidelines and advice were: • Developer–user social distance, • Organisational specialisation and internal rhythm, • Degree of business/mission criticalness of a feature, • Project scope: openness of design space and variance in use practices, and • Relevance of existing knowledge about users. In the following, each dimension will be outlined in more detail. (3a) Developer–user social distance In Habbo, the developer–user social distance started out small, but grew first through differences in age, professional background, and wealth (a consequence of there being more teenagers than professional new media people in Habbo), and then through differences in language, regional and national cultures, ethnicity, lifestyle, and so on (resulting from internationalisation). If the developer–user social distance is small, developers’ informal engagement and personal experience can substitute for more formal methods. It is likely that the developer–user social distance grows with a successful service. A broader developer–user social distance requires more effort (more formal research) from the developers’ side to understand users and reduce uncertainty about variance in use practices. (3b) Organisational specialisation and internal rhythm When organisational specialisation increases, which tends to happen when organisations grow, more effort is needed on communicating knowledge about users and their use practices within the organisation, as not all managers and developers can have deep knowledge about users and use practices. With more specialisation different concurrent rhythms in development emerge: community management practices can have one rhythm, market surveys another, and development a third, or even multiple rhythms if there are side-tracked parallel development with longer timeframes.


101

Doing user research two sprints ahead of development is one reported way of integrating user research with agile development,55 but that is just one specific example of one outcome of a specific combination of internal rhythm with the other mentioned aspects, that is, 3a, 3b, 3c, 3d, and 3e. (3c) Degree of business/mission criticalness Login, registration, payment processes, and other factors enabling a low threshold of use are critical parts of most services. It is a big deal if they do not work optimally. Sulake focused its first formal usability evaluation on these processes. On the other hand, less important features can stand more bugs or longer fixing times. Sulake left low-priority features hanging for a while. The degree of business criticalness of specific feature development in Habbo has varied between countries. Like with any transnational service, Habbo developers experience the dilemmas of language regions and differently sized markets. On the one hand, one could argue that, because of Habbo’s Finnish origins, hotels with close linguistic ties to strong languages in Finland (Finnish, English, Swedish) would shape the service most. However, there are two factors counteracting this simple argument. First, language skills and effective communication among Sulake employees can mediate remote user needs. Second, since some user needs are readily communicated through sales statistics, the market size of a particular hotel country might be more significant than the language in that country. (3d) Project scope: openness of design space and variance in use practices User feedback and use practices have most influence on the features that are under active development. Early on emergent use practices and user feedback were significant, for instance, in the development of furniture ownership rights and their sharing, navigation between rooms, furniture trading mechanisms, moderation, and online discussion about Habbo. After the initial years, the Habbo service concept stabilised for a few years, where user feedback influenced incremental feature development. In 2006–2007 the service concept was broadened with social networking features and user feedback could influence those developments. Habbo service development resembles city development: different city areas are open for development during different times and those who where not city founders have the impression that city centres were filled with houses a long time ago and are hard to change. In the same way, current users may feel that central Habbo features were defined and refined a long time ago, whereas the users involved with Habbo in 2000–2002 could have a different feeling of being pioneers in a broadening ‘Wild West’. Assessing relevant variance in use practices is significant for fitting a technical feature to social practices. This reflects in both in-built technical flexibility and in the selection of relevant user involvement methods. For instance, login, registration, and particular payment options are features with use practices that are tightly scripted with little degree of freedom. On the other hand, decorating a room and moving about in Habbo are very open-ended use practices. For open-ended use 55 Miller,

102

2005; Patton, 2008; Sy, 2007.


practices, technical flexibility is key and user research methods that can tackle open-endedness (observation, interviews, data mining server logs with machine learning algorithms). For tightly scripted use practices, clear interaction sequences are key, as are user research methods with a high degree of control and a priori definitions, like A/B testing and quasi experiments, for instance. (3e) Relevance of existing knowledge about users Sulake’s current situation can be described as having a broad distance between developers and users, high degree of organisational specialisation, and many internal rhythms, but for many purposes the organisation already has the user knowledge it needs. This means that for some development sprints no new user research is needed, nor any questioning of the alignment of design goals and user needs, which are some standard assumptions of what user-centred design should do early in projects. Prior method use history shapes consequential method use. For instance, as the service as a whole had been usability evaluated, the usability specialists at Sulake found no sense in repeating the same evaluation again. Usability evaluations turned towards smaller details of the service after the overall evaluation. The same finding was found in the market survey practices. After the first overall customer segmentation, the following surveys could fine-tune specific issues that remained open in the previous surveys. This practice can be interpreted as a sensible ongoing tailoring of methods that becomes necessary when dealing with as complex and changing phenomena as in this case. For other purposes, use practices have changed or the Sulake’s knowledge about users has become otherwise outdated. This means that for some development sprints, new research is needed in advance. For these reasons it is necessary to assess the relevance of existing knowledge about users. • • •

4.4 Pathways of User-Created Content and Contributions While section 4.1 explored the emergence and co-constitution of Habbo, section 4.2 user categorisations, and section 4.3 change in user involvement practices, this section analyses what happened to user contributions. The fact that users are creative and contribute to social media is not news to anyone anymore, but what pathways do these user contributions take? Did they become features, shape community policies, or were they left as tweaks? User-created content (section 2.5) stirred up a many debates when Blogger, Flickr, Youtube, Myspace, Digg, FeedBurner, Second Life, and similar services became societal phenomena. The service developers had not only created text-, photo-, video-, and audio-sharing systems, but also publication systems. What one user did in and contributed to the service became available to other users and contributed to their use experience. The phenomenon of user-created content has become noted in many articles

4.4 Pathways of User-Created Content and Contributions

103

and reports, e.g., OECD 2007, where various technical, social, economic, and legal drivers have been identified. While serving as a relevant starting point, many early analyses of user-created content risked being too superficial and shallow, e.g., people blog on Blogger, share photos on Flickr, share videos on Youtube, etc. Recent research has noted how the social interaction that takes place on these sites is much more multifaceted and nuanced than the cursory descriptions. People comment on the content shared by others, form groups, create and maintain social networks, get otherwise unavailable information, and act out commonplace rituals of everyday life. However, even the research that opens up the nuances of user practices on these social media sites rarely takes a critical look at the developer–user boundary. Categories such as developers and users are often taken for granted, and there is not much research that looks into (1) what roles users play and the functions that they perform to make software work, and (2) how functions shift between developers and users (e.g., online moderation in Habbo started with user volunteers, but was taken over by employed moderators). The following subsections develop a critical analysis of the developer–user boundary through a number of lenses. First, a content perspective is employed by focusing on the question of what content is created by whom in Habbo. Second, we take a look at the tweaking and subverting of Habbo by some users. Third, we analyse and structure developer reactions to these emergent use practices or user contributions. Finally, we analyse the fansites, their role in community management, and boundary shifts taking place between developers and users.

4.4.1 Content Creation Capacities Section 4.2 gave an overview of different strategies to categorising Habbo users. Some strategies were based on demographics, some on engagement with the Habbo communites (operationalised as visit frequency and age in Habbo), and still others on the characteristics of the online activities. While pondering the idea of usercreated content and what counts as content in Habbo, I created a figure to structure different user-created content. It seemed that temporality and visibility were two important characterising dimensions. Figure 4.8 shows different Habbo content on these two dimensions. Starting in the lower-left corner, we find all the things one can do alone in a hotel room: move around, chat with oneself, interact with some of the furniture (e.g., throw dice, rotate a bottle, get a cup of coffee from the coffee machine, turn on the lights). These actions leave no permanent trail and nobody else in Habbo, except moderators, can relate to them without being present in that particular room. There are some actions that leave permanent trails in a room—for instance, decorating the room with furniture, rugs, floor patterns, and wall colours. These actions, including writing a Post-it note or putting a photo on the wall, are grouped in the box above the lower left corner on the level of resource change. In order to be able to do the above things, one needs to have both a Habbo avatar, a room in Habbo, and furniture. These more fundamentally transformative actions are grouped together in two boxes above the lower-left corner, on the level called lasting change.

104


Figure 4.8: Structuration of User-Created Content in Habbo.

While the above actions can be carried out in solitude, as soon as other Habbo avatars are involved, we move from the left towards the right on the visibility axis. The middle column, friends, indicates interaction and content creation in Habbo together with Habbo avatars, but with audience restriction: not everyone can participate, only Habbo friends or members in the same Habbo group. The list of friends, messaging with friends, and password-restricted rooms are central in making such content creation possible. The righthand column, all, indicates actions that spread further by being visible to all Habbo users in one hotel. This requires toplist functionality or equivalent filtered and centralised communication. Everything in the hotel cannot be broadcast to everyone, but several mechanisms are built in to allow some actions to travel further than other. For instance, the personal rooms are showed in the navigator sorted by number of participants, so rooms with many Habbo avatars are showed first. Towards the top right-hand corner we note more lasting changes—for instance, if a room hosts a seemingly unending series of events, or when a group becomes long-lasting and enduring. The top row in the figure, especially the top-right corner, indicates fundamental changes in the hotel that within the timespan of the hotel seem ‘forever’—for instance, launching a new hotel, a new public room meeting place, new floorplans for personal rooms, new furniture, new in-game levels, points, currencies, or other features. These are things that only paid Sulake developers can do via access to the back-end and development interfaces, which is why the top row has another background colour. Including this level of content shows that content created by developers and users can be analysed on the same dimensions; developer-created


105

content has more visibility and persistency. The figure not only shows the content creation roles of developers and users, but can also be used to show what moderators do in Habbo. Moderators create the rules for what content can be allowed to become more visible and persistent. For instance, what kind of room names are allowed, what events can be marketed, what topics groups can and cannot be discussed and in what tone, and what character names and descriptions are not deemed suitable for the Habbo way. In this sense, the moderators and their policies sit on the lines between cells in this figure, allowing some content to sustain while filtering out other content. Besides users, developers, and moderators there is a fourth common special role in social media: advertisers and campaigns created by them. Not everyone can create an advertising campaign and the service operator usually charges per views or clicks. Advertising campaigns are visible to all and usually last between a week or a month. These characteristics of advertising campaigns led me to put it in the box that intersects all and lasting change. This figure is important because it challenges the clear-cut categories of users and developers, opens up nuances, and places them on a two-dimensional continuum. Not only are users and developers considered, but also moderators and advertisers become visible in this figure. The persistency and visibility dimensions of content are not particular to Habbo, but feature in all social media. For this reason, this figure is important to social media theoretisation and development practice. Instead of starting from a specific definition of users, such as ‘individual that interacts with the system’ (ISO standard), this approach has avoided an a priori definition of stakeholders, but instead leaned on an approach that is grounded in empirical data. Such a methodology of leaning on content creation capacities is here suggested as good practice for social media research. This analysis also contributes to an expanding paradigm on content analysis in Internet research.56

4.4.2 Tweaking and Subverting The previous subsection put users, developers, moderators, and advertisers ‘on the same page’, by studying their different content creation capacities. The figure can be used not only to map how content changes in visibility and in persistance, but also how the results of different users’ actions and activities end up in different boxes. Very early in our research on Habbo we (Johnson–Toiskallio) learned that Habbo users do not just hang around, meet other users, chat, play games made by Sulake, and decorate their virtual hotel rooms like their own room at home. In contrast, articles I and II showed that many users are much more creative. They • bring in themes from child’s play: collecting and trading furniture, play ‘catch me if you can’, orphanages (a place to find participants for pretend play in a mom–dad–child constellation), and other children’s games; • imitate adults, tv formats, and professions: casinos, pretend dating, Idols, The Bachelor, beauty contests, journalist, doctor, policeman, etc.; 56 Herring,

106

2010.


• play with the spatiality of the hotel: furni trap games, Habbo variants of musical chairs, Habbo Soccer, and other sport imitations and transformations • form groups with missions in Habbo; • connect rooms together with teleports to make a wide variety of room constellations; and • find and use glitches in the hotel architecture. In articles V and VII we (Johnson–Sihvonen and Johnson-Hyysalo-Tamminen) analysed in detail two examples of tweaking and subverting Habbo: gothic play and playacting horse girls. In article V we (Johnson–Sihvonen) analyzed various forms of online gothic style in avatars, virtual objects, and places. Players of massively multiplayer online games usually use the available means to achieve the typical goth appearance of pale skin and dark clothes for their avatars. The reproduction and transformation of the gothic style is also rather common in Habbo, although it was initially designed to support benevolent and playful online behaviour. Furthermore, we discovered that performing goth online is not just based on a visual style, as it is also a way of playing in Habbo. Being a goth in Habbo can mean different ways of being noticed and being alone together with like-minded players, as well as provoking the more mainstream players. Actively discouraging contact is one such norm-breaker, and another example is the gathering of ‘off people’ players in a room, where everyone sits silently in the chat. We established that goths in Habbo play an interesting role in the development of the virtual world, as well as in the maintenance of it as a platform for social interaction. From the user’s point of view, Habbo is a place where gothic users can find the company of other gothic users. The user-created gothic scenery—avatars and rooms in gothic style—provide a fun environment for the interaction. Gothic events, rituals, and groups structure the interaction and enable participation in something bigger than just a conversation. Talking about gothic topics can be fun or ironic for some gothic users, but it is also a way of dealing with personally touching and emotionally charged topics. What Habbo goths have from the developers’ point of view is a specific role in the innovation process. Goth users provide the kind of content for the virtual world that interests many teenagers and can be further utilized by the developer company. Some gothic users are consumers—they pay real money for virtual furniture—but there is more to it. For example, gothic users started out by using the ‘Halloween’ furniture line of candles, skulls, and bats for their own purposes. The developers noted the popularity of the ‘Halloween’ line, and in 2007, they incorporated parts of the gothic subculture into the core of Habbo. A gothic line of furniture emerged as a set of its own, which shows the impact of gothic users in the design and innovation processes. Without the content and feedback provided by gothic Habbos, the gothic line of furniture would probably not have been launched by the developers. These two perspectives on goth subculture in Habbo make an interesting duality. On the one hand, Habbo goths have created a provocative anti-mainstream playing style; on the other hand, the developers have managed to turn it into a business benefit. This incorporation of subcultural activities has not meant the end of


107

playing goth, as early subculture theory by Hebdige suggests.57 On the contrary, the gothic subculture is still vibrant. The gothic subculture in Habbo has survived commercialization, which is in line with recent research on the new relationships between gothic subculture and the mainstream.58 In article VII we (Johnson–Hyysalo–Tamminen) report of the story of two horseinterested girls in Habbo. They managed to turn the indoors setting of a hotel into virtual stables through skillful furniture decoration, coordination, and imagination. Horse-riding activities (like any other, such as formula racing, found within Habbo) require a quasi-bounded digital space, or a setting that allows the play to be focused on commonly shared digital objects. The spatial infrastructure is arranged in particular ways to enable interpretative flexibility about already named objects provided by Sulake as basic components of the rooms–tables, chairs, decorations, and such. A new ‘placeness’ can be symbolized through their spatial rearrangement, the result of which is that any object residing within this social space becomes resymbolized. In horse rooms (e.g., ‘stables’ and ‘riding grounds’), for example, the spatial rearrangement indicates that tables and chairs have become new social objects (e.g., stalls or hurdles) that are constitutive ingredients of virtual horse activities. The riding activity itself is socially coordinated by assigning roles to the avatars within the rooms by their location or by active textual communication. Within rooms indicated as settings for equine activities, an avatar can become a ‘horse owner’ or a ‘horse’. Horse ownership is claimed for an undefined period of time by asking the horse for its approval. The role of ‘owner’ is self-claimed. One can just start to perform linguistically most of the care work one would do on an organic horse outside Habbo—grooming and patting, shoeing, and harnessing the horse all happen by letting the others know that this is happening to the horse at the moment. Becoming a horse, on the other hand, is indicated to others by standing within the bounds of the marked stable—a spatial bodily ‘gesture’ that has been turned into a sign of a particular role-taking initiative. The role of being a horse is further symbolically indicated by selecting a brown skin colour for the avatar and by typing the breed, temperament, and age of the role-played horse in the avatar’s description (an identity indicator accessible to others through clicking on the avatar). The role of ‘being a horse’ is often reinforced by making ‘horse sounds’ and communicating the feelings (via undecipherable utterings or descriptions of the sounds and feelings of the horse marked with an asterisk at the beginning and end of the description to separate it from normal conversation), but this kind of communication is not necessarily involved in the case of all ‘horses’—central is that the interavatar communication is a noncommunicative act in its symbolic sense. The horse sounds and silence both act as noncommunicative markers in the intentionally framed human–horse conversation of gestures, or the staged nonsymbolic interaction that paradoxically occurs through symbols. A horse has various ways of accepting a new owner, most often indicated by walking close to the avatar that asked for the permission and responding to the commands the avatar gives to the horse. Here the textual and the nontextual 57 Hebdige, 58 Goodlad

108

1979. and Bibby, 2007.


communication intertwine seamlessly to perform a virtual trans-species agreement on the central relationship needed for ‘riding’. Horse owners can then command the horses for riding sessions on dedicated riding grounds—these are usually other rooms connected to the main stables by teleports, allowing for a spatial bridge between them. The riding grounds are enacted by decorating the connected rooms as different kinds of terrain. As the avatars cannot really ride on each other, the riding activity itself in Habbo happens by walking the avatars side by side. Like being gothic, the actualization of the virtual equine activities draws from not only the quasi-bounded digital space but also entire outside ‘equine worlds’59 consisting of interrelated activities and subworlds in horse breeding, riding, stable keeping, veterinary medicine, international horse shows, Olympic games, and wide publicity in newspapers and TV—mingled with smells of autumn forests and so forth. Just as importantly, the virtual horse subworld in Habbo draws from other virtual pet worlds (significant in popularity, such as www.neopets.com) and ‘real-life’ pet worlds, and hence comes into being as a mixed set of imaginative, communicative, and material resources. Drawing from the equine world and children’s playmaking practices, these teen girls have managed to alter (one could say subvert) the materialities of Habbo and effectively added a layer of imaginative existence to Habbo that never was designed into it. The virtuality of virtual worlds comes into being in multiple ways, some more digital, some relying more on imaginative add-ons to designed components (see Table 4.8 below). Table 4.8: Actualisation of virtual activities requires a mix of both digital and imaginative resources. Resource Computer and business infrastructure Material Habbo features Social worlds and practices that users bring with them

Description Computer networks, service provider’s server software and hardware, user software and hardware, payment mechanisms, business models, regulation Add piece of furni, move and rotate it, change wall and floor colours, speech bubbles, avatars Children’s playmaking, equine worlds, virtual pets, gothic subcultures including material arrangement, coordination, communication, and imagination

Both theses stories of goths and horse aficionados in Habbo are examples of how the Habbo bits and pixels get a new life in the hands of the users. These did not fit the content creation scheme in the previous subsection and were thus worthy of another subsection. The content creation scheme in the previous subsection took a fairly material standpoint and was based on signals of user action available to the technical system. However, imagination in action and styles of chat are aspects of user practices not available to the interactive system, thus they have persistence on a different scale, compared to content creation in subsection 4.4.1. The next subsection structures the different pathways that user tweaks can take in Habbo. 59 Becker,

1982.


109

4.4.3 Developer Reactions to User Contributions Whereas the model of user-created content shows different kinds of content on the dimensions of persistance and visibility, it does not include anything about the dynamic relationships between different kinds of content. During my interviews with developers and users I came across a lot of examples where the users picked up an idea started by developers and turned it into something else. Sometimes it happened the other way around—users brought in something to Habbo that the developers had not considered, but that eventually became a feature or strongly shaped later feature development. I find this slowly paced dialogue between developers and users through the Habbo material very interesting to follow and worth an account. Let us take the example of the bar desk. So far it has been established that one built-in developer–user relationship is that developers create furniture and floorplans and users make rooms from a floorplan and purchased furniture. However, the designed flexibility of a bar desk made it suitable to divide a room in to smaller parts, so users started making labyrinths. This was communicated through the discussion forums on the fansites and became noted through the room popularity mechanisms as labyrinth rooms became popular. Users also realised that creative use of two bar desks and a bar desk door can create a restricted area in one corner of a room. Then it was up to the room owner whom to allow through and perhaps to charge admission if the restricted area or the desired location accessible through teleports in the restricted area was interesting enough. The developers followed the user practices and noted the popularity of room dividers and made both more furniture with similar functions and new floorplans with pre-made sections, even one with an island in the middle. These were published in the furniture catalogue, to be purchased by users and taken into use in various ways. Figure 4.9 shows a schematic of this. In a similar fashion to the bar desk, here are some examples of how users transformed Habbo: • users turned the dice furniture into bingo halls and casinos, and used the dice as the necessary element of chance in some other Habbo games; • whereas developers thought users would use the teleport to create virtual homes, with several interconnected rooms (kitchen, living room, bed room, etc.), users thought of connecting the teleports both in serial (multi-room mazes) and using many of them in the same room to create room hubs (travel agencies, teleporting centres); • users took individual furni and created mass effects by applying many of them: many green rugs on the floor became outdoor grass; a flower and suitable wall colours completed the transformation of an indoor hotel room to an outdoorsy space; • users agreed on that a certain avatar skin colour would mean a pretend play horse; horse movements and thinking communicated with an asterisk; and • users imported professions into Habbo, creating roleplay around professions like (and rooms decorated in that theme) nurse/doctor and hospitals, policemen and police headquarters, teachers and schools, TV-hosts and studios, etc. 110


Figure 4.9: An example of pathways through an interaction arena in Habbo.


111

Based on my empirical data I found four different responses of the developers to these tweaks and creative developments by users. Some user practices 1. covered an area not previously considered and were readily incorporated into technology and became built-in features • Room Categories, Furniture Trading, User Groups, Fansites 2. shaped the design of already planned features • Habbo Pets, new furniture lines (e.g. gothic furniture, Asian furniture) 3. were (so far) left alone as tweaks • Habbo professions, user ‘cooperatives’, roleplaying as horses 4. were incorporated into policy and shaped community policies • hacking, netphishing, fansites In some cases the user practices were completely transformed, and in other cases the user practices were only partially incorporated and live on as creative parallels. For instance, the developers considered virtual pets early on in the design of Habbo: “. . . already then we decided that there should be pets, we should have a moose that sits in front of the TV and drinks beer when left alone, and stuff like that. . . ” (Interview 11 Apr 2005). However, due to the decreasing age of the target user group, the pet feature was not implemented according to the initial vision. The pets became friendly cats and dogs, to be taken care of like tamagotchis. Similarly, a matchmaking feature, which was developed early on for an older target group, was also withdrawn at some point, and neither did the designs of new public rooms relate to pub life or drinking, like the first public rooms. On the other hand, the creative parallel to pets, role-playing as horses (article VII) lived on. The fourth category, how user practices shaped community policies, is outlined in the next subsection.

4.4.4 Fansites in Community Management Based on interviews in 2004 with the community director at Sulake, the head of all regional community managers, and the Finnish Habbo community manager in 2005, it became clear that fansites are an important part of community management as well. The community managers maintain close relations with the biggest fansite authors, and provide them information about coming features and changes in advance. At the time of the interviews, several of the largest Finnish fansite authors had visited the Sulake headquarters in Finland. Not all Habbo website authors write positively about Habbo; some are fairly critical, some are just underground, and others share tips about hacking in Habbo. Also, scamming sites have emerged, where someone copies the layout of Sulake’s website in order to trick users into inputing their usernames and passwords on a fake website. In addition to the community building effects that the fansites60 have, 60 The emergence of Habbo- or Sulake-critical websites and the Habbo scamming websites puts question marks on the label ‘fansite’ for all of the Habbo themed sites created by Habbo users, as some of the authors of these websites are clearly not fans of Habbo or Sulake. However, for reasons of simplicity I call all non-professionally authored Habbo-themed websites fansites.

112


they also pose two risks to Sulake. First, from Sulake’s perspective, some fansites disturb operations as they intentionally share information on how to hack Habbo or perform other Habbo stunts that are prohibited by the rules given by Sulake. Second, fansites like to use Habbo graphics, modify them, and post screenshots. This caused fears at Sulake that parts of the Habbo brand could enter the public domain at some point. To manage these fansite risks, Sulake created the notion of official Habbo fansites and a policy for them. By agreeing to this Habbo Official Fansite policy, the fansite authors promised to follow the rules of Sulake in return for their established status in the Habbo communities. In 2009, there were 130 official fansites all over the world. The policy enabled easier counteractions against fansites that behave badly. For instance, in 2005, when rogue fansites were discovered, the normal procedure at Sulake was to send a formal letter to the Internet service provider of the fansite, threathening legal actions if the fansite content was not removed from the web hotel. At that time, rogue fansites were discovered by the rate of 200 a week. By following the fansite phenomenon over a number of years, it is interesting to note that some of the functions that they perform in the Habbo communities have shifted between developers and active users. During the early years the core of Habbo, the virtual hotel, was more chat-based than today. The hotel featured a sychronous chat, but there was no place in the hotel for hotel visitors to have persistent discussions, share info about popular rooms, or record memories of events. Such features could only by served by asynchronous technologies, such as ordinary websites, discussion forums, so-called online guest books, picture galleries, gallup polls, and other voting mechanisms. The first fansites in the Finnish community emerged a few weeks after the launch of Habbo, which at the time was called Hotel Kultakala in Finland. One of the first popular fan-authored sites was called www.hotellikultakala.com, made by a character named Depis, and featured interviews with Habbo visitors, visitor profile galleries, pictures, guest books, and top lists of hotel rooms and fansites. This site called itself ‘unofficial Kultakala communitysite’,61 an implicit reference to the then official site: Kultakalan Kuvalehti,62 which was produced by a team at Sulake. Two major function shifts took place, one in April 2003 and another in late 2006 and early 2007. In early 2003 it became apparent for the then still relatively small headquarters (a few dozen people) that the online magazine production was too much work. In April 2003, the 21st and last issue of the official online fan magazine was published, while at the same time a weekly online newsletter was started. The last few issues of the fan magazine were co-produced by Sulake and authors from the Finnish fan community. A year later, as the first contest to become official fansites in Finland was held, a few of the fan authors, who had written texts in the official magazine, had started fansites of their own. Except for official information about the hotel that the newsletter provided, the other functions that the official 61 English

translation of ‘Epäviralliset Kultakala yhteisösivut’. kuvalehti can still be found in the Web Archive when searching for www.kultakalankuvalehti.com. Twenty-one issues were published between March 2001 and April 2003. 62 Kultakalan


113

community site performed had shifted from developers to the user community. This fan magazine and fan website reorganisation co-occurred with an increasing number of users, but with a decreasing average age. In the early summer of 2003 a guide written for parents of Habbo users emerged on the official Habbo website. These transformations signalled the end of a close developer engagement with the community. The tone of voice of the newsletter was different from the fan magazine; it was more written for the community, not with the community as it was previously. Inside Sulake, community management became more separated from game development. In late 2006 and early 2007 Sulake launched community platform extensions to the virtual world. All Habbo characters got homepages in the community and features to support groups of users in Habbo were launched. Previously group membership was only a communication signal to others in Habbo through a oneline character description. The change enabled groups of users to share a group homepage and discussion forum. These changes brought a lot of the common fansite features in to the Sulake-operated platform—for instance, visitor profiles and discussion forums. This time functions shifted from the community to the developers. The changes in the Sulake community platform to make it easier to discuss Habbo in Sulake’s service changed the landscape for fansites. Instead of focusing on any kind of persistent discussion, fansites got to focus on more specific discussions and in particular, such discussions that could be too delicate to have on Sulake’s Habbo site: for instance, the exchange values of rare furniture, discussions about the negative side of Habbo, detailed Habbo feature histories, and trading of Habbo furniture outside the Sulake-controlled Habbo market.

4.4.5 Summary User-created content is also a resource for developers to learn about users, which was partially discussed under the heading of User Categorisation (subsection 4.2.6). The following gives more detail about the importance of user-created content to user involvement and wraps up my observations and arguments. Instead of taking categories such as developers and users for granted, this subsection critically analysed the developer–user boundaries in user-created content through a number of lenses: content creation capacity, with respect to developer intentions and reactions, as well as community management. User-created content is not only about the publication of original works, their legal ramifications, and economic value, but it is also useful to user studies and ongoing service design. This is because user-created content is also about • mediated communication with friends, colleagues, family, and strangers; • carrying out everyday life mediated by computerised devices, network technologies, and social media; • one (or many) particular development–use relationship(s) related to the social media service(s) in question; and • consumer relations as an audience for advertisements and campaign participants.

114


(1) In addition to established genre and content analyses of user-created content, the dimensions of content visibility and content persistence seem particularly promising. (1a) An analysis of the visibility and persistence of user-created content reveals variation among users and user-created content similar in genre and type, which can be used for user categorisation. For instance distinctions between public and more underground chatters, or event participants versus playmakers that create the events for others to participate in (subsection 4.2.1). (1b) Broadening the content perspective along the persistence dimension to include both more fleeting content, such as chat, and more longer lasting content, such as features that developers make, reveals a useful framework for understanding usercreated content and relevant stakeholder groups. This visibility/persistence content analysis framework allows for an analysis of content creation capacity, that is, a shared emergent attribute of social media stakeholders such as various user groups, developers, moderators, and advertisers. This perspective, grounded in content creation, helps put different social media stakeholders on ‘the same page’ in analyses and complements a priori definitions of these stakeholders. This contributes to an expanding paradigm on content analysis in Internet research.63 (2) The actualisation of virtual activities requires not only digital content created by users, but also things invisible to the computer system, such as the imaginative resources that users bring with them and draw on. User-created content is not just about a developer-created platform where users can create content and make modifications, but rather users bring their own background, cultures, and personality to the system, without which the service could not work in practice. These new layers of virtual existence are important for both user studies and post-launch service design. (3) Some cases of user-created content turn in to a fruitful co-creation and a slowly paced material dialogue through technology between users and developers, while other user contributions are left as tweaks, or shape community policies. Some, but not all, user-created content becomes a design inspiration for service developers and is transformed into new features. This is important to the design process and the design outcome. Depending on the service lifecycle, the developers are more or less dependent on user-created content. If user-created content is to be actively supported, developers better leave hooks, openness, and flexibility for the users to chime in. Likewise it seems a worthwhile pursuit for developers to follow what users do 63 Herring,

2010.


115

in the service, and what kind of material reactions come from the users. It is not obvious what user-created content and user practices will end up as new features or be left alone by the developers; there are different developer reactions and many potential pathways for user contributions. (4) Even though user intermediaries, such as fansites, were crucial to Habbo’s success early on, the functions they play in the user communities have shifted several times between developers and users over the years. Important user community information was co-created first in the developer-edited fanzine, then on both official and unofficial fansites edited by users, and finally part of the service, as group homepages and discussions became possible within the Habbo service. This story about fansites shows the importance of user intermediaries, especially when developer resources are low, but also the attempts by developers to incorporate crucial parts of the environment. To identify, understand, and possibly support or benefit from these boundary shifts, categories such as developers and users cannot be taken as given, but rather must be understood as mutually constitutive and co-configured.

116


5 Conclusions

Broadly speaking, I argue that both research and practice of user involvement can benefit from a better understanding of design and use context dynamics, user categorisation, rhythms in development, and users’ contributions after market launch. More precisely, the central conclusions of this study are the following: 1. The unclear role of informal engagement and personal experience in changing design and use contexts can be resolved by considering shifts in developer–user social distance. This concept is developed in this study and refers to the degree of uncertainty and familiarity between developers and users based on diversity, differences, participation, and indirect contact. 2. User categories in design are not, and should not, be based solely on representiveness, since development organisations’ user categorisations are cumulative, emergent, and strategic. The adequate balance between strategy and representativeness changes over time in social media development and can be assessed by examining social media characteristics and developer–user social distance. 3. User involvement guidelines should be uncoupled from the assumption of stable and orderly project phases. Instead they should pay attention to neglected dimensions in the design context, such as multiple socio– technical rhythms, developer–user social distance, and the cumulation of knowledge about users. 4. In assessing whether, when, and how to involve users in further service development after market launch, their content creation capacity becomes a prime issue. Content persistence, visibility, tweaks, pathways, and user-operated services are among the key issues in analysing content creation capacity and user contributions. The following two sections elaborates on these conclusions and summarises the main contributions and their implications.

5 Conclusions

117

• • •

5.1 Summary of Contributions This case study contributes to a better understanding of • social media service evolution (section 4.1), • how user categorisation could be developed in social media and beyond (section 4.2), • rhythms and change in user involvement practices (section 4.3), and • how creative use and other user contributions influence social media services (section 4.4).

Shifts in Developer–User Social Distance: Developer Subjectivity and Service Evolution Developer subjectivity, for instance, a developer’s own use of a particular product or service and resulting first-hand experience, is poorly considered in guidelines and other advice on user involvement.1 Much writing on user involvement starts with the assumption that a developer is not a representative user and can therefore not trust his or her own gut feelings with respect to design choices. The other extreme opinion is also common, that is, developers are competent members of a community of practice and their personal experience is perfectly representative. In contrast, I argue that developers can lean on their ideas about use and experience of use, but that it depends on how familiar the developers are with the users and the use practices—what I call here the developer–user social distance. The concept of developer–user social distance emerged from this case, as I had the opportunity to learn about a social media company’s user involvement practices over several years and I became witness to a gradual, but significant, change in how users were involved in design. Early on the developers could draw upon their own experience as players and direct participation in the user communities, but, as the user base expanded internationally and extended to a younger demographic, they were forced to resort to more indirect forms of user involvement. The change in the forms of user involvement became visible as I distinguished and compared a number of stages in the service evolution of Habbo (subsection 4.1.5): Table 5.1: Habbo Service Evolution, from subsection 4.1.5. Concept

Beta

Expansion

Complexity

Competition

1999–2000

2001–2003

2004–2005

2006–2007

2008–2010

The development company gradually adopted more and more advanced user research methods, and after 10 years claimed to be a company driven by user data. For instance, usability evaluations and market surveys entered development 1 Visible for instance in debates about ‘heroic design’, personas, the timing of field studies, and biases towards formal methods in certain user-centred design approaches (chapters 1 and 2).

118

5 Conclusions

practice about four years after market launch. The Personas method was adopted still four years later. Instead of speculating whether earlier adoption of established user involvement methods would have improved the service or made it worse— an unresearchable topic—I decided to investigate how it was possible that the company succeeded without the textbook methods and what instigated the change to adopt them anyway later on. The usual suspect would be ‘more resources available because of organisational growth’, but my research showed that this was only half the truth. As outlined in section 4.1, the developer–user relationship changed significantly. With an increasing number of users, more features, and geographic expansion of the service, the diversity of use practices also increased. The younger demographic of the users brought increasing differences between developers and users. Developers’ active participation in use communities decreased, and volunteer users’ participation in development and moderation waned. The role of the fansites changed as certain discussions about Habbo could be carried out in the developer-provided forums. I conceptualised the above changes in the development as changes in social distance between developers and users. As developed in subsection 4.1.6, shifts in developer–user social distance refers to changes in uncertainty and unfamiliarity of the other group’s practices, resulting from a combination of changes in (1) diversity of use practices, (2) differences between developers and users, (3) direct developer participation in use practices and vice versa (direct user participation in development practices), and (4) indirect contact between developers and users through both social and technical mediators. This case is an example of self-centred design being adequate, but within certain limits. To convey the limits, the sensitising concept of developer–user social distance is proposed. As long as the distance is small, one can posit that self-centred design and informal user engagement can work, but as soon as the developer–user social distance grows, more effort is needed in terms of user involvement to bridge the emerging gaps. It also works the other way around. In many product and service design cases, the initial developer–user social distance is broad; however, as users engage in development and personal contacts develop, the developer–user social distance decreases, which then opens up possibilities for the use of more informal, potentially lighter and more first-hand methods.2 While the context of use has been in focus since the beginnings of user-centred design (sections 2.1 and 2.2), the design context has been found between the lines and in the margins until recent debates.3 The concept of developer–user social distance brings these two, design context and use context, together. It has the potential to overcome what has been described as a ‘heroic view’ of design, where developers are understood in too simplistic notions of either omnipotent heroes or 2 With regard to recent developments in the Habbo story, it is worth noting here that while these more informal methods contributed to the success of Habbo, the adoption of more formal methods have not corresponded to significantly greater success, as indicated by the ideas of usability maturity models. On the contrary, recent decline in monthly users and lapses in moderation casts a shadow over their adequacy, but as tens of other factors contribute to this development, it is difficult to estimate the effectiveness of more formal methods. 3 Dourish, 2006; Svanæs and Gulliksen, 2008.

5.1 Summary of Contributions

119

malevolent devils.4 Applying the concept also might reduce the risk of believing that design inspiration must be located solely in the use context, which a narrow reading of various user-centred design guidelines might lead one to believe. Already in 1991, Grudin highlighted various gaps between developers and users, but despite being well cited, the paper has not received much follow-up in terms of detailed empirical investigations and related conceptual development in HCI.5 This thesis contributes to showing how these gaps come about and develop, even within the context of developing the same product or service. That is important to understand, as undertaken user-centred bridging activities influence the developer–user social distance, which then has a consequence for subsequent user involvement activities. Muller has suggested interpreting HCI as a third space, between developers and users.6 This thesis contributes in putting the focus not only on the shared activities, but also on the different backgrounds from which developers and users come when they meet in third spaces.

User Categorisation Based on Developer–User Social Distance: Cumulative and Strategic This research on user categorisation concerned both the nitty-gritty details for particular studies and the overall strategies for using specific user categories to direct design efforts. Sometimes developers discuss various user categories as their future target groups or audiences; other times different designations of users becomes a question of representing various established user groups in design. The point being that the user is always present in the design process, not in a physical sense, but in the sense that it is impossible to design anything without an idea of how the end result is going to be used (section 3.1). In more traditional design contexts, where developers are far away from users, both geographically and socially, user categorisations (such as typical users, scenarios, personas, user stories, storyboards, and use cases) have an important role throughout the design process. The resulting user categories and representations of use convey the basic elements of the service or product to be (chapter 2). Compared to such contexts, this case study turns user categorisation and representation on its head.7 Early on in the service lifecycle, when the developer–user social distance was small, there was little need for one set of user categories, as representative of the complete range of users and use practices, to drive the design. The service itself, easily accessible to and used by developers, represented users and their activities well. Online forums and web analytics complemented developers’ own first-hand experience. It was not until later on in the service lifecycle, when the developer–user social distance was broader, that more strictly defined user categories became necessary as both a communication device for user knowledge and a strategic steering device in development. It became evident that the development organisation’s user categorisations were cumulative, emergent, and strategic. Focus changed from a ‘cool hangout on4 Stewart

and Williams, 2005. 1991. 6 Muller, 2003. 7 The third research question: How does user categorisation change with social media? 5 Grudin,

120

5 Conclusions

line’ and typical user practices, to catering to younger users and their parents, to cost-efficiency, multi-sided business, and global competition. Whereas early user practices and their (implicit) representations drove business, later the user representations became means to implement business strategies. In contrast to typical psychology-based approaches in HCI, the aim of Sulake’s user categorisations was never to understand users and their practices per se, but the key criteria of the user categorisations varied with respect to changes in the development–use relationship. As this study observed that Habbo development was shaped not just by users and their practices, but also by emerging business models and concerns (e.g., cost savings and revenue generation from user activities and business deals), also the user categories sometimes incorporated business- or design-specific issues. It is therefore suggested that user categories are not characterisations of individuals but rather portray certain characteristics of a development–use relationship. This finding is in line with some previous studies, both on developer–user relations8 and categorisation in general,9 but the consequences have largely been ignored by user research advice. In light of the above, I argue that the social distance between developers and users significantly influences what counts as sensible user categories and user involvement methods. I also contribute a number of ways of how advice about categorising users could take advantage of the opportunities present in a social media design context (subsection 4.2.6), e.g., multiple sets of user categorisations, a combination of service ‘census’ data and samples, and categorisations of user groups and user development pathways. The above arguments are analogous to one particular intellectual journey in the field of anthropology. Early on, the field started by exploring faraway people considered to be different; however, as the methodologies evolved, anthropologists came to understand that any exploration into ‘other people’ is also simultaneously an exploration into the anthropologist’s self and background. The more anthropologists learned about ‘the others’, the more they also learned about themselves. Anthropological analyses have moved from ‘objective characterisations’ of a specific people, to relational accounts of the involved cultures. If we take this development seriously this would imply that anthropology in product and service development could have two parallel tracks, one on users and another on developers and their assumptions. This would enable a comparative study on developers and users, as well as on their social distance.10

Rhythms and Change Beyond One Project: Cumulative Understandings about Users A dominant way of structuring guidelines and other advice on user involvement has been to assume project phases more or less tied to the software lifecycle, e.g. requirements analysis, implementation, testing, etc. (section 2.1). In contrast with 8 Hyysalo,

2009; Stewart and Hyysalo, 2008. and Star, 1999; Suchman, 1994a; Winner, 1980. 10 Barab, Thomas, Dodge, Squire and Newell, 2004; Cooper, Hine, Rachel and Woolgar, 1995; Dourish, 2006; Nyce and Löwgren, 1995. 9 Bowker


121

this literature, in this study I did not find the notion of project and project phases beneficial to structuring user involvement for social media or generalisation across cases. When studying how the developers’ user involvement practices evolved over several years,11 it became apparent that project phases did not structure the use of user research and involvement methods. Because of the immaturity of the social media market, social media advantages to user feedback, as well as the relatively low cost of updating the software service, I found multiple overlapping developments and rhythms in the user involvement activities. Software releases, user research, user experience and usability evaluations were sometimes more synched and sometimes in different pace depending on what made sense in different situations. Similarly, different parts of the software infrastructure evolved at different rates. For instance, some features had a slow rate of change because of low business priority, difficulties in implementation, or reliance on external frameworks, e.g., credit card payment processes. Prior method-use history shaped what was sensible consequential method use— e.g., after the main contours of Habbo had been usability evaluated, usability evaluations turned towards smaller details. Similarly, after overall user mappings, the following enquiries fine-tune specific issues that remained open. This practice can be interpreted as a sensible ongoing tailoring of methods that becomes necessary when dealing with complex and changing phenomena. The key question hence becomes how to enrich the knowledge in the organisation and how to meet the present and long-term key concerns in service development and the organisations doing it. Given this evidence, I argue that guidelines and other advice on user involvement should be uncoupled from the assumption of stable and orderly project phases, both with regard to a notion of a project that always starts from scratch and the idea of a standard type of project that can be stabilized. In a design situation where the focus is on extending an existing service, it may be that no new user research or participation is needed, nor any questioning of the alignment of design goals and user needs—contrary to common assumptions in the communication of participatory and user-centred design principles. On the other hand, changing development contexts may render established knowledge about users obsolete. ‘Text book’ approaches to user involvement and engagement that lean on the notion of a single, stand-alone project have over-simplified product and service development. User involvement should be structured by the actual conditions in the development context. Relevant actual conditions in this study were (subsection 4.3.5) developer–user social distance, organisational specialisation and internal rhythm, degree of business/mission criticalness, project scope, and relevance of existing knowledge about users. This can be compared with similar conditions proposed by Grudin as well as Svanæs and Gulliksen.12 11 The second research question: How do social media developers’ user involvement practices evolve over time? 12 Grudin, 1991; Svanæs and Gulliksen, 2008.

122

5 Conclusions

After Market Launch: Community Contributions and Developer–User Dialogue The literature on user involvement is fairly silent about user contributions after market launch. In many cases only projected future use or limited use during evaluation is considered as influences to a product or service, not actual use (chapter 2). This study presents a number of new openings that develop user involvement into considering use after market launch. Another challenge is that even research that opens up the nuances of user practices in social media rarely takes a critical look at the developer–user boundary, and categories such as developers and users are often taken for granted. In contrast, this study looked beyond these given categories and studied the actual roles of different stakeholders in the ongoing development of a social media service. Based on an extensive analysis of what users do in Habbo and how features change,13 I observed two significant parallels between content created by users and features created by developers. Both user-created content as well as features could be analysed with respect to their persistence (fleeting or longer lasting) and visibility (to a smaller group or all users) in the service. By pursuing this analysis on these dimensions of persistence and visibility, here called content creation capacity (subsection 4.4.1), I found that it distinguished, not only between the activities of various different users and developers, but also between moderators and advertisers. Thus I argue that content creation capacity is a useful framework for understanding the role of user-created content in service development. This concept helps put different social media stakeholders—such as developers, moderators, users, and advertisers—on ‘the same page’ in analyses and complements a priori definitions of these stakeholders. The extended timescale of this study allowed for an analysis of what I called a slowly paced dialogue between developers and users through the Habbo material. I examined how users tweaked and further developed practices around some features offered by the developers and conversely how developers picked up emerging user practices and implemented them as features in the hotel. Two particular cases of user creativity, the emergence of goth subculture and the creation of space for fantasy play by horse-interested girls, demonstrated how users tweak, subvert, and create new layers of virtual existence around a service. In addition, I tracked the pathways by which user activities fed in to further service development, noting that not all user-created content was incorporated in features but that some contributions were left as tweaks or shaped community management policies. These analyses of features and exchanges also contribute to an expanding paradigm on content analysis in Internet research.14 The relationship between the company and its user communities was also subject to continued experimentation and reform. I outlined the roles a variety of users played and the functions that they performed to make the service work, as well as how functions (e.g., moderation and support) shifted between developers and users (sections 4.1, 4.4, Appendix 5.2). Early on, active user volunteers helped 13 The first research question: How do users’ actions in and around a social media service shape its design after market launch? 14 Herring, 2010.


123

out in areas where the service developer company had no resources, but later moderation and discussion were brought in-house and in part automated in an attempt to reduce uncertainties. Key questions to the organisation of user possibilities in influencing service evolution include who hosts, maintains, and controls the rights to activities and outcomes of user-run, developer-run, or interconnected third-party blog/forum resources and services. A fruitful example from this case that addresses the scalability of participation in social media is the form of developer-user collaboration conceptualised as community journalism (sections 4.1.6, 4.2.2, 4.2.5). The developers encourage users to report and write articles about important topics in Habbo. Instead of the developers having to find out what matters most to the users through expensive user research, developers can rely on users doing the work. The role of qualitative and quantitative inquiry is then transformed into a means for source critique of what the users write. • • •

5.2 Implications Are we witnessing the funeral of user-centred design, as some researchers suggest?15 The argument has been that user-centred design stems from a time when the world was much different compared to the one we live in now and that the practice has not kept up with the pace of time. If one understands user-centred design as up-front user research, mainly interviews and observations, then it is easy to dismiss it. However, if one takes the approach suggested in the ISO 9241–210 standard,16 and makes a few modifications, user-centred design seems plausible in the context of social media. The modifications needed would be to (1) update the notion of social environment to mean community activities instead of job function and work practices, (2) note that community responsibilities can determine user goals and tasks in the same way job responsibilities can, (3) add that the principles also apply to re-development and actual use after market launch, and (4) take heed of the discussion on developer–user social distance, e.g., make explicit the special case of how sufficient user involvement can differ when developers also use the service themselves. Perhaps the broadest alternations needed to the standards and other user involvement advice concerns the recommendations for involving users. The characteristics of social media provide a number of direct and indirect ways of involving users (digital trails and feedback mechanisms) not previously considered in the 15 Norman,

2006; Norman, 2005; Zimmerman, 2011. ISO 9241–210 suggests six principles and four design activities. Principles: (1) the design is based upon an explicit understanding of users, tasks, and environments; (2) users are involved throughout design and development; (3) the design is driven and refined by user-centred evaluation; (4) the process is iterative; (5) the design addresses the whole user experience; (6) the design team includes multidisciplinary skills and perspectives. Design activities: (1) understanding and specifying the context of use; (2) specifying the user requirements; (3) producing design solutions; (4) evaluating the design. ISO, 2010. 16 The

124

5 Conclusions

ISO standards.1718 These new ways of involving users also highlight a number of contingency factors not covered by the ISO standards.19 This case study questions the validity of lifecycle stage as a relevant factor and confirms the following contingency factors (of which the ones marked with an asterisk are mentioned in the ISO standard): • degree of specialisation in development organisation20 • degree of business/mission criticalness of the feature in question (*) • variance in use practices (* wide task spectrum) • degree of regular use21 among developers • developer–user social distance • degree of blind corners22 in the service • degree of relevant cultural maturation23 • degree and variety of users’ digital trails that developers can access • degree of the user community primary activities that are covered by the service24 • degree of integration with other services in the social media ecosystem • degree of relevance of existing knowledge about users. Basically what this implies to managers is that ‘being in the beginning of a project’ loses its inherent meaning in relation to user involvement strategy. Instead of assuming that ‘the early stages’ of a project means little knowledge about users and a wide-open design scope, for instance, managers could consider the above contingency factors and use the relevant ones to decide on the next steps for user involvement. The same goes for ‘later stages’ in the project, which is commonly a shorthand for assumptions regarding plenty of user knowledge, closed design 17 The ISO standard on usability methods supporting human-centred design, ISO 16982, covers the following methods for direct involvement: observation of users, performance measurements, critical incident analysis, questionnaires, interviews, thinking aloud, collaborative design and evaluation, and creativity methods. ISO, 2002. 18 The ISO standard on usability methods supporting human-centred design, ISO 16982, covers the following methods for indirect involvement: document-based methods, model-based approaches, expert evaluation, and automated evaluation. ISO, 2002. 19 According to the ISO standard on usability methods supporting human-centred design, ISO 16982, the factors affecting the choice of usability methods are lifecycle stage, characteristics of the users, characteristics of the task to be performed, the product or system itself, the constraints that affect the project, and the degree of expertise in ergonomics available in the development or evaluation team. ISO, 2002. 20 It is not about organisational growth, but the way how the organisation is organised, e.g., functional departmentalisation vs. cross-functional and holographic teams. 21 Use can be actual or imagined, depending on whether the service is launched or not. 22 When an increase in the number of users coincides with an increase of variance in use practices, the developer means to learn about users (visits, community discussion, statistics, and other activity aggregation mechanisms, etc.) can only cover parts of the actual use, which results in increased uncertainty of actual use practices, that is, more blind corners in the service. 23 In the Habbo case the developers leaned on user interface genres from the 1980s, especially video games like Commodore 64 and similar console gaming platforms, and gameplay patterns from those days. Habbo developers drew other aspects from multiplayer online games in the late 1990s, both in copying and avoiding such gameplay design patterns. 24 Compare for instance Habbo with Flickr, where part of the activity is outside the service (photo taking) and part is computer-mediated (sharing, organising, voting, searching, etc.).

5.2 Implications

125

scope, and a fairly complete prototype. In the case of re-design projects there is always plenty of user knowledge in the organisation and a working product or service to evaluate in all project stages—it is not mandatory to wait until after development. These results imply that the same gentle critique towards usability evaluations that Greenberg and Buxton provided (subsection 2.2.1) could be extended towards usability maturity and process guidance in the ISO standard on human-centred lifecycle process descriptions (subsection 2.2.3). While they have a ‘significant role to play when conditions warrant it’, naive application can be ‘ineffective and even harmful’.25 Some of the time it is not safe to start with a field study or user categorisation. This is not against what the usability maturity standard states, because a key process area is not equivalent to an actual process and the standard leaves ample room for selecting what to evaluate,26 but its current format and spirit easily lends itself to misunderstanding. Furthermore, one cannot assume that documented evidence of usability evaluation always trumps informal engagement: in the case of low developer–user social distance, it is precisely the opposite way around. The hope is that user-centred design can be enriched by making what is tacit more explicit, especially regarding the developer–user social distance. If one understands user-centred design to imply upfront user research in all projects, that understanding ignores the experience of the developer, which is more relevant the smaller the social distance between developers and users. The point in usercentred design is to define requirements from the users’ perspectives, which is easy if the developer is close to the user but much more difficult if there are several degrees of separation between the developer and the user. This is one way of stating that the designer-self is relevant and good design is designed from somewhere, not nowhere.27 Svanæs and Gulliksen developed a similar list28 recently with aspects of the design context as risks to usability.29 This work on tactical user-centred design highlights the importance of the design context, in addition to the mentioned well-known use context analysis. With respect to tactical usability, this case study emphasises that we need an analysis of the relationship between the use and design context, for instance, with respect to developer–user social distance and accumulation of user knowledge. A key theme for future research would be to engage in understanding the social distance between developers and users and its consequences to design choices, with regards to design space thinking, product flexibility, and user involvement. We need to know what heuristics seasoned designers use when they make design choices and how developer–user social distance, with associated uncertainty about user practices, influences those design choices. An important question would be 25 Greenberg

and Buxton, 2008, p. 1. Dalton, Anderson, Konrad and Shrum, 2008. 27 Suchman, 2002. 28 The organisations involved, their relations, and agendas; internal factors in the developer organisations; software development methodology and tools; maturity levels; internal factors in the client organisations; customer–developer legal relationships (e.g., contracts, tender); handover issues; organisational stability; lifecycle perspective; conflicting requirement. 29 Svanæs and Gulliksen, 2008. 26 Glazer,

126

5 Conclusions

what are appropriate measures for the social distance between developers and users, but that will require investigations of the size of a research program. By no means does this case live up to the participation criteria of participatory design (subsection 2.3.2), but the research field can still learn from this case. The forms of user participation (Table 4.7) encountered are very broad, compared to what method guides contain—especially with regard to indirect involvement, where users’ actions shaped the design as a side effect, for instance, regarding online community journalism and statistics of furniture sales and user-to-user transactions. This case also reveals how user involvement is different in social media design contexts. Involving users in more traditional contexts has been a much more explicit effort, as developers and researchers have had to go where the users are and ask permission to observe and get feedback. While some ways of involving social media users remain as indirect as before, feedback from social media users is at the same time genuinely direct. What people do in the service leaves digital trails accessible to developers, which can be interpreted as user feedback. Users might not think that they are being involved or formally invited—they will just not return if the service does not meet their needs. However, previously it took longer before a decision by a user to stop using a product or service reached the developers. The findings and conclusions about the strategic and cumulative character of user categorisations also have implications for the first steps in user-centred and participatory design projects and later user research in social media design contexts. Utmost care is needed as regards to embarking on user groups definition ‘expeditions’. First, if there is a low social distance between developers and users, there might not be a great need for well representative user categories, especially if the social media service in question already represents the user practices well. Second, a large number of users and broad diversity in use practices makes sampling and analysis much more difficult compared to other software design contexts. Partial and strategic user categories might be all that is needed, as well as a critical reflection on what to categorise: it need not be individual users at all. Considering a situation after market launch, the opportunities present in a social media design context are manyfold, but need to be applied wisely. Feedback from social media users, in terms of frequent data points of user actions, has a major consequence for user studies. The techniques commonly denoted as evaluation ‘with’ and ‘without’ users need to be supplemented with ‘evaluation with database access to users’. As both use and non-use is logged in the server database, alarms can be set to trigger if use practices change. These alarm triggers can then be used as starting points for subsequent user studies to explain the changes in use practices. However strange, or perhaps disciplinary inconvenient, it may sound, the experience from this case study suggests that acquisition and analysis of database logs should take place before embarking on either quantitative or qualitative studies on social media use after market launch. In broader discussions on participatory democracy, the market has been criticised as not being a fair indicator of citizen opinions, needs, and aspirations. The same applies to Habbo, since not all users can afford to purchase furniture from Sulake. However, since so much of the core community activity is computer-mediated

5.2 Implications

127

and loggable by Sulake, including the second-hand furniture market, the uncertainty of what the not-so-much-purchasing users want is markedly decreased. This is of broader importance, not only to similar social media services, but also to other future products and services, as more and more of society is digitalised. Emergent questions are what data from ordinary users’ everyday actions are logged, under which circumstances, and to the benefit of whom. The risk is that developers use web analytics in not-so-transparent ways, leading to a situation where only developers gets to decide on web analytics and users do not know exactly what is analysed, which increases the power asymmetry between developers and users and further complicates participatory design. Concerning methods of participation, this case sheds light on the dilemmas of a service developer that co-constructs a service that is used by millions of people. It explored the developers’ user categorisation processes, and the strategies that emerged to solve the problem of abstracting relevant user knowledge from the everyday life of individuals in their social contexts. These strategies are very much like localisation and generification processes identified previously,30 but use contexts as the local and software as the generic is too simple, as in the Habbo case there were many local–generic processes taking place at any one moment, e.g., service topology and language localisation, feature categories, and abstractions of users with respect to countries, lifestyles, and consumption practices. A task for future research is to compare localisation and generification processes across cases and application domains. Parallel research by Holzapfel31 gives further weight to some of the findings in this thesis, especially as regards to popular social media services tending to start out with developers who develop for themselves and their friends.32 He also suggests that the characteristics of social media offset the need for user research and discourages early prototyping. This study is more rigid with respect to the boundary conditions for these assertions and provides conflicting evidence on prototyping. On the one hand, certain features in Habbo were redesigned over the years, because it was possible due to the low cost of redesign and distribution, and also because it was needed as the complexity of the service grew. On the other hand, Habbo was based on two prototypes, Mobiles Disco and Lumisota. One apparent reason for relaunching the service was that the primary activity and service scope changed between the services: Mobiles Disco was a disco, Lumisota was a snowball game; whereas Hotelli Kultakala (which became Habbo Hotel) was a hotel. Concerning easy redesign, I learned that there was code that was hard to change and code that was more easily changed (subsection 4.3.3). Moreover, starting in 2006 Sulake also organised online user panels, whose task it was to evaluate design sketches and discuss them in an online forum, a typical example of prototyping with users. The evidence from the Habbo case suggests that the relationships between prototyping and social media is not clear-cut at all. In particular, hypothesis formulation on this topic needs to be specific about whether one discusses (1) interactive prototypes vs. design sketches, (2) hard-to-change fundamental features vs. easily 30 Pollock,

Williams and D’Adderio, 2007. 2008. 32 Also debated in a blog: Porter, n.d. 31 Holzapfel,

128

5 Conclusions

redesigned features, (3) whole service changes vs. A/B testing, (4) designer-created content vs. aggregated content based on the actions of thousands of users, and (5) tightly scripted vs. open-ended features. Based on the pathways of the Habbo service, it made sense to prototype in different design circumstances outlined above. Considering research on user innovations, this study contributes to approaches that emphasize intermediaries and feedback loops between developers and users.33 In line with previous research, this case also highlights a broad variety of intermediate actors between ‘regular’ developers and users who play significant roles, be they volunteer community moderators, fansite authors, NGOs, hackers, mobile phone operators, consumer ombudsmen, and so on.34 Like Stewart and Hyysalo— who broaden the functions of intermediaries to encompass the roles of facilitating, configuring and brokering—this study points to the various roles a variety of different users can take, but in the context of social media.35 While some studies on companies that rely on user communities for innovation point out the role of company-controlled user toolkits,36 this case demonstrated how user activity feeds into further service development without such user toolkits. Instead the solution space was partly open to users due to built-in technical flexibility and open-ended use in the service, as well as user contributions with standard image manipulation tools. Recent research on user innovations has opened up a new research front on a so far neglected side of user innovation, so-called user techniques, or practices that users develop to overcome restrictions with existing technologies.37 The fact that users’ actions, both in line with developer expectations and to the surprise of developers, shape products and services has been long known,38 but this has remained regretfully marginal knowledge. This case contributes to the understanding of what happens to the user contributions once they are communicated by a number of examples, and suggests different pathways. Future research includes finding out what kinds of pathways emerge in different domains, including an analysis of success factors and barriers for the user contributions to turn into user innovations. Whereas this case included more detail on the strategies of the developers to learn about the users, future work could deepen the analysis of the strategies and impact of the users who intentionally wanted to make an impact on the service. This case study discussed, for instance, direct feedback, community journalism, collaborative actions in the hotel, hacking, and tweaking as example means to make an impact. The findings from this social media case gain importance if two technical tendencies accumulate. The first is the increasing incorporation of functionality for group communication in software products and services in general. The second 33 Pollock and Williams, 2009; Stewart and Hyysalo, 2008; Van de Ven, Polley, Garud and Venkataraman, 1999; Williams, Stewart and Slack, 2005. 34 Bessant and Rush, 1995; Howells, 2006. 35 Stewart and Hyysalo, 2008. 36 Hippel, 2001; Jeppesen and Molin, 2003. 37 Stockstrom, Lüthje and Antorini, 2010, point to the example of how Jan Boklöv developed ski jumping by inventing the V-style in 1986, and report that similar user innovations take place within the LEGO field. 38 Hyysalo, 2010; Pollock and Williams, 2009.

5.2 Implications

129

is that developers of other products and services make use of cameras and other sensors. Both of these tendencies provide developers with additional digital trails of user activities and increase, not only the means for feedback and control, but also the demands for accountability and transparency.

130

5 Conclusions


One challenge for stakeholder analysis, as part of product development and research involving users, is to find and develop analytical frames with suitable levels of granularity. The ISO standards provide a framework called the context-of-use model,39 but important issues are left out and it has not been developed for technologies outside the work context. The framework used here comes from the work of Adele Clarke. It builds on a collection of existing models of social analysis developed by Strauss in the 1970s called Grounded Theory.40 This analytical frame, slightly tuned compared to Clarke’s version, includes the following actors: Table 5.2: Analytical Frame Developed in the Prima Project.

39 ISO,

actor

explanation

stakeholders of interest

individual, collective, implicated/silent

social relations and power structures

who is able to interact with whom and in which ways?

discourse formations

how do different stakeholders frame the use of specific applications?

norms, regulations, and rules

which behavioral scripts do different stakeholders suggest?

history, traditions, persistence

how has the application in question evolved, what traditions have emerged, and which information is kept and why?

interactional enablers

how is interaction made possible and what kind of information is hereby suggested?

communicational forms

how do stakeholders interact with each other?

business models

how is continuous work compensated?

non-human actants

what technologies are used to maintain and develop the service, and what technologies are necessary in order to be able to participate in the specific interaction?

1998; ISO, 1999; Maguire, 2001a. 2005; Strauss and Corbin, 1990.

40 Clarke,


131

Habbo Case Summary (April 2007) This summary was created by the author, applying the analytical frame developed in the Privacy-in-the-Making (Prima) project, of which the author was part of 2006– 2010.41 www.sics.se/projects/prima

Individuals • Current Habbo users42 – about 9 million globally, spread out mostly in industrialized (with either a good coverage of broadband access or lots of people) countries as follows: 25% Europe, 25% Asia and Australia, 25% US, 25% Mexico and South America. – 75% of the users are between 13–18 years old. Finland has more younger users, Japan older users. – On average, the users spend 30 minutes online per day. – In Habbo: Habbo staff, Habbo-guides,VIP guests, youth workers • Previous Habbo users not active any more – visible in the traces that they left in Habbo: inactive Habbo avatars (taken Habbo nicknames), empty rooms, Post-it notes – less visible, but still there: the ways in which they influenced Habbo development and features and playing culture • Habbo users’ parents – often the paying customer – children negotiate access with parents • Habbo founders: Sampo Karjalainen and Aapo Kyrölä – other particular Sulake employees • Consumer ombudsman (e.g., of Finland)

Collectives • Sulake – the company and the brand – the original developers (a handful, of which some work on other stuff) – game development: a team of 15–20, consisting of user interface programmers, server programmers, and graphic designers – marketing and user insight – country-specific operations (moderators, hotel managers), customer services – business 41 Bylund,

Johnson, Lehmuskallio, Seipel and Tamminen, 2010. is a practical impossibility to enumerate millions of individual Habbo users, so I resorted to a few aggregate descriptions here. 42 It

132


– sales • Sulake owners • Sulake’s competitors – Stardoll, Neopets, Club Penguin, Piczo, Miniclip, imvu, gaia Online. . . – MySpace, Bebo, IRC-Galleria, Windows Live Messenger, Facebook, The Lounge, YouTube • Sulake business alliances – Sulake has more than 150 payment mechanisms globally – payment: mobile phone operators, youth card companies, credit card companies, banks, and Internet payment systems – distribution of prepaid cards (in Finland: R-Kioski, Tiimari. . . ) – advertisers • Other alliances – Unicef – EOPH and Hubu—Elämä on parasta huumetta ry – Netari—Helsingin nuorisoasiainkeskus • Organised groups in Habbo (cf. guilds) – volunteer guides – groups formed around a fansite, e.g., authoring it or participating in forums – groups formed around a (popular) room, or set of rooms – groups as in ‘Habbo groups’ (since 2007), e.g., http://www.habbo.co.uk/ community

Implicated (Silent) Stakeholders • Other computer users in the same household – in Johnson’s survey (2004), 93% of the respondents reported that they log on from home – siblings and parents often share one home computer and one Internet connection • Significant others • Friends, school classmates, mates from hobbies – some participate in Habbo, some do not • Celebrities in national and pop culture: e.g., – in Finland’s Habbo there is a famous Habbo avatar named Vanhanen, after the Finnish prime minister Matti Vanhanen – in Finland there is a purchasable wall poster of the former president Kekkonen – there used to be a wall poster of the British queen


133

Discursive Constructions and Formations How do different stakeholders frame the use of specific applications? • it’s a hotel—hotel manager, hotel staff, hotel visitors • it’s a game—socialisers, explorers, achievers, killers43 • it’s child’s play—children with imagination and creativity, older people • it’s a chat—chatters do nothing serious, they just chat and don’t want to win or play the game seriously • it’s a community—people come there to meet others with similar interests; one can fulfill oneself as a community member; one starts as a newbie, continues to become a regular well-behaving member; and some very active become appreciated gurus • it’s a dating place—teenagers come there to meet potential girlfriends and boyfriends, to practice online with a different ‘face’; there is less need to manage school reputation • it’s a place for collecting and trading furniture:—similar to other collecting hobbies, such as football player cards, baseball player cards, musicians cards, Magic the Gathering, etc. • it’s a place where one can make up one’s own game—game makers and players (imitating TV-shows, sports, dice-based games) • it’s a place where one can hangout—meet friends, get to know other users • it’s a time-shared place—new mothers and other minority groups become visible during daytime, when the majority of the users are at school; older teenagers later in the evening. • there are threats to the game/community/hotel—hackers, script kiddies, scammers . . . grief players • it’s cool graphics online—well-designed (thoroughly thought out pixels) and enduring (8 years already) • it’s part of the new media scene—early on new media people used to hangout in Habbo • it’s new technology • it’s a new ICQ—the original developers wanted to make a cool new graphical ICQ • it’s business—a success story, employer-employee negotiations (HS, Taloussanomat) • it’s where the money comes from—employee, manager, boss, investor • it’s the future—future users • it’s a service that entices children to consume (HS/TS) • it’s a crime scene—some have stolen Habbo furniture and been sentenced for stealing virtual goods (Wikipedia) 43 Bartle,

134

1997.


• it’s a place for youth work (HS 22.6.2009) • it’s a means for online security education (Taloussanomat 8.2.2007) • it’s a means for advertising (HS 29.8.2005) • it’s a pop culture arena (Taloussanomat 15.1.2009)

Non-human Actants–Technologies What technologies are used to maintain and develop the service, and what technologies are necessary in order to be able to participate in the specific interaction? • users – computer hardware, keyboard, monitor, mouse, network card/modem – router, Internet connection, Internet service provider, at least 256kbps (guess), phone lines or other (wireless or cable) signal distribution network – electricity, distribution network, standard AC (e.g. 220V 50Hz) – operating system, Internet browser, Shockwave plugin • developers (Sulake) – Macromedia Director, Lingo, Java server, Fuse – IDE (integrated development environment), version control, and code management – office software – test-server • live site operators (country-specific operations) – effective servers, lots of bandwidth

Revenue Streams How is continuous work compensated? • Habbo credits – in-game currency – users buy Habbo credits from Sulake that can be exchanged into Habbo furniture • Habbo Club – monthly fee in exchange for more • Advertisements – companies pay Sulake to display ads in public rooms in Habbo – ads can be displayed ‘between rooms’, when loading a room – ads can be displayed on Habbo’s website • Customising Habbo’s Game Engine – Walt Disney Magic Kingdom, Coke Music/Studios/MyCoke (Coca-Cola)


135

• Hollywood Deals Habbo, the largest online virtual world for teens, today with Paramount Pictures Digital Entertainment, announced a licensing agreement to create virtual goods based on titles including ‘The Spiderwick Chronicles,’ ‘Beowulf’ and ‘Mean Girls.’ The deal will provide Habbo merchandising rights throughout the U.S. and Canada for all three properties. The partnership marks one of the first of its kind between a major motion picture studio and a virtual world commodity. http://www.sulake.com/press/ releases/2008-01-29-Virtual_goods_get_a_Hollywood_makeover.html • Venture capital

Social Relations and Power Structures Who is able to interact with whom and in which ways? • different power in relation to the technology – developers—can change Habbo architecture – project managers, art directors, and IT architects have more power than graphic designers, user interface programmers, server programmers – product board (representatives from business, sales, marketing, and development) decides roadmap – hotel manager—can create hotel events, competitions – moderators—can kick out disturbing users – users—can be in Habbo, decorate rooms – VIP guests—popular culture icons visit Habbo for 2 hours to chat with their fans – netari youth workers – EOPH drug info bus • different power in relation to room – room owner – room guests – users with access rights to move furni in the room • different power in relation to the ‘national Habbo community’ – – – – – – – –

136

the authors of the largest Habbo-themed fansites the insiders of the largest Habbo groups/forums the richest (most furni) users the most opinionated users in the forums the best game organisers the owners of the most popular rooms the most innovative room decorators Habbo radio amateur reporters


Norms, Regulations and Rules Which behavioral scripts do different stakeholders suggest? • Habbo Way (created by the developers) – You must not ∗ Tell your password to a Habbo, friend, staff member or stranger. ∗ Tell anyone personal information which could be used to locate you or other people in real life. ∗ Abuse, harass or bully other Habbos. ∗ Use hate speech or make rude comments about a Habbo’s race, religion, gender or sexuality. ∗ Use any programs to hack, script or edit Habbo in any way. ∗ Own, sell or run replica Habbo Hotels (retros). ∗ Steal from or trick Habbos into giving you their passwords, Habbo Credits or furniture. ∗ Discuss or take part in sexual acts with other Habbos. ∗ Act out or roleplay violent acts, even as a joke. ∗ Give away, trade or sell your Habbo account. ∗ Break the law or talk others into breaking the law. – You should ∗ ∗ ∗ ∗

Have fun! Hang out with your friends Make new friends Respect other people’s opinions and beliefs

• Privacy Policy, Terms of Use • design principles – non-violence, easy access and easy play, everyone can play • The hotel metaphor (as interpreted by the author) – ‘you’re a guest, behave as one’ – ‘you pay, it’s not your home’ • different hotel rooms suggest different metaphors (and therefore different scripts) – so called public rooms—created by the developers ∗ welcome lounges, spa, cafés, football stadium, entertainment (library, theatre, cinema), outside spaces (rooftop, gardens), games (Battleball, Snowstorm), Habbo club rooms, dance clubs and pubs, restaurants, lobbies, hallways – work as meeting places – the rooms decorated by the users—guest rooms (room categories from Habbo UK 06/2008) ∗ official hot room Appendix A: Stakeholder Analysis

137

∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗

quest archive chat, chill, and discussion rooms club and group rooms gaming and race rooms help centre rooms hair salons and modelling rooms maze and theme park rooms restaurant, bar, and nightclub rooms school, daycare, and adoption rooms theme rooms trading rooms all other rooms

– moderation ∗ as guides ∗ as policemen ∗ as judges

History, Traditions, Persistence How has the application in question evolved, what traditions have emerged, and which information is kept and why? • no formal process during first two years • since 2003—new releases 2–3 times a year – new features – brief user protests (‘vanha kala takaisin’) • no record of chat history is kept – old and not used Habbo nicknames and rooms have been cleared once or twice (at least in 2003) – previously popular rooms tell their stories on post-it notes • Habbo generations (the author’s interpretation) – first generation of Habbo users: started 2000–2002 and quit by 2005 ∗ had the chance to actively shape Habbo, created many legends – second generation: started 2002–2003 and quit by 2007 ∗ more organised groups (e.g., mafias, armies, and other clubs) were formed – third generation: started 2005 or so, really thinking of quitting now ∗ wasn’t there when Habbo was invented ∗ entered a fairly ready-made world and business – fourth generation: started 2007–8 ∗ takes the social networking side of Habbo for granted

138


Communicational Forms How do stakeholders interact with each other? • users – – – – –

chat (writing) and acting (moving avatar) in a room private messages self-presentation on Habbo home pages forum discussions in Habbo group home pages decorating a Habbo room, giving it a name, marketing it (event arranging, game making) – by creating and authoring Habbo fansites • users to moderators – help channel • developers to users (mass-communication) – weekly newsletters – competitions – Habbo website (rules, norms) • customers to the company – customer service, complaints, resolving of ‘stolen’ furni • users to developers (feedback) – – – – –

since 2007: feedback Habbo groups (e.g., Pilots of the Revolution) early on (2000–2 or so): e-mail fansite forums user–2-user communication (some developers closely follow it) aggregated actions: statistics

• advertisers to users – ads

Interactional Enablers How is interaction made possible, and what kind of information is hereby suggested? • User names, rooms, and virtual properties are conserved between playing sessions – persistent game world (no reboot) • No duplicate names – one can trust that the same person is controlling the same avatar from time to time • Habbo is a social proxy (3rd person perspective, same view to all) – it’s a multi-player game • Navigation Appendix A: Stakeholder Analysis

139

– users use the Habbo navigator to move between rooms – the rooms are divided into ‘public rooms’ and ‘rooms’, which have previously mentioned categories – users can purchase teleports to directly connect two rooms • Norms are propagated – by Habbo official website – fansites (forums) teach how to act in Habbo • Inside a Habbo hotel – One can chat/interact with others in the same Habbo location (Habbo room) – interpersonal distance (in the room) influences propagation of Habbo speech (bubbles) – users can use voice control to say, shout, whisper – room owner can share rights to move furni in room – users can manage the great number of users through a friends list ∗ the friends list provides Habbo location info – availability cues ∗ avatar tags explaining avatar agenda /status /interests ∗ avatar looks sleepy if it hasn’t spoken for a while – – – – – –

social status/group identificators (avatar tags and groups) audience control (door lock) persistant texts (Post-it notes) Habbo currency furniture trading automat chat filter

• data security: encryption of traffic between client and server, password mechanisms • Habbo user homepages – Habbo widgets – report inappropriate content • Paying mechanisms work • Larger than Habbo – national, international, and bilateral regulations, policies, acts, constitution, data protection act, copyright

140


References

Akera, Atsushi (2008). “Communities and Specialized Information Businesses”. In: The Internet and American Business. Ed. by William Aspray and Paul E Ceruzzi. MIT Press (cit. on pp. 28, 33). Akrich, Madeleine (1992). “The de-scription of technical objects”. In: Shaping Technology/Building Society: Studies in Sociotechnical Change. Ed. by Wiebe E Bijker and John Law. MIT Press, pp. 205–224 (cit. on pp. 45, 84). Akrich, Madeleine (1995). “User representations: Practices, methods and sociology.” In: Managing Technology in Society: The Approach of the CTA. Ed. by Arie Rip, Thomas J. Misa and Johan Schot. Pinter Pub Ltd, pp. 167–184 (cit. on pp. 4, 38, 45, 84). Allen, Christopher (2004). Tracing the Evolution of Social Software. http://www. lifewithalacrity.com/2004/10/tracing_the_evo.html (Retrieved 31/01/2012) (cit. on p. 29). Allen, Matthew (2008). “Web 2.0: An argument against convergence”. In: First Monday 13.3. http://firstmonday.org/article/view/2139/1946 (cit. on p. 27). Allen, Robert B. (1990). “User models: theory, method, and practice”. In: International Journal of Man-Machine Studies 32.5, pp. 511–543 (cit. on p. 38). Anderson, Chris (2006). The Long Tail: Why the Future of Business Is Selling Less of More. New York: Hyperion (cit. on p. 3). Andreasen, Morten Sieker, Henrik Villeman Nielsen, Simon Ormholt Schrøder and Jan Stage (2006). “Usability in Open Source Software Development: Opinions and Practice”. In: Information Technology and Control 35.3A, pp. 303–312 (cit. on p. 5). Andriessen, J. H. Erik (2003). Working with Groupware: Understanding and Evaluating Collaboration Technology. London; New York: Springer (cit. on pp. 29, 30, 34). Avison, D. E and G Fitzgerald (1988). Information systems development: methodologies, techniques, and tools. Oxford: Blackwell Scientific Publications (cit. on p. 3). Baecker, Ronald M (1993). Readings in Groupware and Computer-Supported Cooperative Work: Assisting Human-Human Collaboration. San Mateo, Calif: Morgan Kaufmann Publishers (cit. on pp. 29, 30, 34). Baecker, Ronald M. (2008). “Timelines: Themes in the early history of HCI—some unanswered questions”. In: Interactions 15.2, pp. 22–27 (cit. on p. 15).

References

141

Baker, Carolyn (2004). “Membership categorization and interview accounts”. In: Qualitative Research: Theory, Method and Practice. Ed. by David Silverman. 2nd ed. London; Thousand Oaks, Calif: Sage Publications, pp. 162–176 (cit. on p. 56). Baldwin, Carliss, Christoph Hienerth and Eric von Hippel (2006). “How user innovations become commercial products: A theoretical investigation and case study”. In: Research Policy 35.9, pp. 1291–1313 (cit. on p. 23). Balka, Ellen and Ina Wagner (2006). “Making things work: dimensions of configurability as appropriation work”. In: Proceedings of the 2006 20th anniversary conference on Computer supported cooperative work. CSCW ’06. New York, NY, USA: ACM, pp. 229–238 (cit. on p. 22). Bannon, Liam (1992). “From human factors to human actors: the role of psychology and human-computer interaction studies in system design”. In: Design at work. Ed. by Joan Greenbaum and Morten Kyng. Hillsdale, NJ, USA: L. Erlbaum Associates Inc., pp. 25–44 (cit. on pp. 16, 46). Barab, Sasha A, Michael K Thomas, Tyler Dodge, Kurt Squire and Markeda Newell (2004). “Critical Design Ethnography: Designing for Change”. In: Anthropology and Education Quarterly 35.2, pp. 254–268 (cit. on pp. 16, 121). Bargas-Avila, Javier A. and Kasper Hornbæk (2011). “Old wine in new bottles or novel challenges: a critical analysis of empirical studies of user experience”. In: Proceedings of the 2011 annual conference on Human factors in computing systems. CHI ’11. New York, NY, USA: ACM, pp. 2689–2698 (cit. on p. 5). Barkhuus, Louise and Jennifer A. Rode (2007). From Mice to Men - 24 years of Evaluation in CHI. ACM CHI’07 - Alt.CHI. (Archived at http://web.archive.org). http://www.viktoria.se/altchi/ (cit. on p. 15). Bartle, Richard (1997). “Hearts, Clubs, Diamonds, Spades: Players Who Suit MUDs”. In: The Journal of Virtual Environments 1.1 (cit. on p. 134). Bartle, Richard A (2004). Designing Virtual Worlds. Indianapolis, Ind: New Riders Pub (cit. on pp. 31, 56). Battarbee, Katja (2003). “Defining co-experience”. In: Proceedings of the 2003 international conference on Designing pleasurable products and interfaces. DPPI ’03. New York, NY, USA: ACM, pp. 109–113 (cit. on p. 18). Baym, Nancy (1995). “The Emergence of Community in Computer-Mediated Communication”. In: CyberSociety: Computer-Mediated Communication and Community. Ed. by Steve Jones. Thousand Oaks, Calif: Sage Publications, pp. 138– 163 (cit. on p. 36). Baym, Nancy K. (2005). “Introduction: Internet Research as It Isn’t, Is, Could Be, and Should Be”. In: The Information Society 21.4, pp. 229–232 (cit. on p. 59). Beck, Kent, Mike Beedle, Arie van Bennekum, Alistair Cockburn, Ward Cunningham, Martin Fowler, James Grenning, Jim Highsmith, Andrew Hunt, Ron Jeffries, Jon Kern, Brian Marick, Robert C. Martin, Steve Mellor, Ken Schwaber, Jeff Sutherland and Dave Thomas (2001). Manifesto for Agile Software Development. http://agilemanifesto.org/ (Retrieved 31/01/2012) (cit. on p. 17). Becker, Howard Saul (1982). Art Worlds. Berkeley: University of California Press (cit. on p. 109). Benford, Steve, John Bowers, Lennart E. Fahlén, Chris Greenhalgh and Dave Snowdon (1995). “User embodiment in collaborative virtual environments”. In: Proceedings of the SIGCHI conference on Human factors in computing systems. CHI

142

References

’95. New York, NY, USA: ACM Press/Addison-Wesley Publishing Co., pp. 242– 249 (cit. on p. 29). Benkler, Yochai (2006). The Wealth of Networks: How Social Production Transforms Markets and Freedom. New Haven [Conn.]: Yale University Press (cit. on pp. 7, 23, 24, 34). Bessant, John and Howard Rush (1995). “Building bridges for innovation: the role of consultants in technology transfer”. In: Research Policy 24.1, pp. 97–114 (cit. on pp. 69, 129). Bevan, Nigel (2001). “International standards for HCI and usability”. In: International Journal of Human-Computer Studies 55.4, pp. 533–552 (cit. on p. 17). Beyer, Hugh and Karen Holtzblatt (1998). Contextual Design: Defining CustomerCentered Systems. San Francisco, Calif: Morgan Kaufmann (cit. on pp. 5, 16). Beyer, Hugh R. and Karen Holtzblatt (1995). “Apprenticing with the customer”. In: Communications of the ACM 38.5, pp. 45–52 (cit. on p. 34). Bias, Randolph G and Deborah J Mayhew (1994). Cost-Justifying Usability. Boston: Academic Press (cit. on p. 19). Bijker, Wiebe, Thomas Hughes and Trevor Pinch (1987). The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology. Cambridge, Mass: MIT Press (cit. on p. 45). Bjerknes, Gro and Tone Bratteteig (1995). “User Participation and Democracy: A Discussion of Scandinavian Research on System Development”. In: Scandinavian Journal of Information Systems 7.1, pp. 73–98 (cit. on p. 20). Bjerknes, Gro, Pelle Ehn, Morten Kyng and Kristen Nygaard (1987). Computers and Democracy: A Scandinavian Challenge. Aldershot [Hants, England]; Brookfield [Vt.], USA: Avebury (cit. on p. 20). Björgvinsson, Erling, Pelle Ehn and Per-Anders Hillgren (2010). “Participatory design and "democratizing innovation"”. In: Proceedings of the 11th Biennial Participatory Design Conference. PDC ’10. New York, NY, USA: ACM, pp. 41–50 (cit. on p. 24). Blomberg, Jeanette (1987). “Social interaction and office communication: Effects on user’s evaluation of new technologies.” In: Technology and the Transformation of White-Collar Work. Ed. by Robert E Kraut. Hillsdale, N.J: L. Erlbaum Associates (cit. on p. 35). Blomberg, Jeanette, Jean Giacomi, Andrea Mosher, Pat Wall, Douglas Schuler and Aki Namioka (1993). “Ethnographic Field Methods and Their Relation to Design”. In: Participatory Design: Principles and Practices. Lawrence Erlbaum, pp. 123–155 (cit. on p. 16). Blythe, Mark A (2003). Funology: From Usability to Enjoyment. Dordrecht; Boston: Kluwer Academic Publishers (cit. on p. 18). Bogardus, Emory (1925). “Measuring Social Distance”. In: Journal of Applied Sociology 9, pp. 299–308 (cit. on p. 76). Bossen, Claus, Christian Dindler and Ole Sejer Iversen (2010). “User gains and PD aims: assessment from a participatory design project”. In: Proceedings of the 11th Biennial Participatory Design Conference. PDC ’10. New York, NY, USA: ACM, pp. 141–150 (cit. on p. 21). Botero, Andrea and Joanna Saad-Sulonen (2008). “Co-designing for new citycitizen interaction possibilities: weaving prototypes and interventions in the design and development of Urban Mediator”. In: Proceedings of the Tenth An-

References

143

niversary Conference on Participatory Design 2008. PDC ’08. Indianapolis, IN, USA: Indiana University, pp. 266–269 (cit. on p. 22). Bourdieu, Pierre (1984). Distinction: A Social Critique of the Judgement of Taste. Cambridge, Mass: Harvard University Press (cit. on p. 76). Bowker, Geoffrey C and Susan Leigh Star (1999). Sorting Things Out: Classification and Its Consequences. Cambridge, Mass: MIT Press (cit. on pp. 47, 89, 121). Boy, Guy André (1998). Cognitive Function Analysis. Stamford, Conn: Ablex Pub. Corp (cit. on p. 44). boyd, danah m. and Nicole B. Ellison (2007). “Social Network Sites: Definition, History, and Scholarship”. In: Journal of Computer-Mediated Communication 13.1 (cit. on p. 62). Brown, Barry and Marek Bell (2004). “CSCW at play: ’There’ as a collaborative virtual environment”. In: Proceedings of the 2004 ACM conference on Computer supported cooperative work. CSCW ’04. New York, NY, USA: ACM, pp. 350–359 (cit. on p. 29). Bruns, Axel (2008). Blogs, Wikipedia, Second Life, and Beyond: From Production to Produsage. New York: Peter Lang (cit. on pp. 7, 24, 25). Budde, Neil F. (2004). “There’s no such thing as an "average" user”. In: Interactions 11.2, pp. 54– (cit. on p. 83). Button, Graham, ed. (1993). Technology in Working Order: Studies of Work, Interaction, and Technology. London; New York: Routledge (cit. on pp. 16, 46). Buxton, William (2007). Sketching User Experience: Getting the Design Right and the Right Design. San Francisco, CA: Morgan Kaufmann (cit. on p. 18). Bygstad, Bendik, Gheorghita Ghinea and Eivind Brevik (2008). “Software development methods and usability: Perspectives from a survey in the software industry in Norway”. In: Interacting with Computers 20.3, pp. 375–385 (cit. on p. 5). Bylund, Markus, Mikael Johnson, Asko Lehmuskallio, Peter Seipel and Sakari Tamminen (2010). “PRIMA — Privacy research through the perspective of a multidisciplinary mash up”. In: Vem reglerar informationssamhället? Nordisk årsbok i rättsinformatik 2006–2008. Stockholm: Jure (cit. on pp. 60, 132). Bødker, Keld, Finn Kensing and Jesper Simonsen (2004). Participatory IT Design: Designing for Business and Workplace Realities. Cambridge, Mass: MIT Press (cit. on pp. 6, 21). Bødker, Susanne (1989). “A Human Activity Approach to User Interfaces.” In: Human-Computer Interaction 4.3, pp. 171–195 (cit. on p. 16). Campbell, John Edward (2011). “It Takes an iVillage: Gender, Labor, and Community in the Age of Television-Internet Convergence”. In: International Journal of Communication; Vol 5 (2011), pp. 492–510 (cit. on p. 36). Campbell-Kelly, Martin (2003). From Airline Reservations to Sonic the Hedgehog: A History of the Software Industry. History of computing. Cambridge, Mass: MIT Press (cit. on pp. 28, 69, 92, 99). Campbell-Kelly, Martin (2009). “Historical reflections: The rise, fall, and resurrection of software as a service”. In: Communications of the ACM 52.5, pp. 28–30 (cit. on p. 28). Campbell-Kelly, Martin and Daniel D. Garcia-Swartz (2008). “The Rise, Fall, and Resurrection of Software as a Service: Historical Perspectives on the Computer Utility and Software for Lease on a Network”. In: The Internet and American Business. Ed. by William Aspray and Paul E Ceruzzi. MIT Press (cit. on p. 28). 144

References

Carroll, John M (1995). Scenario-Based Design: Envisioning Work and Technology in System Development. New York: Wiley (cit. on pp. 4, 38). Ceruzzi, Paul (1999). “Inventing Personal Computing”. In: The Social Shaping of Technology. Ed. by Donald A MacKenzie and Judy Wajcman. 2nd ed. Buckingham [Eng.]; Philadelphia: Open University Press, pp. 64–86 (cit. on pp. 34, 69). Chan, Adrian (2007). Social interaction design. http://www.gravity7.com/social_ interaction_design.html (Retrieved 31/01/2012) (cit. on p. 36). Chilana, Parmit K., Andrew J. Ko, Jacob O. Wobbrock, Tovi Grossman and George Fitzmaurice (2011). “Post-deployment usability: a survey of current practices”. In: Proceedings of the 2011 annual conference on Human factors in computing systems. CHI ’11. New York, NY, USA: ACM, pp. 2243–2246 (cit. on p. 48). Clarke, Adele (2005). Situational Analysis: Grounded Theory After the Postmodern Turn. Thousand Oaks, Calif: Sage Publications (cit. on pp. 47, 56, 63, 131). Clement, Andrew and Peter Van den Besselaar (1993). “A retrospective look at PD projects”. In: Communications of the ACM 36.6, pp. 29–37 (cit. on p. 21). Clement, Andrew, Terry Costantino, Dan Kurtz and Mike Tissenbaum (2008). “Participatory design and web 2.0: the case of PIPWatch, the collaborative privacy toolbar”. In: Proceedings of the Tenth Anniversary Conference on Participatory Design 2008. PDC ’08. Indianapolis, IN, USA: Indiana University, pp. 51–60 (cit. on p. 22). Clemmensen, Torkil (2005). “Community knowledge in an emerging online professional community: the case of Sigchi.dk”. In: Knowledge and Process Management 12.1, pp. 43–52 (cit. on p. 16). Cockburn, Alistair (2001). Writing Effective Use Cases. The Crystal series for software development. Boston: Addison-Wesley (cit. on p. 38). Cockton, Gilbert (2006). “Designing worth is worth designing”. In: Proceedings of the 4th Nordic conference on Human-computer interaction: changing roles. NordiCHI ’06. New York, NY, USA: ACM, pp. 165–174 (cit. on p. 19). Coleman, David (1997). Groupware: Collaborative Strategies for Corporate LANs and Intranets. Upper Saddle River, NJ: Prentice Hall PTR (cit. on pp. 29, 30, 34). Constantine, Larry L. (2004). “Beyond User-Centered Design and User Experience: Designing for User Performance”. In: Cutter IT Journal (cit. on p. 18). Constantine, Larry L. (2006a). Activity Modeling: Toward a Pragmatic Integration of Activity Theory with Usage-Centered Design. Technical Paper. Laboratory for Usage-centered Software Engineering (cit. on p. 19). Constantine, Larry L. (2006b). Designing Web Applications for Use. http://www. uie.com/articles/designing_web_applications_for_use/ (Retrieved 31/01/2012) (cit. on p. 19). Constantine, Larry L. and Lucy A. D. Lockwood (1999). Software for Use: A Practical Guide to the Models and Methods of Usage-Centered Design. Addison-Wesley Professional (cit. on pp. 18, 19). Cooper, Alan (1999). The Inmates Are Running the Asylum. Indianapolis, IN: Sams (cit. on pp. 6, 11, 18, 38). Cooper, Alan (2003). About Face 2.0: The Essentials of Interaction Design. Indianapolis, IN: Wiley (cit. on pp. 18, 83). Cooper, Geoff, Christine Hine, Janet Rachel and Steve Woolgar (1995). “Ethnography and human-computer interaction”. In: The social and interactional di-

References

145

mensions of human-computer interfaces. Ed. by Peter J. Thomas. New York, NY, USA: Cambridge University Press, pp. 11–36 (cit. on pp. 16, 121). Courage, Catherine and Kathy Baxter (2005). Understanding Your Users: A Practical Guide to User Requirements Methods, Tools, and Techniques. San Francisco, CA: Morgan Kaufmann Publishers (cit. on p. 43). Crabtree, Andrew, Tom Rodden, Peter Tolmie and Graham Button (2009). “Ethnography considered harmful”. In: Proceedings of the 27th international conference on Human factors in computing systems. Boston, MA, USA: ACM, pp. 879–888 (cit. on p. 46). Denzin, Norman K (1970). The Research Act; a Theoretical Introduction to Sociological Methods. Methodological perspectives. Chicago: Aldine Pub. Co (cit. on pp. 57, 58). Dillon, Andrew, Marian Sweeney and Martin Maguire (1993). “A survey of usability engineering within the European I.T. Industry - Current practice and needs”. In: HCI’93. People and Computers VII. Ed. by J. Alty, S. Guest and D. Diaper. Cambridge University Press (cit. on p. 5). DiSalvo, Carl, Jeff Maki and Nathan Martin (2007). “Mapmover: a case study of design-oriented research into collective expression and constructed publics”. In: Proceedings of the SIGCHI conference on Human factors in computing systems. CHI ’07. New York, NY, USA: ACM, pp. 1249–1252 (cit. on p. 22). Dittrich, Yvonne, Christiane Floyd and Ralf Klischewski (2002). Social Thinking— Software Practice. Cambridge, Mass: MIT Press (cit. on pp. 16, 46). Dix, Alan, Janet Finlay, Gregory D. Abowd and Russell Deale (1993). Human– Computer Interaction. Prentice Hall (cit. on p. 15). Dourish, Paul (2006). “Implications for design”. In: Proceedings of the SIGCHI conference on Human Factors in computing systems. CHI ’06. New York, NY, USA: ACM, pp. 541–550 (cit. on pp. 16, 46, 119, 121). Earthy, Jonathan, Brian Sherwood Jones and Nigel Bevan (2001). “The improvement of human-centred processes–facing the challenge and reaping the benefit of ISO 13407”. In: International Journal of Human-Computer Studies 55.4, pp. 553– 585 (cit. on p. 17). Eglash, Ron (2004). Appropriating Technology: Vernacular Science and Social Power. Minneapolis: University of Minnesota Press (cit. on pp. 13, 14). Ehn, Pelle (2008). “Participation in design things”. In: Proceedings of the Tenth Anniversary Conference on Participatory Design 2008. PDC ’08. Indianapolis, IN, USA: Indiana University, pp. 92–101 (cit. on p. 22). Elliott, Margaret S. and Kenneth L. Kraemer (2008). Computerization Movements and Technology Diffusion: From Mainframes to Ubiquitous Computing. Information Today, Inc. (cit. on p. 24). Elm, Malin Sveningsson (2009). “Question Three: How Do Various Notions of Privacy Influence Decisions in Qualitative Internet Research”. In: Internet inquiry: conversations about method. Ed. by Annette N. Markham and Nancy K. Baym. Los Angeles: Sage Publications, pp. 69–87 (cit. on p. 60). Elverdam, Christian and Espen Aarseth (2007). “Game Classification and Game Design: Construction Through Critical Analysis”. In: Games And Culture 2.2007, pp. 3–22 (cit. on p. 62).

146

References

Engestrom, Yrjo and David Middleton, eds. (1996). Cognition and Communication at Work. Cambridge; New York, NY, USA: Cambridge University Press (cit. on p. 16). Erickson, Thomas (2011). Social Computing. Ed. by Mads Soegaard and Rikke Friis Dam. Encyclopedia of Human-Computer Interaction. http://www.interactiondesign.org/encyclopedia/social_computing.html (Retrieved 31/01/2012) (cit. on p. 29). Eriksén, Sara (2002). “Localizing Self on the Internet: Designing for "Genius Loci" in a Global Context”. In: Social Thinking—Software Practice. Cambridge, MA, USA: MIT Press, pp. 425–449 (cit. on p. 43). Estes, Janelle, Amy Schade and Jakob Nielsen (2009). Social Media User Experience: Improving Notifications, Messages, and Alerts Sent Through Social Networks and RSS. Tech. rep. Nielsen Norman Group (cit. on p. 36). Fallows, James (1999). “The American Army and the M-16 rifle”. In: The Social Shaping of Technology. Ed. by Donald A MacKenzie and Judy Wajcman. 2nd ed. Buckingham [Eng.]; Philadelphia: Open University Press, pp. 382–394 (cit. on p. 46). Federated Media Publishing, O’Reilly Media Inc. and UBM Techweb (2011). Web 2.0 Summit. http://www.web2summit.com (Retrieved 31/01/2012) (cit. on p. 25). Fischer, Claude S (1992). America Calling: A Social History of the Telephone to 1940. Berkeley: University of California Press (cit. on p. 77). Fitzpatrick, Geraldine (2002). “The Locales Framework: Making Social Thinking Accessible for Software Practitioners”. In: Social Thinking—Software Practice. Cambridge, MA, USA: MIT Press, pp. 141–160 (cit. on p. 89). Fleischmann, Kenneth R. (2006). “Do-it-yourself information technology: Role hybridization and the design–use interface.” In: Journal of the American Society for Information Science & Technology 57.1, pp. 87–95 (cit. on pp. 34, 35). Flick, Uwe (2007). Managing quality in qualitative research. The sage qualitative research kit. Thousand Oaks, CA: Sage Publications (cit. on p. 58). Flick, Uwe (2009). “Quality Criteria in Qualitative Research”. In: An introduction to qualitative research. 4th ed. Los Angeles: Sage Publications. Chap. 28, pp. 384– 399 (cit. on p. 58). Floyd, Christine, Wolf-Michael Mehl, Fanny-Michaela Resin, Gerhard Schmidt and Gregor Wolf (1989). “Out of Scandinavia: Alternative Approaches to Software Design and System Development.” In: Human-Computer Interaction 4.4, pp. 253– 350 (cit. on p. 20). Forlizzi, Jodi and Katja Battarbee (2004). “Understanding experience in interactive systems”. In: Proceedings of the 5th conference on Designing interactive systems: processes, practices, methods, and techniques. Cambridge, MA, USA: ACM, pp. 261–268 (cit. on p. 18). Franke, Nikolaus, Eric von Hippel and Martin Schreier (2006). “Finding Commercially Attractive User Innovations: A Test of Lead-User Theory”. In: Journal of Product Innovation Management 23.4, pp. 301–315 (cit. on p. 23). Freeman, Stephanie (2011). “Constructing a Community: Myths and Realities of the Open Development Model”. PhD Thesis. University of Helsinki (cit. on pp. 35, 36).

References

147

Friedman, Andrew L. and Dominic S. Cornford (1989). Computer Systems Development: History Organization and Implementation. John Wiley & Sons (cit. on p. 69). Friedman, Batya (1996). “Value-sensitive design”. In: Interactions 3.6, pp. 16–23 (cit. on p. 19). Friedman, Batya (1997). Human Values and the Design of Computer Technology. CSLI lecture notes no. 72. Stanford, Calif.; Cambridge; New York: CSLI Publications; Cambridge University Press (cit. on pp. 19, 46). Fulton Suri, Jane (2003). “Empathic Design: Informed and Inspired by Other People’s Experience”. In: Empathic Design: User Experience in Product Design. Ed. by Ilpo Koskinen, Tuuli Mattelmäki and Katja Battarbee. IT Press, pp. 51–57 (cit. on p. 38). Følstad, Asbjørn, Håvard D. Jørgensen and John Krogstie (2004). “User involvement in e-government development projects”. In: Proceedings of the third Nordic conference on Human-computer interaction. NordiCHI ’04. New York, NY, USA: ACM, pp. 217–224 (cit. on p. 5). Gaffney, Gerry (2006). Personas and outrageous software - an interview with Alan Cooper. http://www.infodesign.com.au/uxpod/alancooper (Retrieved 31/01/2012) (cit. on p. 4). Garrett, Jesse James (2011). The Elements of User Experience: User-Centered Design for the Web and Beyond. Berkeley, CA: New Riders (cit. on p. 36). Gaver, William W. (1991). “Technology affordances”. In: Proceedings of the SIGCHI conference on Human factors in computing systems: Reaching through technology. CHI ’91. New York, NY, USA: ACM, pp. 79–84 (cit. on p. 16). Geels, Frank W. and Johan Schot (2007). “Typology of sociotechnical transition pathways”. In: Research Policy 36.3, pp. 399–417 (cit. on pp. 27, 28, 48). Gillmor, Dan (2004). We the Media: Grassroots Journalism by the People, for the People. Beijing; Sebastopol, CA: O’Reilly (cit. on pp. 7, 24). Gilmore, David J., Gilbert Cockton, Elizabeth Churchill, Sari Kujala, Austin Henderson and Monty Hammontree (2008). “Values, value and worth: their relationship to HCI?” In: CHI ’08 extended abstracts on Human factors in computing systems. CHI EA ’08. New York, NY, USA: ACM, pp. 3933–3936 (cit. on pp. 19, 46). Ginsburg, Suzanne (2011). Designing the iPhone User Experience: A User-Centered Approach to Sketching and Prototyping iPhone Apps. Upper Saddle River, NJ: Addison-Wesley (cit. on p. 36). Glazer, Hillel, Jeff Dalton, David Anderson, Michael D. Konrad and Sandra Shrum (2008). CMMI or Agile: Why Not Embrace Both! Technical Note CMU/SEI-2008TN-003. Software Engineering Institute / Carnegie Mellon University (cit. on pp. 17, 126). Goodlad, Lauren M. E. and Michael Bibby (2007). Goth: Undead Subculture. Duke University Press (cit. on p. 108). Goodman-Deane, Joy, P John Clarkson, Patrick Langdon and Susannah Clarke (2008). “Categorising Design Methods”. In: Contemporary Ergonomics 2008. Ed. by Philip Bust. Taylor & Francis, pp. 273–278 (cit. on pp. 13, 43). Gould, John D. and Clayton Lewis (1985). “Designing for usability: key principles and what designers think”. In: Communications of the ACM 28.3, pp. 300–311 (cit. on p. 15).

148

References

Granovetter, Mark S. (1973). “The Strength of Weak Ties”. In: American Journal of Sociology 78.6, pp. 1360–1380 (cit. on p. 76). Gray, Wayne D. and Marilyn C. Salzman (1998). “Damaged Merchandise? A Review of Experiments That Compare Usability Evaluation Methods.” In: HumanComputer Interaction 13.3, pp. 203–261 (cit. on p. 15). Greenbaum, Joan and Morten Kyng, eds. (1991). Design at Work: Cooperative Design of Computer Systems. Hillsdale, NJ, USA: Lawrence Erlbaum Associates (cit. on p. 20). Greenberg, Saul and Bill Buxton (2008). “Usability evaluation considered harmful (some of the time)”. In: Proceeding of the twenty-sixth annual SIGCHI conference on Human factors in computing systems. Florence, Italy: ACM, pp. 111–120 (cit. on pp. 5, 15, 46, 91, 126). Grudin, J. (2005). “Three faces of human-computer interaction”. In: Annals of the History of Computing, IEEE 27.4, pp. 46–62 (cit. on p. 15). Grudin, Jonathan (1991). “Interactive Systems: Bridging the Gaps Between Developers and Users”. In: Computer 24.4, pp. 59–69 (cit. on pp. 76, 87, 120, 122). Grudin, Jonathan (1993). “Interface: an evolving concept”. In: Communications of the ACM 36.4, pp. 110–119 (cit. on p. 1). Gulliksen, Jan, Inger Boivie and Bengt Göransson (2006). “Usability professionalscurrent practices and future development”. In: Interacting with Computers 18, pp. 568–600 (cit. on p. 5). Gulliksen, Jan, Inger Boivie, Jenny Persson, Anders Hektor and Lena Herulf (2004). “Making a difference: a survey of the usability profession in Sweden”. In: Proceedings of the third Nordic conference on Human-computer interaction. NordiCHI ’04. Tampere, Finland: ACM, pp. 207–215 (cit. on p. 5). Gulliksen, Jan, Bengt Göransson, Inger Boivie, Stefan Blomkvist, Jenny Persson and Åsa Cajander (2003). “Key principles for user-centred systems design”. In: Behaviour and Information Technology 22, pp. 397–409 (cit. on p. 6). Gulliksen, Jan, Ann Lantz and Inger Boivie (1999). “User centered design—problems and possibilities: a summary of the 1998 PDC & CSCW workshop”. In: ACM SIGCHI Bulletin 31, pp. 25–35 (cit. on pp. 20, 21). Gutwin, Carl and Saul Greenberg (2002). “A Descriptive Framework of Workspace Awareness for Real-Time Groupware”. In: Computer Supported Cooperative Work (CSCW) 11.3, pp. 411–446 (cit. on p. 30). Gärtner, Johannes (1998). “Participatory Design in Consulting”. In: Computer Supported Cooperative Work (CSCW) 7.3-4, pp. 273–289 (cit. on p. 20). Gärtner, Johannes and Ina Wagner (1996). “Mapping Actors and Agendas: Political Frameworks of Systems Design and Participation”. In: Human-Computer Interaction 11.3, pp. 187–214 (cit. on p. 21). Habborator.org (2011). Habbo Hotel Fact Sheets. http : / / www. habborator. org (Retrieved 31/01/2012) (cit. on p. 73). Hackett, Edward J, Olga Amsterdamska, Michael Lynch and Judy Wajcman, eds. (2008). The Handbook of Science and Technology Studies. 3rd ed. Cambridge, Mass: MIT Press; Published in cooperation with the Society for the Social Studies of Science (cit. on p. 46). Hackos, JoAnn T. and Janice C. Redish (1998). User and Task Analysis for Interface Design. Wiley (cit. on pp. 11, 43).

References

149

Hagen, Penny and Toni Robertson (2010). “Social technologies: challenges and opportunities for participation”. In: Proceedings of the 11th Biennial Participatory Design Conference. PDC ’10. New York, NY, USA: ACM, pp. 31–40 (cit. on pp. 22, 29, 36). Haigh, Thomas (2008). “Protocols for Profit: Web and Email Technologies as Product and Platform”. In: The Internet and American Business. Ed. by William Aspray and Paul E Ceruzzi. MIT Press, pp. 105–158 (cit. on pp. 28, 33). Hales, Mike (1994). “Where are designers? Styles of design practice, objects of design and views of users in CSCW”. In: Design issues in CSCW. Ed. by D Rosenberg and C Hutchinson. London: Springer-Verlag, pp. 151–178 (cit. on p. 44). Harper, R. H. R. (2000). “The Organisation in Ethnography – A Discussion of Ethnographic Fieldwork Programs in CSCW”. In: Computer Supported Cooperative Work (CSCW) 9.2, pp. 239–264 (cit. on p. 16). Hassenzahl, Marc and Noam Tractinsky (2006). “User experience – a research agenda”. In: Behaviour & Information Technology 25.2, pp. 91–97 (cit. on p. 18). Hebdige, Dick (1979). Subculture: The Meaning of Style. London: Routledge (cit. on p. 108). Heiskanen, Eva, Sampsa Hyysalo, Tanja Kotro and Petteri Repo (2010). “Constructing innovative users and user-inclusive innovation communities”. In: Technology Analysis & Strategic Management 22.4, pp. 495–511 (cit. on p. 44). Henderson, Austin and Morten Kyng (1991). “There’s no place like home: continuing design in use”. In: Design at Work: Cooperative Design of Computer Systems. Ed. by Joan Greenbaum and Morten Kyng. Hillsdale, NJ, USA: Lawrence Erlbaum Associates, pp. 219–240 (cit. on p. 22). Herring, Susan C (2007). “A Faceted Classification Scheme for Computer-Mediated Discourse”. In: Language@Internet 4.1 (cit. on p. 30). Herring, Susan C. (2010). “Web content analysis: Expanding the paradigm”. In: The International Handbook of Internet Research. Ed. by J. Hunsinger, M. Allen and L. Klastrup. Berlin: Springer Verlag, pp. 233–249 (cit. on pp. 106, 115, 123). Hertzum, Morten, Torkil Clemmensen, Kasper Hornbæk, Jyoti Kumar, Qingxin Shi and Pradeep Yammiyavar (2011). “Personal Usability Constructs: How People Construe Usability Across Nationalities and Stakeholder Groups”. In: International Journal of Human-Computer Interaction 27.8, pp. 729–761 (cit. on p. 43). Hienerth, Christoph (June 2006). “The commercialization of user innovations: the development of the rodeo kayak industry”. In: R&D Management 36.3, pp. 273– 294 (cit. on p. 23). Himanen, Pekka (2001). The Hacker Ethic, and the Spirit of the Information Age. New York: Random House (cit. on p. 34). Hine, Christine (2009). “Question One: How Can Qualitative Internet Researchers Define the Boundaries of Their Projects?” In: Internet inquiry: conversations about method. Ed. by Annette N. Markham and Nancy K. Baym. Los Angeles: Sage Publications, pp. 1–20 (cit. on p. 59). Hippel, Eric von (1988). The Sources of Innovation. New York: Oxford University Press (cit. on pp. 6, 23, 44). Hippel, Eric von (2005). Democratizing Innovation. The MIT Press (cit. on pp. 6, 23, 37, 44).

150

References

Hippel, Eric von, Stefan Thomke and Mary Sonnack (1999). “Creating Breakthroughs at 3M.” In: Harvard Business Review 77.5, pp. 47–57 (cit. on pp. 6, 23). Hippel, Eric von (2001). “User toolkits for innovation”. In: Journal of Product Innovation Management 18.4, pp. 247–257 (cit. on pp. 23, 129). Hippel, Eric von and Ralph Katz (2002). “Shifting Innovation to Users via Toolkits”. In: Management Science 48.7, pp. 821–833 (cit. on p. 24). Holtzblatt, Karen, Jessamyn Burns Wendell and Shelley Wood (2005). Rapid Contextual Design: A How-to Guide to Key Techniques for User-Centered Design. San Francisco: Elsevier/Morgan Kaufmann (cit. on p. 16). Holzapfel, Nicolas (2008). “User-centred design and the user-driven web”. M.A. Interactive Media. University of the Arts London (cit. on pp. 2, 36, 37, 128). Hoogma, Remco, René Kemp, Johan Schot and Bernhard Truffer (2002). Experimenting for Sustainable Transport: The Approach of Strategic Niche Management. Transport, development and sustainability. London: Spon (cit. on p. 46). Hornbæk, Kasper (2006). “Current practice in measuring usability: Challenges to usability studies and research”. In: International Journal of Human–Computer Studies 64.2, pp. 79–102 (cit. on p. 15). Howells, Jeremy (2006). “Intermediation and the role of intermediaries in innovation”. In: Research Policy 35.5, pp. 715–728 (cit. on pp. 69, 129). Hughes, Thomas Parke (1987). “The Evolution of Large Technological Systems”. In: The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology. Ed. by Wiebe E Bijker, Thomas Parke Hughes and Trevor J Pinch. Cambridge, Mass: MIT Press, pp. 51–82 (cit. on pp. 71, 73–75). Humphreys, Lee (2005). “Reframing Social Groups, Closure, and Stabilization in the Social Construction of Technology”. In: Social Epistemology 19, pp. 231–253 (cit. on p. 47). Hussain, Zahid, Wolfgang Slany and Andreas Holzinger (2009). “Current State of Agile User-Centered Design: A Survey”. In: HCI and Usability for e-Inclusion. Lecture Notes in Computer Science. Springer, pp. 416–427 (cit. on pp. 5, 17). Hutchins, Edwin (1995). Cognition in the Wild. Cambridge, Mass: MIT Press (cit. on p. 16). Hyysalo, Sampsa (2004). Uses of innovation: wristcare in the practices of engineers and elderly. University of Helsinki, Department of Education (cit. on pp. 4, 8, 45). Hyysalo, Sampsa (2006). “Representations of Use and Practice-Bound Imaginaries in Automating the Safety of the Elderly”. In: Social Studies of Science 36.4, pp. 599–626 (cit. on p. 47). Hyysalo, Sampsa (2009). “Learning for learning economy and social learning”. In: Research Policy 38.5, pp. 726–735 (cit. on p. 121). Hyysalo, Sampsa (2010). Health Technology Development and Use. New York: Routledge (cit. on pp. 8, 24, 43, 45, 48, 129). Iivari, Juhani, Rudy Hirschheim and Heinz K. Klein (2001). “A Dynamic Framework for Classifying Information Systems Development Methodologies and Approaches”. In: Management Information Systems 17.3, pp. 179–218 (cit. on p. 3). Isbister, Katherine and Kristina Höök (2009). “On being supple: in search of rigor without rigidity in meeting new design and evaluation challenges for HCI practitioners”. In: Proceedings of the 27th international conference on Human factors in

References

151

computing systems. CHI ’09. New York, NY, USA: ACM, pp. 2233–2242 (cit. on p. 22). ISO (1998). Ergonomic requirements for office work with visual display terminals (VDTs) – Part 11: Guidance on usability. Tech. rep. ISO 9241-11:1998. International Organization for Standardization (cit. on pp. 43, 131). ISO (1999). Human-centred design processes for interactive systems. Tech. rep. ISO 13407:1999. International Organization for Standardization (cit. on pp. 6, 17, 131). ISO (2000). Ergonomics of human-system interaction – Human-centred lifecycle process descriptions. Tech. rep. ISO/TR 18529:2000. International Organization for Standardization (cit. on pp. 11, 17). ISO (2002). Ergonomics of human-system interaction – Usability methods supporting human-centred design. Tech. rep. ISO/TR 16982:2002. International Organization for Standardization (cit. on pp. 2, 5, 11, 90, 125). ISO (2010). Human-centred design for interactive systems. Tech. rep. International Organization for Standardization (cit. on pp. 18, 124). IXDA (2008). ACD/UCD Round Two. http://www.ixda.org/node/18087 (Retrieved 31/01/2012) (cit. on p. 19). Jacob, Elin (2004). “Classification and categorization: a difference that makes a difference”. In: Library Trends 52.3, pp. 515–540 (cit. on p. 4). Jacobsen, Niels Ebbe, Morten Hertzum and Bonnie E. John (1998). “The evaluator effect in usability tests”. In: CHI 98 conference summary on Human factors in computing systems. CHI ’98. New York, NY, USA: ACM Press, pp. 255–256 (cit. on p. 15). Jenkins, Henry (2006). Convergence Culture: Where Old and New Media Collide. New York: New York University Press (cit. on pp. 7, 24, 25). Jeppesen, Lars Bo and Lars Frederiksen (2006). “Why Do Users Contribute to Firm-Hosted User Communities? The Case of Computer-Controlled Music Instruments.” In: Organization Science 17.1, pp. 45–63 (cit. on p. 23). Jeppesen, Lars B.O. and Måns J. Molin (2003). “Consumers as Co-developers: Learning and Innovation Outside the Firm.” In: Technology Analysis & Strategic Management 15.3, pp. 363–383 (cit. on pp. 23, 129). Johansen, Robert (1988). Groupware: Computer Support for Business Teams. New York: London: Free Press; Collier Macmillan (cit. on pp. 29, 30, 34). John, Bonnie E (2003). “Information Processing and Skilled Behavior”. In: HCI Models, Theories, and Frameworks: Toward a Multidisciplinary Science. Ed. by John Carroll. Morgan Kaufmann, pp. 11–26 (cit. on p. 44). Johnson, Mikael (2006a). “Categorising users in the design of Habbo Hotel”. In: ProACT 2006 – International Innovation Pressure Conference. Tampere, Finland (cit. on p. 59). Johnson, Mikael (2006b). “Uses of history in the design of Habbo Hotel”. In: ProACT 2006 – International Innovation Pressure Conference. Tampere, Finland (cit. on p. 59). Johnson, Mikael (2006c). “Visiotypes in the membership categorisation practices of Habbo Hotel”. In: Public Imageries Today Conference. The Finnish Museum for Photography and the PhD programme "Image. Body. Medium. An Anthropological Perspective." at the New Media School in Karlsruhe. Helsinki, Finland (cit. on p. 59). 152

References

Johnson, Mikael (2007). “Habbo-artefaktien kotouttaminen”. In: Teknologian kulttuurinen ja yhteiskunnallinen tutkimus. Ed. by Petteri Repo and Jari Aro. Työselosteita ja esitelmiä 106. Kuluttajatutkimuskeskus. Helsinki, Finland (cit. on p. 59). Johnson, Mikael (2009). “How The Teenage Invasion Changed Habbo or AgeSensitive Design In Inclusive Society”. In: The Good, The Bad, The Challenging: The user and the future of information and communication technologies: a transdisciplinary conference. Organized by COST Action 298, Participation in the Broadband Society. Ed. by Bartolomeo Sapio, Leslie Haddon, Enid Mante-Meijer, Leopoldina Fortunati, Tomaž Turk and Eugene Loos. Copenhagen, Denmark (cit. on p. 59). Johnson, Mikael (2010). “Localisation and Generification Processes in Social Media: Case Habbo Hotel”. In: Workshop: The role of users in the intertwined changes of technology and practice. Helsinki Collegium for Advanced Studies. Helsinki, Finland (cit. on p. 59). Johnson, Mikael and Sakari Tamminen (2007). “Doing Gender in Habbo Hotel”. In: 4th European Gender & ICT Symposium. University of Helsinki. Helsinki, Finland (cit. on p. 59). Johnson-Lenz, Peter and Trudy Johnson-Lenz (1991). “Post-mechanistic groupware primitives: rhythms, boundaries and containers”. In: International Journal of Man-Machine Studies 34.3, pp. 395–417 (cit. on p. 29). Jokela, Timo (2004). “Evaluating the user-centredness of development organisations: conclusions and implications from empirical usability capability maturity assessments”. In: Interacting with Computers 16.6, pp. 1095–1132 (cit. on p. 5). Jordan, Patrick W (2000). Designing pleasurable products: an introduction to the new human factors. London; New York: Taylor & Francis (cit. on p. 18). Kaplan, Andreas M. and Michael Haenlein (2010). “Users of the world, unite! The challenges and opportunities of Social Media”. In: Business Horizons 53.1, pp. 59–68 (cit. on p. 25). Kaptelinin, Victor (1992). “Can mental models be considered harmful?” In: Posters and short talks of the 1992 SIGCHI conference on Human factors in computing systems. CHI ’92. New York, NY, USA: ACM, pp. 107–108 (cit. on p. 46). Kaptelinin, Victor, Bonnie A. Nardi and Catriona Macaulay (1999). “Methods & tools: The activity checklist: a tool for representing the ’space’ of context”. In: interactions 6.4, pp. 27–39 (cit. on p. 16). Keinonen, Turkka (2010). “Protect and Appreciate – Notes on the Justification of User-Centered Design”. In: International Journal of Design 4.1, pp. 17–27 (cit. on p. 18). Kensing, Finn and Jeanette Blomberg (1998). “Participatory Design: Issues and Concerns”. In: Computer Supported Cooperative Work (CSCW) 7.3-4, pp. 167–185 (cit. on p. 20). Kim, Amy Jo (2000). Community Building on the Web: Secret Strategies for Successful Online Communities. Berkeley, CA, USA: Peachpit Press (cit. on p. 36). Klein, Hans K. and Daniel Lee Kleinman (2002). “The Social Construction of Technology: Structural Considerations”. In: Science, Technology, & Human Values 27.1, pp. 28–52 (cit. on p. 47). Klepper, Steven (1997). “Industry Life Cycles”. In: Industrial and Corporate Change 6.1, pp. 145–182 (cit. on pp. 27, 48).

References

153

Koskinen, Aleksi (2006). “Elämistä, olemista, pelaamista - keitä Habbo Hotelin käyttäjät ovat ja kuinka heidän kanssaan tehdään nuorisotyötä?” B.Sc. Thesis. Mikkeli Polytechnic (cit. on p. 52). Kotler, Philip (1997). Marketing Management: Analysis, Planning, Implementation, and Control. 9th ed. Upper Saddle River, NJ: Prentice Hall (cit. on pp. 4, 37). Krug, Steve (2006). Don’t Make Me Think!: A Common Sense Approach to Web Usability. Berkeley, Calif: New Riders Pub (cit. on p. 36). Kuniavsky, Mike (2003). Observing the User Experience: A Practitioner’s Guide to User Research. Morgan Kaufmann (cit. on p. 18). Kuutti, Kari (1996). “Activity theory as a potential framework for human-computer interaction research”. In: Context and Consciousness: Activity Theory and HumanComputer Interaction. Ed. by Bonnie A Nardi. Cambridge, Mass: MIT Press, pp. 17–44 (cit. on p. 16). Kvale, Steinar (1995). “The Social Construction of Validity”. en. In: Qualitative Inquiry 1.1, pp. 19–40 (cit. on p. 58). Kvale, Steinar (2007). Doing interviews. The sage qualitative research kit. Thousand Oaks, CA: Sage Publications (cit. on p. 58). KZero (2009). The Virtual Worlds Radar. http : / / www. kzero . co . uk / radar. php (Retrieved 31/01/2012) (cit. on p. 71). Lamb, Roberta and Rob Kling (2003). “Reconceptualizing users as social actors in information systems research”. In: MIS Quarterly 27.2, pp. 197–235 (cit. on p. 44). Larusdottir, M. K., E. Bjarnadottir and J. Gulliksen (2010). “The Focus on Usability in Testing Practices in Industry”. In: Human Computer Interaction Symposium, World Computer Congress 2010, pp. 98–109 (cit. on p. 17). Latour, Bruno (1987). Science in Action: How to Follow Scientists and Engineers Through Society. Cambridge, Mass: Harvard University Press (cit. on p. 67). Latour, Bruno (1992). “Where Are the Missing Masses? The Sociology of a Few Mundane Artifacts”. In: Shaping Technology/Building Society: Studies in Sociotechnical Change. Ed. by Wiebe E Bijker and John Law. Cambridge, Mass: MIT Press, pp. 225–258 (cit. on p. 47). Latour, Bruno (2005). Reassembling the Social: An Introduction to Actor-NetworkTheory. Clarendon lectures in management studies. Oxford; New York: Oxford University Press (cit. on p. 76). Latour, Bruno and Steve Woolgar (1979). Laboratory Life: The Social Construction of Scientific Facts. Beverly Hills: Sage Publications (cit. on p. 47). Law, Effie Lai-Chong, Virpi Roto, Marc Hassenzahl, Arnold P.O.S. Vermeeren and Joke Kort (2009). “Understanding, scoping and defining user experience: a survey approach”. In: Proceedings of the 27th international conference on Human factors in computing systems. Boston, MA, USA: ACM, pp. 719–728 (cit. on p. 18). Lee, Yee (2011). FrameThink: How Facebook Ships Code. http://framethink.blogspot. com/2011/01/how- facebook- ships- code.html (Retrieved 31/01/2012) (cit. on p. 41). Lehdonvirta, Vili (June 2009). “Virtual item sales as a revenue model: identifying attributes that drive purchase decisions”. In: Electronic Commerce Research 9.1, pp. 97–113 (cit. on p. 28).

154

References

Lehmuskallio, Asko, Sakari Tamminen and Mikael Johnson (2008). “Managing Privacy on Social Network Sites”. In: Medienamateure: Wie verändern Laien unsere visuelle Kultur? Siegen, Germany (cit. on p. 60). Leonard-Barton, Dorothy (1995). Wellsprings of Knowledge. Harvard Business School Press (cit. on pp. 13, 14, 28). Levy, Steven (1984). Hackers: Heroes of the Computer Revolution. Garden City, N.Y: Anchor Press/Doubleday (cit. on p. 34). Lindsay, Christina (2003). “From the Shadows: Users as Designers, Producers, Marketers, Distributors, and Technical Support”. In: How Users Matter: The CoConstruction of Users and Technology. Ed. by Nelly Oudshoorn and Trevor Pinch. Cambri: MIT Press. Chap. 1, pp. 29–50 (cit. on pp. 34, 35). Livingstone, Sonia (2008). “Taking risky opportunities in youthful content creation: teenagers’ use of social networking sites for intimacy, privacy and selfexpression”. In: New Media & Society 10.3, pp. 393–411 (cit. on p. 59). Lüthje, Christian (2003). “Customers as co-inventors - An empirical analysis of the antecedents of customer-driven innovations in the field of medical equipment”. In: 32nd EMAC Conference, Glasgow, UK (cit. on p. 23). Lüthje, Christian, Cornelius Herstatt and Eric von Hippel (2005). “User-innovators and "local" information: The case of mountain biking”. In: Research Policy 34.6, pp. 951–965 (cit. on pp. 6, 23). Mackay, Hugh, Chris Carne, Paul Beynon-Davies and Doug Tudhope (2000). “Reconfiguring the User: Using Rapid Application Development”. In: Social Studies Of Science 30.5, pp. 737–757 (cit. on pp. 35, 44, 45). MacKenzie, Donald (1999). “Theories of technology and the abolition of nuclear weapons”. In: The Social Shaping of Technology. Ed. by Donald A MacKenzie and Judy Wajcman. 2nd ed. Buckingham [Eng.]; Philadelphia: Open University Press, pp. 419–442 (cit. on p. 46). MacKenzie, I. Scott (2003). “Motor Behavior Models for Human-Computer Interaction”. In: HCI Models, Theories, and Frameworks: Toward a Multidisciplinary Science. Ed. by John Carroll. Morgan Kaufmann, pp. 11–26 (cit. on p. 44). Maguire, Martin (2001a). “Context of Use within usability activities”. In: International Journal of Human-Computer Studies 55.4, pp. 453–483 (cit. on pp. 18, 131). Maguire, Martin (2001b). “Methods to support human-centred design”. In: International Journal of Human-Computer Studies 55.4, pp. 587–634 (cit. on pp. 5, 11, 12, 17). Malaby, Thomas M (2007). “Beyond Play: A New Approach to Games”. In: Games And Culture 2.2, pp. 95–113 (cit. on p. 62). Mander, Jonathan (2001). To The Point - uusmedian kasvattajaseura. http://www. rekaksois . com / research / 305 / to - the - point - uusmedian - kasvattajaseura (Retrieved 31/01/2012) (cit. on p. 64). Mansfield, Tim, Simon Kaplan, Geraldine Fitzpatrick, Ted Phelps, Mark Fitzpatrick and Richard Taylor (1997). “Evolving Orbit: a process report on building locales”. In: Proceedings of the international ACM SIGGROUP conference on Supporting group work: the integration challenge. GROUP ’97. New York, NY, USA: ACM, pp. 241–250 (cit. on p. 89). Markham, Annette N. and Nancy K. Baym, eds. (2009). Internet inquiry: conversations about method. Los Angeles: Sage Publications (cit. on p. 59).

References

155

Mayhew, Deborah J (1999). The usability engineering lifecycle: a practitioner’s handbook for user interface design. San Francisco, Calif.: Morgan Kaufmann Publishers (cit. on p. 11). Merikivi, Jani (2007). Netari.fi - nuorisotyön arvo. Tech. rep. Nuorisotutkimusverkosto/Nuorisotutkimusseura (cit. on p. 52). Merkel, Cecelia B., Lu Xiao, Umer Farooq, Craig H. Ganoe, Roderick Lee, John M. Carroll and Mary Beth Rosson (2004). “Participatory design in community computing contexts: tales from the field”. In: Proceedings of the eighth conference on Participatory design: Artful integration: interweaving media, materials and practices - Volume 1. PDC 04. New York, NY, USA: ACM, pp. 1–10 (cit. on p. 22). Messinger, Paul R., Eleni Stroulia and Kelly Lyons (2008). “A Typology of Virtual Worlds: Historical Overview and Future Directions”. In: Journal of Virtual Worlds Research 1.1 (cit. on p. 62). Miller, L. (2005). “Case study of customer input for a successful product”. In: Agile Conference, 2005. Proceedings. IEEE, pp. 225–234 (cit. on p. 102). Molich, Rolf, Meghan R Ede, Klaus Kaasgaard and Barbara Karyukin (Jan. 2004). “Comparative usability evaluation”. In: Behaviour & Information Technology 23, pp. 65–74 (cit. on p. 15). Molich, Rolf, Ann Damgaard Thomsen, Barbara Karyukina, Lars Schmidt, Meghan Ede, Wilma van Oel and Meeta Arcuri (1999). “Comparative evaluation of usability tests”. In: CHI ’99 Extended Abstracts on Human Factors in Computing Systems. New York, NY, USA: ACM Press, pp. 83–84 (cit. on p. 15). Muller, Michael, David R. Millen and Carol Strohecker (2001). “What makes a representative user representative? a participatory poster”. In: CHI ’01 extended abstracts on Human factors in computing systems. CHI EA ’01. New York, NY, USA: ACM, pp. 101–102 (cit. on p. 83). Muller, Michael J. (2003). “Participatory design: the third space in HCI”. In: The human-computer interaction handbook: fundamentals, evolving technologies and emerging applications. L. Erlbaum Associates Inc., pp. 1051–1068 (cit. on p. 120). Muller, Michael J and Sarah Kuhn (1993). “Participatory design”. In: Communications of the ACM 36, pp. 24–28 (cit. on pp. 11, 12). Myers, Brad, Jim Hollan, Isabel Cruz, Steve Bryson, Dick Bulterman, Tiziana Catarci, Wayne Citrin, Ephraim Glinert, Jonathan Grudin and Yannis Ioannidis (1996). “Strategic directions in human-computer interaction”. In: ACM Computing Surveys 28.4, pp. 794–809 (cit. on p. 15). Mørch, Anders I. and Nikolay D. Mehandjiev (1999). “Tailoring as Collaboration: The Mediating Role of Multiple Representations and ApplicationUnits”. In: Computer Supported Cooperative Work (CSCW) 9.1, pp. 75–100 (cit. on p. 22). Nardi, Bonnie A (1993). A Small Matter of Programming: Perspectives on End User Computing. Cambridge, MA: MIT Press (cit. on p. 22). Nardi, Bonnie A, ed. (1996). Context and Consciousness: Activity Theory and HumanComputer Interaction. Cambridge, Mass: MIT Press (cit. on p. 16). Nardi, Bonnie A (1997). “The Use of Ethnographic Methods in Design and Evaluation”. In: Handbook of HCI. Ed. by Martin Helander, Thomas K. Landauer and Prasad V. Prabhu. 2nd. Elsevier Science, pp. 361–366 (cit. on p. 16). Nardi, Bonnie A. and James R. Miller (1991). “Twinkling lights and nested loops: distributed problem solving and spreadsheet development”. In: International Journal of Man-Machine Studies 34.2, pp. 161–184 (cit. on p. 35). 156

References

Nelson, Ted (1982). “A new home for the mind”. In: Datamation 28, pp. 168–173 (cit. on p. 33). Nerokala.com (2004). Kysy Habbo-Oppaalta ja Staffilta. (Archived at http://web. archive.org). http://nerokala.com/forum/index.php?topic=6197.0 (cit. on p. 70). Nielsen, Jakob (1993). Usability Engineering. Academic Press Inc (cit. on pp. 11, 15). Nielsen, Jakob (2006a). Corporate Usability Maturity: Stages 1–4. http://www.useit. com/alertbox/maturity.html (Retrieved 31/01/2012) (cit. on p. 17). Nielsen, Jakob (2006b). Corporate Usability Maturity: Stages 5–8. http://www.useit. com/alertbox/process_maturity.html (Retrieved 31/01/2012) (cit. on p. 17). Nielsen, Jakob (2008). Bridging the Designer-User Gap. http://www.useit.com/ alertbox/designer-user-differences.html (Retrieved 31/01/2012) (cit. on pp. 76, 87). Norman, Don (2006). “Why doing user observations first is wrong”. In: Interactions 13.4, 50–ff (cit. on pp. 5, 124). Norman, Donald A (1988). The Psychology of Everyday Things. New York: Basic Books (cit. on pp. 16, 44, 76, 87). Norman, Donald A (2005). “Human-centered design considered harmful”. In: Interactions 12, pp. 14–19 (cit. on pp. 19, 46, 124). Norman, Donald A. and Stephen W. Draper (1986). User Centered System Design: New Perspectives on Human-computer Interaction. Hillsdale, NJ, USA: Lawrence Erlbaum Associates (cit. on pp. 15, 19). Noyes, Janet M and Christopher Baber (1999). User-Centered Design of Systems. London: Springer (cit. on p. 77). Nyce, James M. and Jonas Löwgren (1995). “Toward foundational analysis in human-computer interaction”. In: The social and interactional dimensions of human-computer interfaces. Ed. by Peter J. Thomas. New York, NY, USA: Cambridge University Press, pp. 37–47 (cit. on pp. 16, 121). Nørgaard, Mie and Kasper Hornbæk (2006). “What do usability evaluators do in practice?: an explorative study of think-aloud testing”. In: Proceedings of the 6th conference on Designing Interactive systems. DIS ’06. New York, NY, USA: ACM, pp. 209–218 (cit. on p. 5). OECD (2007). Participative Web and User-Created Content: Web 2.0, Wikis and Social Networking. Tech. rep. Organisation for Economic Co-operation and Development (cit. on p. 25). OECD (2009). Rethinking e-Government Services: User-centred Approaches. Tech. rep. Organisation for Economic Co-operation and Development (cit. on p. 36). O’Reilly, Tim (2005). What Is Web 2.0. http://oreilly.com/web2/archive/what-isweb-20.html (Retrieved 31/01/2012) (cit. on p. 25). Osterwalder, Alexander and Yves Pigneur (2010). Business Model Generation: A Handbook for Visionaries, Game Changers, and Challengers. Hoboken, NJ: Wiley (cit. on pp. 3, 24, 28). Oudshoorn, Nelly (1999). “The decline of the one-size-fits-all paradigm, or, how reproductive scientists try to cope with postmodernity”. In: The Social Shaping of Technology. Ed. by Donald A MacKenzie and Judy Wajcman. 2nd ed. Buckingham [Eng.] ; Philadelphia: Open University Press, pp. 325–340 (cit. on p. 46). Oudshoorn, Nelly and Trevor Pinch (2003). How Users Matter: The Co-Construction of Users and Technology. Cambridge: MIT Press (cit. on p. 43).

References

157

Oulasvirta, Antti and Pertti Saariluoma (2006). “Surviving task interruptions: Investigating the implications of long-term working memory theory”. In: International Journal of Human-Computer Studies 64.10, pp. 941–961 (cit. on p. 44). Park, Robert Ezra (1924). “The Concept of Social Distance As Applied to the Study of Racial Attitudes and Racial Relations”. In: Journal of Applied Sociology 8, pp. 339–344 (cit. on p. 76). Patton, Jeff (2008). Twelve emerging best practices for adding UX work to Agile development. http://www.agileproductdesign.com/blog/emerging_best_agile_ ux_practice.html (Retrieved 31/01/2012) (cit. on p. 102). Payne, Stephen J. (2003). “Users’ mental models: the very ideas”. In: HCI Models, Theories, and Frameworks: Toward a Multidisciplinary Science. Ed. by John Carroll. Morgan Kaufmann, pp. 11–26 (cit. on p. 44). Pelkonen, Tommi (2005). History on the Finnish Digital Media. http://pelkonen. wordpress.com/2005/07/14/history-on-the-finnish-digital-media/ (Retrieved 31/01/2012) (cit. on p. 64). Perez, Carlota (2010). “Technological revolutions and techno-economic paradigms”. In: Cambridge Journal of Economics 34.1, pp. 185–202 (cit. on pp. 27, 48). Philip, Ross and Chris Rourke (2006). Beyond usability testing: user-centred design and organisational maturity. http : / / www. mercurytide . co . uk / news / article / beyond-usability-testing/ (Retrieved 31/01/2012) (cit. on p. 17). Pietiläinen, Reetta (2004). “Kommunikointi ja toiminta Habbo Hotel Kultakalassa”. Pro gradu. University of Helsinki (cit. on p. 52). Pinch, Trevor J. and Wiebe E. Bijker (1984). “The Social Construction of Facts and Artefacts: Or How the Sociology of Science and the Sociology of Technology Might Benefit Each Other”. In: Social Studies of Science 14.3, pp. 399–441 (cit. on pp. 64, 65). Pinch, Trevor J and Wiebe E Bijker (1987). “The Social Construction of Facts and Artifacts: Or How the Sociology of Science and the Sociology of Technology Might Benefit Each Other”. In: The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology. Ed. by Wiebe E Bijker, Thomas Parke Hughes and Trevor J Pinch. Cambridge, Mass: MIT Press, pp. 17– 50 (cit. on pp. 47, 64). Piskorski, Mikolaj Jan, Thomas R. Eisenmann, David Chen and Brian Feinstein (2011). Facebook. Harvard Business School Case 808-128. Harvard Business School (cit. on p. 25). Pollock, Neil and Robin Williams (2009). Software and Organisations: The Biography of the Enterprise-Wide System or How SAP Conquered the World. London; New York: Routledge (cit. on pp. 8, 43, 48, 129). Pollock, Neil, Robin Williams and Luciana D’Adderio (2007). “Global Software and its Provenance”. In: Social Studies of Science 37.2, pp. 254–280 (cit. on pp. 38, 128). Porter, Joshua (n.d.). Personas and the Advantage of Designing for Yourself. http: / / bokardo . com / archives / personas - and - the - advantage - of - designing - for yourself/ (Retrieved 31/01/2012) (cit. on p. 128). Postigo, Hector (2003). “Emerging Sources of Labor on the Internet: The Case of America Online Volunteers”. In: International Review of Social History 48.Supplement S11, pp. 205–223 (cit. on pp. 34, 36, 69).

158

References

Postigo, Hector (2007). “Of Mods and Modders”. In: Games and Culture 2.4, pp. 300– 313 (cit. on p. 34). Preece, Jenny, Yvonne Rogers and Helen Sharp (2002). Interaction Design: Beyond Human-Computer Interaction. New York, NY: J. Wiley & Sons (cit. on p. 18). Pruitt, John and Tamara Adlin (2006). The Persona Lifecycle: Keeping People in Mind Throughout Product Design. Amsterdam; Boston: Elsevier: Morgan Kaufmann Publishers (cit. on pp. 18, 38, 97). Rheingold, Howard (1993). The Virtual Community: Homesteading on the Electronic Frontier. The MIT Press (cit. on pp. 34, 36). Righi, Carol and Janice James (2007). User-Centered Design Stories: Real-World UCD Case Files. Amsterdam: Elsevier/Morgan Kaufmann (cit. on pp. 5, 42). Rogers, Yvonne (2004). “New theoretical approaches for human-computer interaction”. In: Annual Review of Information Science and Technology 38.1, pp. 87–143 (cit. on pp. 5, 15, 16). Rogers, Yvonne (2009). “The Changing Face of Human-Computer Interaction in the Age of Ubiquitous Computing”. In: HCI and Usability for e-Inclusion. Ed. by A. Holzinger & K. Miesenberger. Vol. USAB 2009, LNCS 5889. Lecture Notes in Computer Science. Springer, pp. 1–19 (cit. on p. 15). Rosenberg, Richard S (1997). The Social Impact of Computers. San Diego, Calif: Academic Press (cit. on p. 24). Ross, M, J Nowicki, D Solomon, L. Yarbrough and C. Schwendeman (2000). Designing the User Experience, the Usability Life Cycle. Usability Professionals’ Association, poster (cit. on p. 11). Rubin, Jeffrey (1994). Handbook of Usability Testing: How to Plan, Design, and Conduct Effective Tests. New York: Wiley (cit. on p. 15). Salovaara, Antti, Mikael Johnson, Kalle Toiskallio, Sauli Tiitta and Marko Turpeinen (2005a). “Nurturers: A creative player segment and their motivations in multiplayer games.” In: Creative Gamers – Exploring Participatory Culture in Gaming. Hypermedia Laboratory, University of Tampere, Finland (cit. on p. 59). Salovaara, Antti, Mikael Johnson, Kalle Toiskallio, Sauli Tiitta and Marko Turpeinen (2005b). “Playmakers in multiplayer game communities: their importance and motivations for participation”. In: Proceedings of the ACE 2005. Proceedings of the 2005 ACM SIGCHI International Conference on Advances in computer entertainment technology. Valencia, Spain: ACM, pp. 334–337 (cit. on p. 59). Sarvas, Risto and David M. Frohlich (2011). From Snapshots to Social Media - The Changing Picture of Domestic Photography. Springer (cit. on pp. 25, 43). Schaffer, Eric (2004). Institutionalization of Usability: A Step-by-Step Guide. Boston, Mass: Addison-Wesley (cit. on p. 17). Scholz, Trebor (2008). “Market Ideology and the Myths of Web 2.0”. In: First Monday 13.3 (cit. on p. 25). Schuler, Douglas and Aki Namioka (1993). Participatory Design: Principles and Practices. Mahwah, NJ, USA: Lawrence Erlbaum Associates (cit. on p. 20). Seale, Clive, Giampietro Gobo, Jaber F. Gubrium and David Silverman (2004). “Quality and Credibility”. In: Qualitative research practice. Ed. by Clive Seale, Giampietro Gobo, Jaber F. Gubrium and David Silverman. London; Thousand Oaks, Calif: SAGE, pp. 378–379 (cit. on p. 58). Sengers, Phoebe, Kirsten Boehner, Shay David and Joseph ’Jofish’ Kaye (2005). “Reflective design”. In: Proceedings of the 4th decennial conference on Critical

References

159

computing: between sense and sensibility. CC ’05. New York, NY, USA: ACM Press, p. 49 (cit. on p. 46). Sengers, Phoebe, John McCarthy and Paul Dourish (2006). “Reflective HCI: articulating an agenda for critical practice”. In: CHI ’06 Extended Abstracts on Human Factors in Computing. New York, NY, USA: ACM Press, pp. 1683–1686 (cit. on p. 46). Shapiro, Dan (1994). “The limits of ethnography: combining social sciences for CSCW”. In: Proceedings of the 1994 ACM conference on Computer supported cooperative work. CSCW ’94. New York, NY, USA: ACM, pp. 417–428 (cit. on p. 16). Sharp, Helen, Yvonne Rogers and Jenny Preece (2007). Interaction Design: Beyond Human-Computer Interaction. Chichester; Hoboken, NJ: Wiley (cit. on p. 18). Shirky, Clay (2003). A Group Is Its Own Worst Enemy. Keynote on Social Software, O’Reilly Emerging Technology conference, Santa Clara, April 24, 2003. Published July 1, 2003 on the "Networks, Economics, and Culture" mailing list. http://shirky.com/writings/group_enemy.html (Retrieved 31/01/2012) (cit. on p. 29). Shirky, Clay (2008). Here Comes Everybody: The Power of Organizing Without Organizations. New York: Penguin Press (cit. on pp. 7, 24, 25). Shneiderman, Ben (1998). Designing the User Interface: Strategies for Effective Human-Computer-Interaction. 3rd ed. Reading, Mass: Addison Wesley Longman (cit. on p. 43). Sihvola, Sebastian (2005). “Netari.fi Nuorisotyö Habbo Hotelissa”. B.Sc. Thesis. HUMAK, University of Applied Sciences (cit. on p. 52). Sihvonen, Tanja (2009). Players Unleashed! Modding The Sims and the Culture of Gaming. Doctoral thesis (monograph) (cit. on pp. 34, 79). Smith, Marc A and Peter Kollock (1999). Communities in Cyberspace. London; New York: Routledge (cit. on p. 36). Spillers, Frank (2006). Designing for the ’average user’. Demystifying Usability blog. http://www.demystifyingusability.com/2006/04/designing_for_t.html (Retrieved 31/01/2012) (cit. on p. 83). Stewart, James and Sampsa Hyysalo (2008). “Intermediares, Users and Social Learning in Technological Innovation”. In: International Journal of Innovation Management 12.3, pp. 295–325 (cit. on pp. 1, 35, 45, 69, 121, 129). Stewart, James and Robin Williams (2005). “The wrong trousers? Beyond the design fallacy: Social learning and the user”. In: Handbook of critical information systems research: Theory and application. Ed. by D. Howcroft and E. Trauth. Cheltenham, UK: Edward Elgar, pp. 195–221 (cit. on pp. 43, 86, 120). Stockstrom, Christoph, Christian Lüthje and Yun Mi Antorini (2010). “User-generated techniques: The case of the Adult Fans of LEGO”. Role of Users Workshop, Helsinki, August (cit. on p. 129). Stolterman, Erik (2008). “The Nature of Design Practice and Implications for Interaction Design Research”. In: International Journal of Design 2.1 (cit. on p. 16). Straub, Kath, Mona Patel, Arno Bublitz and Javier Broch (2009). The HFI UX Maturity Survey. Tech. rep. Human Factors International. http://www.humanfactors. com/UXMSurvey-2009/ (cit. on p. 17). Strauss, Anselm L (1978). “A Social World Perspective”. In: Studies in Symbolic Interaction 1, pp. 119–128 (cit. on p. 56). 160

References

Strauss, Anselm L and Juliet M Corbin (1990). Basics of Qualitative Research: Grounded Theory Procedures and Techniques. Newbury Park, Calif: Sage Publications (cit. on p. 131). Suchman, Lucy (1987). Plans and Situated Actions: The Problem of Human-Machine Communication. Cambridge [Cambridgeshire]; New York: Cambridge University Press (cit. on p. 16). Suchman, Lucy (1994a). “Do categories have politics?” In: Computer Supported Cooperative Work (CSCW) 2.3, pp. 177–190 (cit. on pp. 47, 89, 121). Suchman, Lucy (1994b). “Working relations of technology production and use”. In: Computer Supported Cooperative Work (CSCW) 2.1-2, pp. 21–39 (cit. on pp. 35, 43). Suchman, Lucy (2002). “Located accountabilities in technology production”. In: Scandinavian Journal of Information Systems 14.2 (cit. on pp. 35, 46, 126). Sulake (2012). Quick Habbo facts. http : / / www. sulake . com / habbo/ (Retrieved 31/01/2012) (cit. on pp. 9, 10). Svanæs, Dag and Jan Gulliksen (2008). “Understanding the context of design: towards tactical user centered design”. In: Proceedings of the 5th Nordic conference on Human-computer interaction: building bridges. Lund, Sweden: ACM, pp. 353– 362 (cit. on pp. 17, 119, 122, 126). Sy, Desiree (2007). “Adapting Usability Investigations for Agile User-Centered Design”. In: Journal of Usability Studies 2.3, pp. 112–132 (cit. on p. 102). Tapscott, Don and Anthony D. Williams (2006). Wikinomics: How Mass Collaboration Changes Everything. Portfolio Hardcover (cit. on p. 3). Tapscott, Don and Anthony D Williams (2010). Macrowikinomics: Rebooting Business and the World. New York: Portfolio/Penguin (cit. on pp. 25, 26, 34). Thomas, Peter J. (1995). The social and interactional dimensions of human-computer interfaces. Cambridge University Press (cit. on pp. 16, 46). TSR (n.d.). The Sims Resource. http : / / www. thesimsresource . com (Retrieved 31/01/2012) (cit. on p. 79). Turner, Fred (2005). “Where the Counterculture Met the New Economy: The WELL and the Origins of Virtual Community”. In: Technology and Culture 46.3, pp. 485–512 (cit. on p. 34). Tushman, Michael L. and Philip Anderson (1986). “Technological Discontinuities and Organizational Environments.” In: Administrative Science Quarterly 31.3, pp. 439–465 (cit. on pp. 27, 48). Usability Professionals’ Association (n.d.). What is User-Centered Design? http : //www.usabilityprofessionals.org/usability_resources/about_usability/what_ is_ucd.html (Retrieved 31/01/2012) (cit. on p. 1). UsabilityNet (2003). UsabilityNet: usability resources for practitioners and managers. http://www.usabilitynet.org/ (Retrieved 31/01/2012) (cit. on pp. 5, 17). Van de Ven, Andrew H., Douglas E. Polley, Raghu Garud and Sankaran Venkataraman (1999). The Innovation Journey. New York: Oxford University Press (cit. on pp. 27, 28, 48, 129). Venturi, Giorgio and Jimmy Troost (2004). “Survey on the UCD integration in the industry”. In: Proceedings of the third Nordic conference on Human-computer interaction. NordiCHI ’04. Tampere, Finland: ACM, pp. 449–452 (cit. on p. 5). Venturi, Giorgio, Jimmy Troost and Timo Jokela (2006). “People, Organizations, and Processes: An Inquiry into the Adoption of User-Centered Design in In-

References

161

dustry.” In: International Journal of Human-Computer Interaction 21.2, pp. 219– 238 (cit. on p. 17). Voss, Alex, M. Hartswood, Rob Procter, M. Rouncefield, Roger Slack and Monica Büscher, eds. (2009). Configuring User-Designer Relations: Interdisciplinary Perspectives. London: Springer (cit. on pp. 8, 21, 22). Vredenburg, Karel, Scott Isensee and Carol Righi (2002). User-centered design: an integrated approach. Upper Saddle River, NJ: Prentice Hall PTR (cit. on p. 11). Vredenburg, Karel, Ji-Ye Mao, Paul W. Smith and Tom Carey (2002). “A survey of user-centered design practice”. In: Proceedings of the SIGCHI conference on Human factors in computing systems: Changing our world, changing ourselves. CHI ’02. New York, NY, USA: ACM, pp. 471–478 (cit. on p. 5). Ware, Colin (2003). “Design as Applied Perception”. In: HCI Models, Theories, and Frameworks: Toward a Multidisciplinary Science. Ed. by John Carroll. Morgan Kaufmann, pp. 11–26 (cit. on p. 44). Westrup, C. (1997). “Constituting users in requirements techniques”. In: Information systems and qualitative research. Ed. by A.S. Lee, J. Liebenau and J.I. DeGross. Proceedings of the IFIP TC8 WG 8.2 international conference. London, UK, UK: Chapman & Hall, Ltd., pp. 183–204 (cit. on pp. 43, 45). Williams, Robin, James Stewart and Roger Slack (2005). Social learning in technological innovation: Experimenting with information and communication technologies. Cheltenham, UK: Edward Elgar (cit. on pp. 4, 8, 22, 35, 43, 45, 129). Winner, Langdon (1980). “Do Artifacts Have Politics?” In: Daedalus 109.1, pp. 121– 136 (cit. on pp. 89, 121). Winner, Langdon (1982). “Energy Regimes and the Ideology of Efficiency”. In: Energy and Transport: Historical Perspectives on Policy Issues. Ed. by George H Daniels and Mark H Rose. Sage focus editions 52. Beverly Hills: Sage Publications, pp. 261–277 (cit. on p. 46). Winner, Langdon (1993). “Upon Opening the Black Box and Finding It Empty: Social Constructivism and the Philosophy of Technology”. In: Science, Technology, & Human Values 18.3, pp. 362–378 (cit. on pp. 47, 48). Wixon, Dennis, Karen Holtzblatt and Stephen Knox (1990). “Contextual design: an emergent view of system design”. In: Proceedings of the SIGCHI conference on Human factors in computing systems: Empowering people. CHI ’90. New York, NY, USA: ACM, pp. 329–336 (cit. on p. 16). Woolgar, Steve (1991). “Configuring the user: The case of usability trials”. In: A sociology of monsters: Essays on power, technology and domination. Ed. by John Law. London: Routledge, pp. 57–102 (cit. on pp. 35, 45). Woolrych, Alan, Kasper Hornbæk, Erik Frøkjær and Gilbert Cockton (2011). “Ingredients and Meals Rather Than Recipes: A Proposal for Research That Does Not Treat Usability Evaluation Methods as Indivisible Wholes”. In: International Journal of Human-Computer Interaction 27.10, pp. 940–970 (cit. on p. 17). Zimmerman, John (May 2011). “Killing off user-centered design”. In: interactions 18.3, pp. 10–11 (cit. on p. 124).

162

References

Publications

I


165

II


187

III


209

IV


239

V


263

VI


287

VII The Virtuality of Virtual Worlds, or What We Can Learn from Playacting Horse Girls and Marginalized Developers 299

163

ARTICLE

I


Mikael Johnson and Kalle Toiskallio (2005). "Fansites as Sources for User Research: Case Habbo Hotel". IRIS’28 Conference. The 28th Information Systems Research Seminar in Scandinavia. Kristiansand, Norway, August 2005. Reprinted with permission.

165

Fansites as Sources for User Research: Case Habbo Hotel Mikael Johnson and Kalle Toiskallio Helsinki University of Technology Software Business and Engineering Laboratory Email addresses: @hut.fi Abstract. The online chat and game environment Habbo Hotel has become a hit among Finnish high school teenagers, as about 25% of them visit Habbo at least once a month. Like in the case of other popular computer games such as The Sims, fandom websites have emerged around Habbo since its launch in August 2000. The fansites are important because they shape the community by providing the arena for public Habbo discussions, which mediates user opinions to a large audience. In this paper we investigate the fansite content to find out what one can learn about the social world of Finnish Habbo users. We identified 173 Finnish fansites, of which we selected 23 fansites aimed at a large audience. These fansites publish news, hints, reviews, lists, graphics, and other articles about essential Habbo topics. Based on this fansite content, it was possible to distinguish 8 different user groups and 11 popular activities. Therefore we draw the conclusion that fansites are useful sources for user research as they complement other sources informing information systems design.

Introduction The growth of networked computers in home and entertainment contexts has increased the importance of virtual communities. These emerging contexts of use challenge existing information systems research with their large numbers of users in non-work contexts, where fun and enjoyment (Blythe et al. 2003) is more important than usefulness. For instance 430 000+ people around the world play Everquest, a medieval fantasy game (Taylor 2003), and four million Koreans play

1

the game Lineage (Woodcock 2005). The popularity of the games is also demonstrated by the emerging web-based fan culture around the games. Fansites have emerged in the U.S. around the single-player game The Sims, where one can create virtual characters and manage their careers and lives. A website called The Sims Resource has a fansite index, where the top four fansites have more than 5 million hits. The fansites offer the latest about the Sims, including news, articles, reviews and lots of extensions and modifications to the game for downloading (The Sims Resource 2005). In this paper we investigate the fansites around the online chat and game environment Habbo Hotel. We distinguish fansites from personal homepages, since the fansite authors use their nickname from Habbo, thereby not revealing their real names. Although some fansites are created by individuals, most of them are made by small groups, therefore resembling what Döring (2002) calls collective personal home pages. Habbo is designed for teens and has become a very popular and fast growing teen website in sixteen countries1. About three million unique users visit Habbo each month, and roughly 22 million Habbo characters have been created (Sulake 2005). It is not a game as such, but a very flexible collaborative game environment that provides the resources for play. The users themselves can generate their goals (collecting furniture, making friends, competing, etc.) or choose to focus on the chat and hang around with friends. Since fansites studies are only beginning to emerge, the research question is fairly general at this stage: Are they useful as sources for user research? We do not focus on technical issues behind the websites, but rather on two subquestions: 1) What kind of content can be found on the fansites? 2) What can we learn about user groups and popular activities? Our empirical data originates from the Finnish community around Habbo. We concentrate on the official and active fansites, while also noting fansites on a break, inactive, and inaccessible ones. In our Habbo survey from Summer 2004, about 50% of the respondents visit some fansite at least once a week. The survey showed that the fansites are broadly known among Habbo users and many visit them as often as the hotel. The fansites play an integral part in the community and are therefore an important object of research. Fansite analysis might also prove to be a cost-effective user research method in situations like Habbo where there are a large number of geographically distributed users, of which many gather around popular fansites.

Previous Work Computer games have been researched for more than a decade from technical, aesthetic and socio-cultural perspectives by a growing number of researchers with different backgrounds (Aarseth 2003, Brown & Bell 2004, Dourish 1998). 1 Mainly in Western Europe and North America, but also Australia, Japan, and Singapore are included.

2

Fansites have not been a research topic yet, since most research has focused on the gameplay itself and the question of what constitutes a good game (Konzack 2002). Some attention has been paid to the ’playground’ where the gameplay takes place, and concepts like off-game communication and social playability have emerged (Järvinen et al. 2002, Ermi et al. 2004). While the origins of Habbo can be traced back to research on textual online chats and games like MUDs and MOOs (Cherny 1999, Curtis 1994, Dibbel 1993, Mnookin 1996), this paper leans on research in computer-mediated communication and user research. User research (Hackos & Redish 1998, Beyer & Holtzblatt 1998, Kujala 2002, Kuniavsky 2003) can be seen as a subtopic of the larger theme participation in information systems development – a core topic of information systems research since the 1960’s (Markus & Mao 2004). Usually user research is based on the idea that knowing more about the users will help designers create better products that meet the user needs. User research typically includes identifying user groups, understanding their context of use (ISO 9241-11 1998) by visiting user sites and making interviews and observations of the users in their own environment. However, as Henriksen (2002) reported, the notion of one bounded field site is problematic when investigating distributed information systems. As a consequence she turned to a multi-sited research design, where the research sites are regarded as an open-ended space of possibilities. We follow this line of thinking in our investigation of the Habbo fansites and regard them as important first-hand sources for user research, complementing other sources and methods (mainly different kinds of interviews and observations). Research on computer-mediated communication (CMC) has also been published in the field of information systems: for instance anxiety and CMC (Brown et al. 2004) and web home page complexity (Geissler et al. 2001). Nicola Döring (2002) has made a systematic review of about thirty home page studies. She summarises how theories of identity, self-presentation and CMC have been applied to personal home pages. To integrate the diverse empirical findings, she structures the findings by looking at homepages as media products with production processes, product characteristics, and reception processes. Focusing here on the product characteristics, the homepages can be classified based on whether they are available, under construction, or completed. The realized homepages can be divided into expressive (the topic is one’s own person) and instrumental (another topic without explicit reference to one’s own person). The core elements of personal homepages are e-mail links, external links, a welcome message, 1-4 graphics, and a brief biography. This classification and the core elements inspire our fansite analysis, especially the list of fansite elements. One specific type of personal homepages that has recently attracted research attention is the weblogs. Weblogs, or just blogs, are frequently updated websites where the authors post their opinions and links to other sites, commenting on topics close to their heart. The news section on the fansites are very similar to

3

blogs, which makes blogs relevant here. The scholarly research on blogs is emerging mostly at scientific conferences such as Computers and Writing. Some blogs have a large audience and interact with other media. Drezner & Farrel (2004) report on the influence of blogs on the American political debate. According to Gurak et al. (2004), what characterises the blogs as a genre is their form and function: “all posts to the blog are time-stamped with the most recent post at the top, creating a reverse chronological structure governed by spontaneity and novelty”. Kuniavsky (2004) noted that blogs are one source to others’ hard work and they can contain important, but unedited, insights. Our research is also informed by fandom research that originates in the studies of scifi magazines beginning in the 1920’s and 1930’s in the U.S. (Jenkins 1992). These magazines that fans corresponded through were precursors to current fan magazines, or fanzines, which in turn can be seen as precursors to the fansites were discussing in this article.

Case Habbo Hotel Based on experiences from two Internet chat rooms (Mobiles Disco and Lumisota), the game development company Sulake Labs launched the first Habbo Hotel in Finland in August 2000. People familiar with Internet technologies describe Habbo as an audiovisual chat environment on the web that can be accessed with a web browser with the Shockwave plug-in. Others talk about a virtual hotel where one can hang out and make new friends. When checking in to the virtual hotel one creates one’s own cartoon like Habbo avatar (Figure 1) that can walk, dance, eat, drink and chat in the cafés (Figure 2), restaurants, swimming pools and games rooms. Besides experiencing these common rooms in the hotel, one can decorate and furnish a room of one’s own.

Figure 1. A Habbo avatar.

Compared to popular massively multiplayer online games (MMOGs), Habbo’s 3 million users seems to be fairly large number, only the Korean Lineage with 4 million subscriptions is in the same magnitude. Everquest has about 0.4 million users and Ultima Online and a few others have about 0.2 million users

4

(Woodcock 2005). Habbo is also huge compared to other non-competitive virtual worlds that are marketed as ’virtual getaways’. The number of active users in Active Worlds, There, and Second Life is measured in tens of thousands (Activeworlds Corp. 2005, Forterra 2004, Woodcock 2005).

Figure 2. A café in Habbo.

It is difficult to exactly measure the number of Finnish Habbo users, since the popular measuring techniques are counting the number of browsers from with users log on to Habbo. The organization that publishes Finnish web use statistics, TNS Gallup, reports about 90 000 users per week (09/2005) or about 278 000 users per month (03/2004). Their statistics show that the Habbo website has been among the top 20 most visited sites in Finland since 08/2003. According to our research in Summer 2004 95% of the users are between 10 and 19 years old. Assuming about 160 000 users would imply that about 25% of the Finnish 10-19 years old population visit Habbo. Unique selling points of Habbo, compared with the above mentioned MMOGs, could be that it is intended for teenagers, since one’s avatar is a boy or a girl – not a grown-up. Habbo is fairly easy to start, since it is not a separate program, but works within the web browser. Also, one does not have to choose lots of character

5

capabilities or attributes for one’s avatar, just gender and clothes. Habbo has no subscription fee, which is usually difficult for teenagers to pay, it is free to chat, but some extra hotel services cost money. The site is a hotel, which is different from the outdoor setting of most MMOGs. From a parent perspective Habbo is a non-violent and moderated environment, which makes it a reasonably safe place for kids. However, these are also the reasons for why one might not want to visit Habbo. There are a lot of kids there who prefer talking to their friends and not to adults. If more options in changing avatars and the virtual world is wanted, other virtual worlds provide more alternatives. Also, one has to follow the Habbo Way, which basically means being nice and polite, and not swear nor cheat, which is far from the ‘anything goes’ attitude of some games.

Research Method & Data We follow the strategies for ethnographic research on virtual communities outlined by Paccagnella (1997) and Jones (1997). Fansites exhibit a fair amount of interpretative flexibility – “they can mean different things to different individuals or different groups, and their use continues to be interpreted and reinterpreted with the passing of time” (Paccagnella 1997). It also means that fansites can be studied from many different perspectives, we could discuss their role in the social worlds of Habbo, their potential for user feedback, their technical implementation, or the copyright issues related to them. Here we choose to focus on what we can learn about the users and their social worlds around Habbo from a user research perspective. We are interested in what kind of user groups emerge and what kinds of activities in Habbo are popular. Even though this is not an identical study, we present a similar kind of argument as Bøving & Simonsen did a year ago in SJIS (2004). They presented an approach to studying the use of web-based information systems, focusing on web usage statistics from http log files on web servers as data. Their argument was that results from http log analysis might support other data collection methods such as surveys, interviews, and observation. They showed that http log analysis could be valuable in both quantitative and qualitative research. Here in our paper, we will argue that fansite analysis can fill a similar role, complementing the mentioned sources for user research. The term virtual community has become fashionable, unfortunately creating more confusion than order in the academic literature. Paccagnella considers it a problematic scientific concept, because there is not a commonly accepted definition of what a community really is. He ends up using the term virtual community as a “metaphor to indicate the articulated pattern of relationships, roles, norms, institutions, and languages developed on-line” (1997). The important thing here is to note that a fansite is not a virtual community, it is a just

6

a ‘window’ through which we look at the virtual community. We can elaborate on this by discussing Jones approach to cyber-archaeology. Jones (1997) notes that one must distinguish between the virtual community and its place on the Internet. He defines a set of minimum conditions for the place, which he calls a virtual settlement: 1) a minimum level of interactivity, referring to a dialogue in messages, where later messages refer to earlier ones 2) a variety of communicators, more than two individuals communicating on 3) a virtual common-public-space where a significant portion of interactive group-computer-mediated communications occur 4) a minimum level of sustained membership, thereby excluding temporary and fleeting aggregations of people, e.g. Smart Mobs (Rheingold 2002) Jones considers finding a virtual settlement proof for the existence of a virtual community. Looking at the fansites alone, we find that they do not qualify as virtual settlements, since the level of interactivity is fairly low and only a few fansites feature a discussion forum, which could be an interactive virtual public space. However, together with the interaction in the virtual public spaces of Habbo we find that there is evidence of virtual settlements. Fansites differ from traditional sources for user research, because they are accessible without researcher intervention. This can be a benefit compared to the problems arising from the presence of the researcher in the field doing participatory observation, for instance. This obviously causes new ethical issues, but it reduces the danger of distorted data by the presence of the researcher. The ethical issues concern how to threat electronic discussions, for instance would the researcher need permission to analyse the fansites? Here we adopt Paccagnella’s advice, the public discussions on the fansites are considered public acts deliberately intended for public consumption – therefore no more precautions than common in other studies of everyday life in public spaces or mass media communications are necessary. It is ok to quote fansites providing that the source is mentioned, and names need not be held confidential as in private interviews. Some preliminary qualitative background data was drawn from semi-structured contextual theme interviews with a couple of moderators (15 and 22-years old), and an active player (11-years old), the community director of Sulake Labs, and from the conversations and correspondence with a handful of the producers. The second background data stems from a web-based questionnaire with 27 questions, of which 21 were multiple-choice and 6 were open. A link to the web form was put on the Finnish Habbo frontpage. 10 613 visitors answered during the two weeks the link was there (June 22 - July 6, 2004). This data is, however, used in this article only as descriptive background figures, because here we focus on the fansites. The more analytic usage of it will be published elsewhere. Data collection. For this particular fansite study we have identified 173 Finnish fansites in total, of which 5 became official fansites in June 2004. As of

7

December 2004 we found that 32 of the rest were active, 12 on a break, 59 had no updates during the last three months, and 65 were inaccessible or the content was removed. Many of the fansites have a name with ’kala’ in it, which means fish in English. The reason is that the Finnish Habbo started out as Hotelli Kultakala (Hotel Goldfish) and anything annexed with ’kala’ seems to be cool. We first analysed the official fansites (Kala-Stamo, Monitori, Nefekala, Nerokala, Pixeli), one of the biggest closed fansites (Paratiisikala), and 4 random fansites (Aalto, Palmu, Rausku, Wannabe-kala) in detail. This body of articles on the web forms the basis for our qualitative content analysis, resulting in the list of fansite elements. After this, we validated and improved the list based on the contents of 13 additional fansites. Finally, we clicked through the remaining 85 accessible fansites to check for new elements that our list did not include. We found none, which confirmed our results.

Habbo Fansites The most popular fansites are usually made by a small team of Habbo fans with different expertise and roles. One designs the layout and look & feel, another one writes the stories and a third one has the technical skill to publish them on the web. We have not systematically studied the age and gender of the fansite authors, because many fansite authors remain anonymous and only reveal their nickname in Habbo. However, some do, and among them we note that some of the youngest fansite authors are 10 and 11 years old (e.g. Sekofisu). Many are between 13 and 15 years old. The most active fansites have smaller updates (like news and rumours about Habbo) several times per week, but publish reviews and articles once a week or bi-weekly. Some fansites group their articles together and publish them as an issue of a webmagazine. The webmagazines seem to follow a rhythm of one issue per one or two months. The form and content of the fansite is created with a wide variety of software tools. The simplest way is maybe using the html editing facilities on a free web hosting service like Freewebs.com. Another way is to write in Word and save the article in the html format, after which it is uploaded to the Internet service provider that hosts the website. Html-savvy use Notepad or an html-editor. A few fansite teams have skilled webgurus and set up sites using scripting techniques (e.g. php) and databases (e.g. MySQL). Most fansite functionality is based on text and images on static webpages, but some allow the fansite audience to participate through guest books, gallup polls, and forums. Figures 3 and 4 show examples of two fansites.

8

Figure 3. Kala-Stamo, one of the official fansites.

Figure 4. Pixeli, another official fansite.

9

Results In this section we present the results to the first research subquestion: what kind of content can be found on the fansites? The following Habbo-related topics seem to be the common elements on the fansites. They are based on a bottom-up classification and examples are given for each category. After the summarizing table, in the following subsections, we answer the other subquestion regarding what we can learn about user groups and popular activities.

Fansite Elements News and rumours. The fansites are very active news and rumour distributors. News items are often placed in a corner of the frontpage. A hot topic in April 2004 was the changed way of ordering Habbo coins through SMS and with which mobile phone operators it worked. The news also included the Habbo happenings during Easter: searching for a bunny, drawing competitions, background images and e-postcards. A software bug in the swimming pool caused some buzz. (Pixeli) Guest books and gallup polls provide the fansite audience a way to participate comment on the fansite. Fansite visitors post greetings and polite comments on the good work as well as rude ones. Some try to market their own homepages or fansites, and others ask questions about the fansite development. (Nerokala, Pixeli). For instance Paratiisi-Kala asks if the readers have cheated furniture, and 28% of 1931 respondents say yes. Other fansites. Some of the studied fansites comment on each other. Reviewing ‘official’ Habbo stuff made by Sulake is one thing, but reviewing other fansites is a delicate matter. For instance, in an article at Nefekala it is evident that the article & fansite author is offended by a harsh review of Nefekala at Pixeli. Clearly Nefekala ‘gives back’ in their review of Pixeli. The official fansites have large link collections to other fansites, between 40 and 50 Finnish fansites were linked. Nerokala and Pixeli also included ratings of the other fansites, and Nerokala even provide click through statistics. Hints. Once one has learned the basics in the hotel, one starts to notice certain things that are not mentioned in the official tutorials. How do they make those special characters in the room names, and how does one change the look on the face of the Habbo character? The hints provide answers to these questions and more. For instance what the curious shorthand writing common in chats mean – LOL means laughing out loud, and so on. Reviews and lists. Paratiisi-Kala provides excellent reviews of all public spaces in the hotel, e.g. what one can do in each room, how the room is designed, how many Habbo characters will fit simultaneously, the name of the bots, etc. All the fansites give information on what virtual pets there are, what they can do and everything about furniture. What furniture one gets when joining the Habbo Club, rare collection items, seasons-related furniture, etc.

10

Secrets. These are things that one can impress one’s Habbo friends with. How to move furniture faster with the Alt-key. Walking through chairs. Making furniture float in the air. Talking to the bartender bots, ordering special drinks with secret phrases, etc. (Pixeli). Guidelines. Some articles on the fansites show serious effort to set boundaries for acceptable behaviour in Habbo. They go much further than the official rules of the Habbo way. Nerokala presents 20+1 rules in an article on how to become ‘fish’-credible. It starts with a rule to read the character description before starting a conversation, to avoid asking unnecessary questions. The second rule advices not to ask for age nor location right away, because it’s ‘childish and stupid.’ It continues by stating how ‘silly, embarrassing, and forbidden’ all sorts of bad manners and cheating are. Pixeli has an article in the same way of writing, how to be ‘idle’ in the hotel without disturbing others, not shouting in public spaces, making friends that matters (not just ‘collecting’), as well as basing discussions on arguments, instead of yes-no debates. Histories. Two major histories are told on the fansites, the history of Habbo and the history of that particular fansite. The Habbo history presented by Nerokala runs along two intertwined threads. The first explains how the hotel structures were developed by Sulake, which rooms and furniture items were created when. The other thread describes the development of the Habbo community and how the hotel structure changes were received. The fansite authors usually present their fansite history and themselves on a ‘staff’-page. It includes stories of what personal reasons lay behind the decision the make the fansite and who were invited to the editorial staff. Advice on making a fansite is given, where to publish them, being original compared to previous fansites, and even how to make money with the aid of the fansite. Fashion and celebrities. Pixeli has an article ‘Fashion for the early Winter’, where Muotiminna gives hints on what’s fashionable for boys and girls. Boys are recommended a long-sleeve shirt, and girls are recommended to show their belly below their t-shirt. Paratiisikala (no. 8) has done a Habbo interview with snowgirl90 and .:Eminemi:., who recommend the hottest thing: the recently introduced public room ‘Kahvila Kaneli’ and the drink ‘kuplamehu’ (bubblejuice). Unfashionable furniture are digi-tv sets and Dew-sofas. Nerokala has published extensive lists and screenshots of celebrities like 23 Habbo staff members and 97 of the about two hundred moderators. International Habbo. All four studied fansites maintain links to Habbo Hotels in other countries. Especially the British Habbo is important, since the Habbo Club was introduced there a few weeks before Finland. This means that one can go there and peek on what the next Habbo Club furniture will be. Similarly the Finnish Habbo was exceptionally popular abroad in February 2004, when it was the first hotel where the virtual pets were announced. Nerokala also tells the story about Veijo’s great Habbo-adventure to the British Habbo.

11

Meetings IRL. Since Habbo is intended to be a safe place for everyone, 1-on-1 meetings outside Habbo are discouraged. Instead, following the Habbo Way, open meetings for groups of Habbo players are encouraged. Fansites provide the information for when and where to meet. Competitions. The fansites present competitions both in Habbo and on the fansite. Inside Habbo it can be an official competition, such as the ‘find a bunny’ contest during Easter arranged by the Hotel Manager, or an unofficial competition invented by a hotel visitor. Some fansites organise competitions and lotteries regarding the fansite, to engage fansite readers in their fansite. Pictures. Edited screenshot pictures are an integral part of many fansites. Pixeli even provides a drawing school on how to make good pixel graphics. Table 1 presents the above outlined fansite elements in more detail. The fansite ID columns refer to a particular fansite listed in Appendix 1. Lots of pluses imply that a fansite element is very typical. Very few pluses imply a fansite element that is unique to only a few fansites, which is either a big strength in profiling or an element that all others consider unnecessary. The fansite elements in between can also be important, signifying emerging elements that only some have implemented. This list reflects the selection of official and active fansites aimed at a large audience in the community. Fansite type Fansite ID News & rumours Participation guest book gallup polls e-mail address feedback form disc. forum competitions stories by others irc link toplist vote file upload Links other fansites int. Habbo int. fansites other category forums Hints, secr., guidel. Reviews and lists public spaces priv. guest rooms HabboClub furni rare furniture virtual pets other fansites hobbas/staff/eoph Histories Habbo history Fansite history Fashion & celebs Interviews Magazine

Official AB CDE + + + + +

Active F GH I J K L MN - - + + + + + + +

Break OP Q + + -

Inactive RS T UVW + + - + + -

∑ 18

+ + + + -

+ + + + + + + -

+ + + + + + -

+ + + + + + + -

+ + + + -

+ + +

+ + + + + -

+ + + + + +

+ + + + + -

+ + -

+ + + + -

+ + + + + + + -

+ + + + + + -

+ + + + -

+ + + + + -

+ + + + +

+ -

+ + + -

+ + + + -

+ + -

+ + + + + -

-

+ + -

17 11 21 12 7 14 9 1 2 3

+ + + + +

+ + + +

+ + + +

+ + + + +

+ + + +

+ +

+ + + +

+ +

+ + + +

+ -

+ + + + +

+ + + +

+ + + + +

+ + + -

+ + +

+ + + +

+

+ + + +

+

+ -

+ +

+ + + +

+ +

20 16 3 12 3 20

+ + + + -

+ -

+ + + -

+ + + + +

+ + -

+ + + -

+ -

+ + +

+ + -

-

+ + + + -

+ + -

+ + -

+ -

+ -

+ -

-

+ -

+ + + + -

+ -

+ + + + + + -

+ -

-

7 5 8 9 2 15 2

+ + +

+ + +

+ + -

+ + -

+ + + -

+ + -

+ + -

+ + + +

+ + + +

-

+ + + -

+ + +

+ + + + +

+ -

+ + + -

+ + + -

-

-

+ + +

-

+ + +

+ -

+ -

3 8 16 16 8

12

Graphics Pics, screenshots + + Drawing school + Habbo Fiction + About the fansite and authors Fansite staff + + Application form - FAQ + + Page counter + Updates + + Banners + + Sulake's © notice + + Downloads + + Real life Meetings IRL - RL / blog - + IRL Gallery - E-cards - +

+ - + - - + - - -

+ + + + - + + + + - - - - - - - - - - - - - - - - -

+ + + + + - -

- + - + + + - - - - - - - - + - -

18 4 3

+ + + + -

+ + + + + + +

+ + + + + + +

+ + -

+ + + + + + -

+ + + + + + +

+ + + + + + +

+ + + -

+ + + + + +

+ + + + + + + +

+ + + + + + -

+ + + + + -

+ + + + + +

+ + + + + + + +

+ -

+ + + + + + -

+ + -

+ + + + + -

+ + + + +

+ + + -

+ + + -

23 6 12 14 21 19 14 11

-

-

+ + -

-

-

+ -

+ -

-

+ -

-

-

-

-

+ -

-

-

-

-

+

+ -

-

2 5 1 2

Table I. Detailed list of fansite elements.

User Groups Here we turn to the second subquestion: what can we learn about user groups and popular activities? We identified two articles on two big fansites that were discussing stereotypical Habbo users. These stereotypes can be seen as potential user groups in Habbo. In the Nerokala article ‘Kalaluokat’ Andna suggests that the Habbo characters can be classified into novices, amateurs, pros, and superhabbos. In this classification a novice is someone who asks simple questions, does not know the Habbo rules, and doesn’t know how to get furniture. An amateur has more ‘streetcredibility’ in Habbo, where as a pro can often be a Hobba moderator. A superhabbo is a celebrity of some sort. This classification is fairly simple, has only one way to proceed, and doesn’t include a differentiation between chatting, playing games, and trading furniture. However, it includes a potential explanation why some cheat and become cheaters: simply because nobody is interested in answering the simple question of how one obtains furniture, the novice may start stealing in lack of knowledge of other ways to get furniture. In the latest Habbo-magazine from Paratiiskala (no. 8), Toivo writes about ‘Kalan osastot’. He differentiates between four kinds of Habbo characters: furniture-traders, competitive players, VIPs, and chatters. Toivo suggests that furniture-traders are those who mainly communicate about trading, and chat with the objective to get as much different furniture as possible. He also suggests that a furniture-trader’s age is between 9 and 13. The competitive players create various quizzes for others in their guest rooms and throw dice in the casinos with furniture as stake. The VIPs create rooms with free access to friends only, others have to pay with furniture. VIPs collect rare furniture items, especially gramophones. Chatters are of age 15 to 40, don’t care about furniture or games, and like to spend time in the public rooms.

13

Summing these with previous fansite knowledge we identify the following eight user groups: 1) Furniture traders and collectors, 2) Chatters (in public rooms), 3) Gang-members and VIPs, 4) Hobba-supervisors, 5) Cheaters, 6) Quizz-makers and players, 7) The Hotel manager, and 8) Celebrities. This is a crude generalisation, since we continuously learn about more and less peculiar ways of using Habbo. For instance, one group of friends use Habbo a few times a year to gather there and catch up. Also not unheard of is parents leaving messages to and communicating with their children through Habbo.

Popular Activities Many are enticed of the fact that one can design the guest room to imitate anything. They enjoy designing to rooms to imitate for instance popular tv-shows like ‘Who wants to be a Millionaire’, or just arranging a quizz, or a bingo. Another popular event has been miss Habbo beauty contests. Some just do it for fun, others participate in or create games to get more furniture. Usually it costs some cheap piece of furniture to enter the competition, and the winner gets the lot, except for what the organiser keeps. Another popular competition is the furniture trap. People are invited into a room, where the organiser moves the furniture around so that less and less space is available. The game ends when the characters are trapped and cannot move anywhere in the room. The winner is the last one to be trapped. Paratiisikala reports in the magazine (no. 7) from early 2004 about one of the newest competitions, Habbo sumo-wrestling. The idea being to get the other character ‘mixed up’. Both players try to step into the same square at the same time. The one who stays in the square gets a point. The wrestling-ring is 4x4 squares big. A common game seems to be 1-on-1 and first to reach 10 points. Monitori.org reports about many different games in Habbo. For instance, they report that there are more than ten formula tracks in Habbo. The idea of a formula track is to run through a set of rooms connected by teleports, and the winner is the first to get the last room, from which the competition also started. Another game is called Habbo-Hockey, where people play in teams. Moving is restricted to two squares per turn, and shooting or passing is simulated by complicated sets of rules of how to roll the dice. Still another game is called CheatersGate (Peeloportti), which is similar to the formula tracks, but full of furniture obstacles on the way. The largest track owned by -Holy- is reported to be 25 rooms big. Rutzki from Aalto writes about different room types. Some types not yet mentioned include: - Casinos: where people gamble with their furniture. Everybody gets a dice and the winner is the one with the highest score on the dice. In many cases there is a VIP room, for those who give the owner a small piece of furniture.

14

- Dating: First the boys and girls line up into distinct queues, then a few boys and girls at a time ask each other questions in the date area of the room. They vote some participants out, ending up with a couple, who then can go into a quiet room to start dating - Talk shows: A host discusses with invited guests, often Habbo celebrities. - Clubs and hotels: by buying a ticket from the owner you get to the ‘better side of the fence’. - Orphanages: Some pretend to be orphans, others pretend to be parents. The room owner ‘connects’ the orphan with a suitable pair of parents, after which they can pretend to be a family. As we can see here, the diversity and creativity of the activities in Habbo is huge. Providing a comprehensive list of popular activities is challenging, but Rutzki’s summary is a good start: Trading furniture, Casinos, Dating, Beauty contests, Competitions, Dice games, Team sports, Formula tracks, Talk shows, Clubs & Hotels, Orphanages.

Conclusion & Discussion In this paper we have answered the research subquestion of what kind of content can be found on the fansites. We have provided detailed examples of different fansite elements and summary table based on 23 fansites. From a user research perspective we posed the second subquestion, what can we learn about the user groups and popular activities? Based on the fansite articles we were actually able to distinguish 8 different user groups and 11 major types of popular activities in the social world of Finnish Habbo users. These user groups and activities are only preliminary, because they have not yet been triangulated with other research data. The point here though, is that it is possible to get a grasp of the user groups and activities by analysing fansite articles. Therefore we can answer the main research question with a strong yes – fansites are useful sources for user research. The fansites around Habbo are important because they complement other sources for user research. Usually user research involves fieldwork methods like interviews, observation, focus groups, questionnaires, diary methods, etc. However, in the case of Habbo, we face a significant sampling problem with the large number of users: how do we know that the opinions of the interviewed users are representative? One way to find out that is of course to do a survey to get a large enough number of observations. However, in our exploratory user research, we realised that the fansites are very valuable. We first found the fansites as one of our interviewed Habbo users pointed them out to us as an important source for his Habbo knowledge. Soon enough we found ourselves reading the fansites and learning Habbo tips & tricks that would have taken months of play to find out. The fansites are not just any source of information, in many cases the content is written for a wide audience, representing a consensus. Also, they are public

15

discussions, which means that people put more thought and effort into them compared to private discussions. However, reading published information on fansites also brings on new challenges. Finding trustworthy information and interpreting it requires use of source critique. The texts are obviously not written for the same purpose as the researcher is reading them. The author and the fansite that published the article give important hints to how trustworthy it is. For instance, if an article is published on an official fansite, the author cannot be too critical towards the Habbo producer. On the other hand, if it is published on a more independent site, it can be critical, but the opinions might not be widely supported. In this way the fansites resemble ordinary mass media, some are more critical and some are more mainstream. Although the fansites are accessible without researcher intervention, the amount of articles is so large that they cannot all be included. The role of the user researcher becomes visible in finding and selecting the articles to report. Our study has concentrated deliberately on one country only, because the setting of one main site and the fansites around it enables us to concentrate on meaningful details compared to multinational and multilingual settings. It is difficult to assess whether our fansite sample is representative of the whole Finnish Habbo fansite landscape, because there is no comprehensive list of all fansites. However, our sample includes all official fansites and about one third of the active fansites, which makes it representative of the relatively small number of big and active fansites aimed at a large audience and being fairly close to the producer. And, correspondingly, our sample is not representative of the smaller and more underground fansites. Another question is how much the Habbo fansites represent fansites around other online games. Obviously, unlike in Habbo, if persistent texts can be found inside the games, the fansites probably look different. Then there is one need less for fansites. Also, in many online games, modifications (mods) play an important role, and sharing these mods are an important and potential activity for fandom. Habbo does not have mods, even if sharing Habbo pictures is an important communication theme. However, as around many online games, also Habbo furniture, especially rare ones, are traded on eBay. Discussing the economical aspects of the fansites would be an interesting theme. The most obvious future work is of course the international comparisons between the fansites around the national Habbo Hotels. Also important would be to analyse the different profiles of the fansites, focusing on how the fansites position themselves in relation to the other fansites. We also intend to reflect on the larger role of the fansites in the Habbo social world, since some of them seem to have strong possibilities to influence. From a user research viewpoint, comparing the results from the fansite analysis to questionnaire and interview results will reveal more details about when and where to use fansite analysis in information system design.

16

Acknowledgments This study is carried out in the context of the Mobile Content Communities project, funded by the National Technology Agency of Finland (TEKES). This research would not have been possible without the important people at Sulake Labs, thanks to Sampo Karjalainen and Rebecca Norman.

References Aarseth, E. (2003). “Playing Research: Methodological approaches to game analysis”, Proceedings of Melbourne DAC 2003. Activeworlds Corporation (2005). Company Information, http://www.activeworlds.com/info/index.asp April 4, 2005. Beyer, H. and Holtzblatt, K. (1998). Contextual design: Defining customer-centered systems. Morgan Kaufmann, San Francisco, CA, USA. Blythe, M. A., Overbeeke, K., Monk, A. F., and Wright, P. C. (eds.) (2003). Funology: from usability to enjoyment. Kluwer, Amsterdam, The Netherlands. Brown, B. and Bell M. (2004). “CSCW at play: ‘there’ as a collaborative virtual environment”, Proceedings of CSCW 2004, 350-359. Brown, S. A., Fuller, R. M., and Vician, C. (2004). “Who’s afraid of the virtual world? Anxiety and computer-mediated communication”, Journal of the Association for Information Systems, vol. 5, no. 2, 79-107, February 2004. Bøving, K. B. & Simonsen, J. (2004). “Http log analysis: an approach to studying the use of webbased information systems”, Scandinavian Journal of Information Systems, vol.16, 145–174. Cherny, L. (1999). Conversation and Community: Chat in a Virtual World. CSLI Publications, Stanford, CA, USA. Curtis, P. (1994). “Mudding: Social Phenomena in Text-Based Virtual Realities”. Intertek, vol. 3, no. 3, 26-34. Dibbel, J. (1993). “Rape in Cyberspace: How an Evil Clown, a Haitian Trickster Spirit, Two Wizards, and a Cast of Dozens Turned a Database into a Society”, Village Voice, vol. 38, no. 51, 36-42, 1993. Dourish, P. (1998) “The State of Play”, Computer Supported Cooperative Work Journal (JCSCW), vol. 7, no. 1-2. 1-7, January 1998. Drezner, D. W. & Farrell, H. (2004). “The Power and Politics of Blogs”. Presented at the 2004 American Political Science Association. August 2004. http://www.danieldrezner.com/research/blogpaperfinal.pdf April 4, 2005 Döring, N. (2002). “Personal Home Pages on the Web: A Review of Research”, Journal of Computer-Mediated Communication, vol. 7, no. 3, April 2002. Ermi, L., Heliö, S., Mäyrä, F. (2004). Pelien voima ja pelaamisen hallinta. Lapset ja nuoret pelikulttuurien toimijoina. Hypermedia Laboratory Net Series 6. University of Tampere. Forterra Systems (2004). Press Release, 20.4.2004. http://www.there.com/pr_webbyAwards2004.html April 4, 2005. Geissler, G., Zinkhan, G., Watson, R. T. (2001). “Web Home Page Complexity and Communication Effectiveness”, Journal of the Association for Information Systems, vol. 2, article 2, 1-47, April 2001.

17

Gurak, A., Antonijevic, S., Johnson, L., Ratliff, C., and Reyman, J. (2004) (eds.). “Into the Blogosphere - Rhetoric, Community, and Culture of Weblogs”. Online edited collection of scholarly articles. http://blog.lib.umn.edu/blogosphere/ April 4, 2005. Hackos, J. T. and Redish, J. C. (1998). User and Task Analysis for Interface Design. Wiley, New York, NY, USA. Henriksen, D.L. (2002). “Locating virtual field sites and a dispersed object of research”, Scandinavian Journal of Information Systems, vol. 14, no. 2, 31-45. ISO 9241-11 (1998). Ergonomic Requirements for Office Work with Visual Display Terminals (VDTs) - Guidance on Usability. International Standard, The International Organization for Standardization, 1998. Jenkins, H. (1992). “Strangers No More, We Sing: Filking and the Social Construction of the Science Fiction Fan Community”, in Lewis A. L. (ed.) The Adoring Audience, Routledge, London, UK, 208-233. Jones, Q. (1997), “Virtual-Communities, Virtual Settlements & Cyber-Archaeology: A Theoretical Outline”, Journal of Computer-Mediated Communication, vol. 3, no. 3. Järvinen, A., Heliö, S., and Mäyrä, F. (2002). Communication and Community in Digital Entertainment Services. Prestudy Research Report. Hypermedia Laboratory Net Series 2. University of Tampere. Konzack, L. (2002). “Computer Game Criticism: A Method for Computer Game Analysis”, in CGDC Conference Proceedings, Frans Mayra (ed.), Tampere University Press 2002, 89100. http://imv.au.dk/~konzack/tampere2002.pdf Kujala, S. (2002). “User Studies: A Practical Approach to User Involvement for Gathering User Needs and Requirements”. Doctor of Philosophy Dissertation, Department of Computer Science and Engineering, Helsinki University of Technology. Kuniavsky, M. (2003). Observing the User Experience: A Practitioner’s Guide to User Research. Morgan Kaufmann, San Francisco, CA, USA. Markus, M. L. and Mao, J-Y. (2004). “Participation in Development and Implementation – Updating An Old, Tired Concept for Today’s IS Contexts”, Journal of the Association for Information Systems, vol. 5, no. 11-12, 514-544, December 2004. Mnookin, J. L. (1996). “Virtual(ly) law: The emergence of law in LambdaMOO”, Journal of Computer-Mediated Communication, vol. 2, no. 1. Paccagnella, L. (1997). “Getting the Seats of Your Pants Dirty: Strategies for Ethnographic Research on Virtual Communities”, Journal of Computer-Mediated Communication, vol. 3, no. 1. Rheingold, H. (2002). Smart Mobs: The Next Social Revolution. Perseus Books Group, Cambridge, MA, USA. Sulake (2005). “Benchmark Capital invests in Sulake”, Sulake press release January 19, 2005. http://www.sulake.com/pressroom_releases_19012005_1.html April 4, 2005. Taylor, T. L. (2003). “Multiple Pleasures: Women and Online Gaming”, Convergence, vol. 9, no.1, 21-46, Spring 2003. The Sims Resource (2005). “Fansites index”. http://www.thesimsresource.com April 4, 2005. TNS Gallup (2004). “Monthly Site Rankings of Finnish Websites“, http://www.gallupweb.com/redmeasure/MonthReport/default.asp April 4, 2005. TNS Gallup (2005). “Weekly Site Rankings of Finnish Websites”, http://www.tns-gallup.fi/webtopList/ April 4, 2005. Woodcock, B. S. (2005). “An Analysis of MMOG Subscription Growth”, MMOGCHART.COM 15.0 February 2005. http://www.mmogchart.com/ April 4, 2005.

18

Appendix 1: List of Fansites Mentioned, December 2004. First in the list are the official fansites, after that the fansites are grouped based on how often they are updated. The "page count" column indicates the cumulative number of frontpage hits. It is only indicative, since the counters count from the day the counter was started or restarted, which might be different from when the fansite was launched. However, the order of magnitude is important, e.g. whether the fansite has one hundred, one thousand, ten thousands, or a hundred thousand hits. Also, the list is not exhaustive, it show only 23 selected fansites of 173. The inactive fansites were found based on fansite lists on older and inactive fansites. Many inactive fansites have been archived, e.g. check www.archive.org. Official fansites Kala-stamo Monitori Nefekala Nerokala Pixeli

www.kala-stamo.net www.monitori.org www.kotinet.com/nefekala www.nerokala.com www.pixeli.net

Total: 5

Active fansites E-kala Habbostaff Kriisipalvelu Lampi Lieskakala Munkki Navigaattori Palmu Sekofisu

koti.mbnet.fi/juhoppi/e koti.mbnet.fi/habbosta www.kriisipalvelu.net lampi.urli.net www.freewebs.com/lieskakala/frame.htm koti.mbnet.fi/munkkitk www.navigaattori.com www.palmu.biz kotisivu.dnainternet.net/paavokor/sekofisu

On a break Aalto H-Hotel Hyvitysturva

koti.mbnet.fi/rutzki/aalto koti.mbnet.fi/h-hotel/index www.hyvitysturva.fi

Page count 485862 N/A 315849 781308 ~300000

ID A B C D E

N/A N/A 2561 3421 N/A 27502 3403 21807 1448

F G H I J K L M N

N/A 9264 N/A

O P Q

1781 N/A 5595 135905 N/A N/A

R S T U V W

Total: 32

Total: 12

Inactive fansites (no updates for 3 months) Total: 59 Boxi koti.mbnet.fi/b0xi KultakalanKuvalehti kultakalankuvalehti.com Loisto nwps.ws/~kultakala/html ParatiisiKala koti.mbnet.fi/toiwo Rausku www.kuukso.com/~maztere/rausku/ Wannabe-kala koti.mbnet.fi/ggstudio/wanabikala

Table II. List of Fansites Mentioned, December 2004.

19

ARTICLE

II


Mikael Johnson and Kalle Toiskallio (2007). "Who are the Habbo Hotel Users—and What are they Doing There?" Nordic Consumer Policy Research Conference 2007. Helsinki, Finland, October 2007. Reprinted with permission.

187

Proceedings of the Nordic Consumer Policy Research Conference 2007

Who are the Habbo Hotel Users – and What are they Doing There? Mikael Johnson Helsinki Institute for Information Technology Helsinki University of Technology and University of Helsinki Helsinki, Finland [email protected] Kalle Toiskallio Intelligent Traffic Unit WSP Finland Ltd. Helsinki, Finland [email protected]

Abstract Who are the Habbo Hotel users, and what is so important in this virtual world that one quarter of the Finnish 10-15 years old population like to spend time there regularly? We present user categorisations based on empirical data from participant observations, a survey, user interviews, and fansite articles. Our results show a learning curve of about 3-4 months in average regarding Habbo behaviour and opinions, and we found some correlations between demographics (mainly age and gender) and online activities. The study found no single motivation for Habbo use, but instead points towards diversity, openended use, and creative content production. The users take part in the innovation process, since what active users do in Habbo becomes both a fundamental part of the use experience and a source of inspiration for the Habbo developers.

1. Introduction This is an account of our research driven by our curiosity towards the Habbo Hotel users, who they are and why Habbo fascinates. Habbo is many things: a software product for digital communities; a social networking site for children and young teenagers; a playful and non-violent virtual world where parents can trust their children to be reasonably safe, etc. However, Habbo is also an example of a change in technical innovation that has gained an increasing amount of attention during the past five or ten years. In many cases, technology users have not had much of an impact in technology production compared to other actors involved in the production of technology: technology sponsors, innovators, vendors, service providers, educators, architects, designers, tech support, etc. The users do not hold a very favourable position, since traditionally the innovation is seen to flow from the different developers via the other mediating stakeholders to the user (technology push), and not vice versa. The interests of the users are not the ones that are served first, because in most cases business and technology production come first, giving the interests of the users a very challenging position. (Kling 1977) On the other hand, examples of successful products emerging from close cooperation between producers and customers have led to an increased attention to user-driven innova1

Proceedings of the Nordic Consumer Policy Research Conference 2007 tion (Nordic Council of Ministers 2006, Hyysalo et al. 2007). User needs are in the center of such approaches as participatory design (Greenbaum & Kyng 1991, Schuler & Namioka 1993), user-centred design (Beyer & Holtzblatt 1998, Norman & Draper 1986), usability engineering (Nielsen 1993) and lead-user design (von Hippel 1988). In some cases, the users take matters in their own hands by modifying existing products, a few even forming own companies. In other cases, researchers and product development companies have developed user-centred methods ranging from early user involvement, making flexible products that support customisation (Laukkanen 2005, Nardi 1993), to representing and modeling users and their activities (Bødker 1998, Maguire 2001). However, also value-production has changed as computing technology has become cheaper and more pervasive. Community builders and researchers (Hagel & Armstrong 1997, Kim 2000, Preece 2000) argue that software for communities gain a great deal of value after the launch of the product/service/community. Effort is needed from the community moderators and value is created by the content-producing community members. This means that the traditional maintenance phase in the software lifecycle, when many considered the software project to be completed, has changed to something important, because now there are both significant costs and value involved. This shift in the design-use relationships raises many new questions. Producers need to know when and where to adjust their processes to meet the new demands on the production and collaboration with users. How can large amounts of simultaneous users be supported technically and socially, as well as represented in design? Revenue models need to take into account content produced by users, and the roles of consumers and users in innovation policies need to be attended to. One way of approaching this problem area is to study who the Habbo Hotel users are and what they do there. In the Mobile Content Communities research project that started in 2003 (Turpeinen & Kuikkaniemi 2007), the authors followed the development of the Habbo Hotel: the users and communities, new features, and the design-use collaboration. Sulake Corporation, the producer of Habbo, was an industry partner in the research project and based in Helsinki, which enabled research access to product development. Early in the project Habbo become an interesting case study for several reasons: • its innovative revenue model with no entrance fee, but rather micro-payments for online activities, which makes it possible for the users to try out the product first, and pay more if the product quality continues to meet expectations • the producers' success in making Habbo a reasonably safe place for its young users • its unique position on the virtual world market: retro, playful, and non-violent During the project, Habbo continued to fascinate the authors, because of its • active users taking part in the content production (Salovaara et al. 2005) • user-created Habbo-themed websites (Johnson & Toiskallio 2005) • developers' responsible role in governing different users' interests (Johnson 2007) • input to the discussions on the concept of community (Johnson & Toiskallio 2007) This article aims to open up the category of Habbo users – what different users are there and how do they differ? We studied demographics and online activities with both quantitative and qualitative methods. The next section gives a brief background to Habbo and the starting points for the study: multiplayer games studies and user research.

2

Proceedings of the Nordic Consumer Policy Research Conference 2007 2. Background Habbo has its origins in the experiences from multimedia cdroms and two Internet chat rooms, Mobiles Disco and Lumisota. In terms of Internet technologies, Habbo is a graphical chat environment on the Web accessible with a web browser with the Shockwave plug-in. This chat environment is designed as a virtual hotel where people can hang out and make new friends. When checking in to the virtual hotel one creates one’s own cartoon-like Habbo avatar that can walk, dance, eat, drink and chat in the cafés, restaurants, swimming pools and games rooms. Besides experiencing these common rooms in the hotel, one can decorate and furnish a room of one’s own. In contrast to many online games, there is no entrance fee to the virtual world, which allows the majority of users to chat for free. The profit model is based on micropayments in the hotel instead. Virtual furniture, minigames, and membership in the Habbo club are bought with so called Habbo credits. These credits can be purchased with pre-paid cards, bank transactions, or special text messages that add a specified amount of money to the customer’s mobile phone bill.

Figure 1. Screenshot from Habbo Hotel (Sulake 2006)

The research approach in this report is based in user research and games studies, research fields that aim at understanding users and players. User research (Hackos & Redish 1998, Beyer & Holtzblatt 1998, Kuniavsky 2003) is based on the idea that knowing more about the users will help designers create better products that meet the user needs. User research typically includes identifying user groups, understanding their context of use (ISO 924111 1998) by visiting user sites and making interviews and observations of the users in their own environment. However, the underlying assumption of many user research models is a product development model where the user needs are fairly well stabilised. Design for digital communities on the other hand does not assume an existing target market; especially online entertainment products strive to facilitate the formation of new communities around the product. This, and because entertainment products are used in a leisure context and usability is not concerned with gameplay, is why it is difficult to apply mainstream usability and user research models to understanding Habbo users. Games studies have researched computer games for more than a decade from technical, aesthetic and socio-cultural perspectives by a growing number of researchers with different backgrounds (Aarseth 2003). Closest to Habbo is the research on MUDs (textual chat environments) (Cherny 1999) and virtual worlds (Bartle 2003, Book 2004, Brown & Bell 2004, Mulligan & Patrovsky 2003, Yee 2002). Most attention has been paid to the gameplay itself, the question of what constitutes a game, and structural game elements (Kon3

Proceedings of the Nordic Consumer Policy Research Conference 2007 zack 2002). There is some research on player models (see below) and also other motivations to participate. For instance, the “Children and Young People as Players of Game Cultures” -project (Ermi et al. 2004) report that 75 % of the 10-12 y. children (N=284) play digital games at least once a week, that owning games is a good argument for inviting friends, and that games aid in learning English. Playing online games is not only a question of the game itself, but rather a mix of online and offline matters. The question why do people play massive multiplayer games has been asked many times in games studies. Table 1 presents four player models that have acted as inspiration for our Habbo research. Bartle’s (2003) infamous socializers-achievers-explorers-killers model has been applied in many cases, but since Habbo is a non-violent game environment1, the killer category does not fit2. Following the hermeneutic rule that if the parts change, then the whole needs to be revisited, we engaged in looking for other models. Table 1. Player models for virtual worlds (from Bartle 2003). Bartle 1997

Farmer 1992

Hedron 1998

Yee 2002

MUD

Habitat (graphical chat)

Ultima Online

Everquest

Socializers Achievers Explorers Killers

Passives Actives Motivators Caretakers Geek Gods

Survival Competence Excel Prove Mastery Seek New Challenges Everything Is One

Achievement Grief Leadership Relationship Immersion

Farmer’s model of passives, actives, motivators, caretakers, and geek gods from Habitat seemed to fit Habbo fairly well. It made sense to understand motivators as those few who create games for the others in Habbo (playmakers in Salovaara et al. 2005), caretakers as the Habbo moderators, and geek gods as the Sulake game developers. However, Farmer’s model lacked a lot of context and demographics to satisfy us, which motivated us to move forward with the aim of finding suitable player categories for Habbo.

3. Methods and Research data When talking about users we made the following initial distinction: demographics and Habbo activity. This distinction was elaborated on during our survey work (Table 3), but it enabled us to break down the initial who-is-the-user question in to three different questions: 1) what is the demographics of the Habbo users? 2) What are the popular activities in Habbo, and 3) Does the demographics correlate with the Habbo activity? Table 2 summarised the empiric data in the Habbo case. Here we focus on the survey data, fansite articles, and player interviews. For data analysis, we have used both quantitative methods (frequencies, cross-tabulation, cluster analysis) and qualitative methods (qualitative content analysis and membership categorisation analysis). 1 Because of the non-violent characteristics of Habbo compared to Ultima Online and Everquest, we have not discussed Hedron’s and Yee’s models in relation to Habbo. 2 If we stretch killing towards grief play (disturbing other players), one could potentially apply it to Habbo as well, but since Habbo is a moderated environment and grief play is not acceptable, that still leaves the killer category in an illegitimate position compared to the other motivations.

4

Proceedings of the Nordic Consumer Policy Research Conference 2007 Table 2. Empiric data in the Habbo case. Main empiric data

Background data

3

survey : Habbo visitor profile, 06/2004 • N = 10000, 21 questions with structured answer alternatives, 6 open questions

First-hand Habbo experiences / participant observation • once or twice per month sporadic visits to Habbo Hotel

fansite (Habbo-themed website) articles • Habbo’s own amateur “mediaworld”, 2004, selected articles from ~25 fansites with 5-50 articles each, 150 MB downloaded material

project meetings • at Sulake once every 2-3 months • 5 recorded meetings, ~10 informal

developer interviews, 04-05/2005 • 10x 1,5-3h open-ended thematic interviews, some development documents • client developers, server developers, graphic designers, ADs, project managers

Confidential work with Sulake • community manager survey, N=4, 2003 • user feedback for release 9, 2006 2 TKK student assignments • data gathered by students • usability test, fansite starter kit

player interviews, 10/2005 • 12 players interviewed in 2 individual, 2 pair, and 1 group interviews

Informal second-hand experiences • friends, colleagues in Habbo • professional media articles

artefact analysis of Habbo Hotel • stereotypic images, affordances

4. Results The explorative survey (see subsections 4.1 and 4.2) suggested that we needed to go deeper into what the popular activities in Habbo were, to pinpoint the different motivations to participate. Our fansite studies (4.3) and player interviews revealed the diversity in the emergent Habbo activities. We describe the emergent use in section 4.5, but before them we also describe the pre-defined Habbo categories made by the developers (4.4).

4.1. Statistics: Habbo Visitor Profile Because there were no data available on the Habbo visitors, we decided fairly early in the project to do a visitor profile survey, to get some quantitative background data for our otherwise mainly qualitative analysis. We were interested in both demographics and online Habbo activites, so we created a model of how we understood the Habbo visitor (Table 3). Our objectives was to find out which of these attributes were related to each other, for instance if age or gender says anything about what the visitors do in Habbo. The survey was carried out as a web-based survey and a link to the web form4 was put on the Habbo frontpage, under the Habbo News section. We asked 27 questions, of which 21 had structured answer alternatives and the last 6 were open. The survey was open for two weeks, during 22.6.2004-6.7.2004, and we got roughly ten thousand Habbo responses (N=10 613). During this time around 160 000 users visited Habbo, which gives us a re3 4

Questions and answer frequencies available at http://mc2.soberit.hut.fi We used a survey research service provided by Digium Research, see http://www.digium.fi

5

Proceedings of the Nordic Consumer Policy Research Conference 2007 sponse rate of about 6%. We discarded about 300 answers, of which some were intended as jokes, but typically it was empty or double answers, since the survey service did not restrict multiple answers from the same computer. Table 3. Model of the Habbo Visitor Traditional background Age, gender, region

Other activities

Online Habbo activities

School, hobbies, other games, friends

Chatting, furniture, decorating rooms, creating games, making friends

Habbo fansites

Habbo background

Favourite, visit frequency, reading & writing, forum discussions

Age in Habbo, visit frequency, network connection, logon place

We did three major kinds of analyses: a) frequencies, b) cross-tabulation, and c) cluster analysis. We start with an overview of the frequencies (4.1.1), continue with findings from the cross-tabulation (4.1.2), and end with the visitor clusters (4.1.3). 4.1.1. Overview Table 4 shows an overview of the answer frequencies to some background questions. Table 4. Overview of survey answer frequencies Variable

Description

age

75% of the respondents were between 10-14 year old.

gender

Roughly 50-50, equal amounts of boys and girls.

region

All over Finland, not only big towns or countryside.

access location

93% log on from home

Habbo age

years in Habbo: % of respondents:

0-1 29%

1-2 26%

Visit frequency

everyday 46%, a few times a week: 40%

Visit freq. fansites

more than half visit at least once a week

2-3 27%

3+ 17%

4.2. Findings Sulake has reported5 that 90% of the Habbo visitors don’t pay. However, this doesn’t mean that these non-paying visitors are not interested in Habbo furniture. Based on preliminary interviews and fansite stories we had got the impression that it is possible to have nice rooms and make careers in Habbo without spending money, for instance by trading or receiving gifts. Our survey showed that a large part of the visitors get furniture through donations, competitions, or as rewards for favours. Of the 26%6 (2557) who said that they don’t buy anything, 57% (1451) still receives furniture donations, 48% (1229) trades furniture, and 31% (782) gets furniture items as pay for “work” in Habbo. Thus, we 5

Press release 25.10.2004. http://www.sulake.com/press/archive Disclaimer: The percentages should be read as percentages of the answers to the survey, which is biased towards active Habbo visitors. The share of all users who don’t buy anything is much larger than 26%. 6

6

Proceedings of the Nordic Consumer Policy Research Conference 2007 can draw the conclusion that from “I don’t spend money in Habbo” does not follow “I’m not interested in furni”. In other words, also people who don’t spend their own money in Habbo contribute to Habbo’s virtual economy through transactions and value creation. Before we did our study, we had good reasons to believe that two factors very much influence Habbo behaviour and attitudes: habbo-age and visit frequency. Our data showed that this is indeed the case. Habbo-age has a strong influence on Habbo attitudes. Figure 2 shows a few trends. We can distinguish different types of curves: a) a rising trend that stabilises at 4-6 months (green, yellow, blue), b) a decreasing trend that stabilises at 4-6 months (red, lila), and c) a decreasing trend all the way (orange, brown).

Figure 2. The importance of Habbo-age.

From these curves we can draw two conclusions. First, since it is possible to distinguish trends at all means that Habbo-age is a really important variable. Second, the data suggests that there is a learning curve of about 3-4 months, after which opinions stabilise. In addition, the slight change in the later parts of the curve can be interpreted as a decrease of participation in Habbo, foreshadowing the end of their Habbo careers.

Figure 3. The importance of visit frequency.

The other variable that clearly influences the Habbo career is the visit frequency (Figure 3). The data suggests that those who visit Habbo often are more likely (decreasing trends) to arrange events, trade furniture, meet with Habbo-friends outside Habbo. And, those who visit Habbo rarely are more likely (rising trends) not to want people in their rooms, not have heard about gangs nor fansites, not like spending time in their own rooms. The two above figures show that common sense and research data agrees. More time spent in Habbo is related to being active (arrange events, trade furniture, etc.) there. The learning curve was very interesting to note. But is there any underlying variable that can 7

Proceedings of the Nordic Consumer Policy Research Conference 2007 predict time spent in Habbo? Can we say that people of a certain gender, age, or from a certain region are more likely to be active in Habbo? We will go through these questions (section 4.2.2), but before that we will turn to other patterns in our data. A cluster analysis will show the relevant patterns in the data, which factors make a difference.

4.2.1. Cluster Analysis The aim of the cluster analysis was to create a manageable number of player clusters to explain the data. (Figure 4). Since this was an explorative questionnaire, the emerging cluster dimensions were considered more important than the exact percentages.

Figure 4. Habbo user groups based on cluster analysis

8

Proceedings of the Nordic Consumer Policy Research Conference 2007

Three criteria were important in the analysis: a) a strive for groups that one can explain, b) no logical conflicts within groups, e.g. non-buyers and buyers in different groups, c) as large part of the data as possible included, e.g. anything that explains more than 50% is good. The SPSS-statistic program provided the functionality: Two-Step Cluster. We ran around 30 different cluster analyses to find out what would be a sensible number of groups to meet the above criteria. The first sensible cluster result is shown in Figure 4, and the significant variables that explain the particular cluster are marked with two stars (**). We ended up with 6 groups that we named Oldtimers, Playmakers, Silent majority, Gang-members, I don’t pay, and Older people. These groups explained 72% of our data. However, these clusters were based on all variables, which meant that the background variables (gender, age, etc.) took over. For instance, when we tried to create 7 clusters instead of 6, the next emerging group would have been an all-girls group. To remedy this issue, we decided to analyse background variables separate from habbo activity variables. An analysis suggested that two dimensions were more distinguishing than others: 1) privacy – publicity, and 2) arranging events – not spend time in own room. Figure 5 shows our previous clusters put on these dimensions. In the next subsection we analyse in more detail two of the background variables usually considered important, age and gender.

Figure 5. The two most important ingame dimensions.

4.2.2. A Brief Note7 on Age and Gender in Habbo “...girls and boys have different things, boys collect all those furni there, and girls chat” (Pair interview 10.6.2004) Early in the project, in the first user interviews, we encountered very categorical statements like the one above about boys and girls in Habbo. Because no use statistics was 7

A gender perspective can be much more than this “within categories” analysis. For a study on gendered self-presentation in Habbo, see Pietiläinen (2004). For a useful contextualisation of gender and online communities from a constructivist perspective, including discussions of hegemonic masculinity and normative feminity, see Sveningsson Elm (2007).

9

Proceedings of the Nordic Consumer Policy Research Conference 2007 available to confirm or contrast these gender stereotypes, we decided to put them to a test in our survey. Table 5 below shows a cross-tabulation of the topics in the survey and gender. It is based on the question “I am”, with the answer alternatives “a boy”, “a girl”, “a man”, “a woman”. The age for when someone becomes a man or a woman was not predefined, in contrast, the respondents could define themselves as boy, girl, man, or woman. The survey showed fairly similar distributions in getting furniture. One difference is that 74% of the boys compared to 65% of the girls report that they get furniture by trading. Another difference is that a slightly larger share of the girls receive furniture as gifts (54% vs. 47%). Based on this survey, rather than finding furniture trading something that mostly boys do, we find that a majority of both boys and girls trade furniture in Habbo. Table 5. Gender similarities and differences. Boys

Girls

Men

Women

Number of respondents (N) Percentage of all respondents

4672 46%

4744 46%

497 5%

317 3%

Age (median) Age (ave)

13 12.8

12 12.6

15 16.7

15 17.2

Computer in my room

42%

24%

55%

32%

No login restrictions

41%

32%

60%

53%

Does not know network connection

18%

41%

11%

27%

Visit frequency - everyday - a few times / week

47% 39%

45% 42%

47% 37%

42% 36%

Habbo-age (years) - 0-1 - 1-2 - 2-3 - 3+

31% 26% 26% 17%

31% 28% 29% 12%

25% 17% 27% 31%

33% 23% 25% 20%

How do you get furniture? (multichoice) - I buy them - I receive them as gifts - by trading - by working (as salary) - by cheating - I’m not interested

66% 47% 74% 30% 4% 10%

64% 54% 65% 27% 1% 15%

61% 44% 62% 25% 8% 18%

63% 51% 58% 26% 2% 21%

Own room activity (multichoice) - organise events - invite friends - don’t feel comfortable

39% 61% 44%

50% 62% 45%

35% 52% 51%

35% 52% 51%

Wants with room? (multichoice) - furniture - friends - fame - fansite fame - something else

33% 65% 48% 14% 13%

29% 68% 26% 6% 22%

28% 58% 46% 19% 21%

29% 56% 24% 5% 30%

Spends time with regular gang

38%

26%

38%

34%

Never needed to call a moderator

27%

39%

29%

44%

Visits fansites at least once a week

50%

34%

47%

27%

10

Proceedings of the Nordic Consumer Policy Research Conference 2007 The similarities in Habbo-age and visit frequency is striking. Among the few differences we note that, like in many previous gender studies (Nurmela 2002), boys tend to have easier access to services (Habbo) compared to the girls. Also familiar from research on gender stereotypes (Svahn 1999), more boys than girls report that they strive for fame. Despite this, a larger share of the girls (50%) reported that they arrange events to get others to visit their room (boys 35%). For some reason the Habbo-themed websites created by users, so called fansites, seem more popular among boys than girls. While leaving a detailed exploration of age for future work, we note that it seems like the men and women tend to be less social compared with the average. This is understandable as they are a marginal group in Habbo.

4.2.3. Bias towards active players As with surveys in general, also this survey gathered answers from the committed and active community members. This is visible in the table below. Table 6. Checking Habbo-age and visit frequency is a way of determining bias. 0-1y

1-2y

2-3y

3+y

Every day

15%

11%

12%

8%

A few times a week

11%

12%

11%

5%

Once a week

1%

1%

1%

1%

A few times a month

1%

2%

2%

1%

Less than once a month

0%

1%

1%

1%

4.2.4. Open questions after the survey Although we were satisfied with the survey and what we learned from doing it, the survey still left some questions open. First, we targeted the survey to answer the question “who are the Habbo users”, not what motivates them to go there. Second, the survey was successful in describing a snapshot of the Habbo community, but community research (Kim 2000) tell us that the motivations for community members to participate change during their career in the community. To answer these questions we studied the Habbo fansites and interviewed Habbo visitors. We will come back to these questions in the conclusion. Third, when doing the explorative survey we did not fully understand the diversity of all the Habbo activities going on. During our analysis we learned that a Habbo “event”, as we called it in the survey, can be broken down in to many dimensions. For instance, the event (or series of events, making it more like an activity) can be described in terms of large or small, long-term or temporary, visible or underground, open or closed. Furthermore, the group activity leading up to the event can be hierarchic or democratic, valuing uniqueness or competition, and the purpose of buying furniture and decorating the room varies from just for fun to becoming rich, to imitating real world games. The point being, in the survey, where people answered that they would create an event to get people to visit their rooms, we lacked data on what kind of event they were talking about. Our fansite studies made us wiser in the above respects, as the next section shows.

11

Proceedings of the Nordic Consumer Policy Research Conference 2007 4.3. Learning from the Habbo Fansites We studied 173 Finnish Habbo fansites in 2004. The focus was on what could be learned about the Habbo visitors and their Habbo practices from user research (Hackos & Redish, 1998; Kuniavsky, 2005), focusing on the membership categories visible through the fansites. Since the fansites are accessible without research intervention, the risk of distorting the data by the presence of the researchers is reduced. The most popular fansites are usually made by a small team of Habbo fans with different expertise and roles. For example, one designs the look and layout of the site, another one writes the stories, and a third has the technical skill to publish the site on the Web. The most active fansites have small updates (such as news and rumors about Habbo) several times per week, but publish reviews and articles once a week or bi-weekly. Some fansites group their articles together and publish them as an issue of a Web magazine. The Web magazines seem to follow a rhythm of one issue per one or two months. The fansite contents were classified and a list of common fansite elements was produced (Table 4). Table 4. Common Habbo Fansite Elements (based on Johnson & Toiskallio, 2005). Fansite Elements Description News and rumors

Fansites are convenient for Habbo visitors who want to reach a large audience, a fast way of spreading information about Habbo happenings (e.g., competitions, pop idols visiting Habbo), new features, news about Sulake Corp.

Participation

The fansite audience is provided ways to comment on the fansite through discussion forums, guest books, polls, etc..

Links

The fansites link to relevant Habbo places: other fansites, and to the hotels in other countries.

Hints, secrets, guidelines

Fansites teach newcomers both basic and advanced tricks with which to impress others. Guidelines on acceptable behavior are frequent.

Reviews and lists

The fansites keep track of the features and possibilities in Habbo: public spaces, different furniture items, pets, etc.

Histories

Two major histories are told on the fansites: the history of Habbo and the history of that particular fansite

Fashion and celebrities

Habbo “journalists” interview Habbo celebrities, avatars who have become famous in Habbo, and report on fashionable clothing and activities.

Graphics

Edited screenshot pictures are an integral part of many fansites, some even provide pixel graphics drawing schools.

Habbo fiction

A few fansites write fictional stories about characters in Habbo.

About

Who comprises the fansites staff, number of visitors, updates, banners, etc.

Real life

Habbo meetings “in real life,” stuff not about Habbo that is important to teenagers, as well as blogs, e-cards, etc.

Based on the fansites, we clustered the hotel residents into eight groups: a) furniture traders and collectors, b) chatters (in public rooms), c) gang-members and VIPs (insider groups not open to everyone), d) supervisors with administration powers, e) cheaters, f) quiz-makers and players, g) the hotel manager (a Sulake employee), and h) celebrities. Similarly, 11 popular activities were identified: trading furniture, casinos, dating, beauty contests, competitions, dice games, team sports, formula tracks, talk shows, clubs & ho12

Proceedings of the Nordic Consumer Policy Research Conference 2007 tels, and orphanages. More important than the exact details of these listings are two observations about Habbo that they convey: the diverse and commonplace qualities of Habbo. First, there is not one particular Habbo activity that attracts all Habbo visitors, but many different ones. Second, the activities going on in Habbo resemble games with rules and pretend play familiar from schoolyards, playgrounds, youth clubs, and so on. The Habbo visitors and their practices seem to be strongly influenced by the fansites. They complement Sulake’s official Web site by providing more detailed information about the hotel from an experienced visitor’s point of view. Hints, secrets, and guidelines, and stories about Habbo fashion influence the boundaries for acceptable behavior in Habbo. The fansites improve the Habbo visitors’ awareness of the fan cultures around Habbo, and also reproduce and reinforce social positions (like potential Habbo career paths or legitimized visitor groups).

4.4. Pre-defined Habbo Categories The Habbo aesthetics de-emphasise bodily differences between visitors. All avatars look like they are of the same length, height, and age. The visitors can choose their own clothing, hair styles, and skin colours from a pre-defined set (Figure 6).

Figure 6. Self-presentation possibilities for Habbo avatars. Left: the avatar of one of the authors. Right: an avatar from Pietiläinen (2004).

There are two avatar categories visible in the user interface in Figure 6, girl and boy. Unlike many other MMOGs, the Habbo designers chose not to use categories such as race, nationality, region, or alignment (good, evil, etc.). At the time of the study, there was a dozen special Habbo badges that show the special status of certain Habbos. Some of these the visitors can buy, some of these are earnable, and some of them signify volunteers or other workers. Table 10 shows a summary of these badges.

4.5. Emergent Habbo Categories By combining elements from the pre-defined sets, the Habbo visitors have chosen their own clothing styles. Table 7 shows examples of Habbo clothing styles. 13

Proceedings of the Nordic Consumer Policy Research Conference 2007 Table 7. Some examples of common clothing styles (Kriisipalvelu.net 20058) …

Punk

Gothic

Teinix (teenie)

Wannabe

Own style

…

…

Bright colours, irokese

White skin, dark clothes

Pastel colours, round cheeks

Strict clothing, smart behaviour

Doesn’t care about fashion

…

However, Habbo visitors are not only clustered based on their looks, it is also about their textual description, the room decoration theme, their style of chatting, and behaviour patterns. A common way of talking about Habbo avatars is through a their profession. Some Habbo visitors pretend to have a profession in Habbo, which involves decorating their room and behaving according to the selected profession. Some fansites have written articles on this role-playing phenomenon and discussed the following professions: journalist, nurse, tv-show host, bartender, pharmacist, actor, police, doctor, fireman, postman, veterinarian (Nerokala 2005, DJ Joneppe 20029).

Figure 7. Two examples of visitors imitating TV-show formats in Habbo: Habbo Idols (left) and The Bachelor (right).

Imitating TV-show formats is another popular way of creating a hotel guest room theme. Figure 7 shows two examples, Habbo Idols (a room visited during a user interview 19.10.2005) and Habbo Bachelor (Unelmien Poikamies in Finnish) reported on the fansite Kriisipalvelu.net (May 2006). We have also found mentions of other tv-shows such as Greed, Do you want to be a millionaire, Big Brother, Survivor, America’s Next Top Model, and different Dating-formats. 8 9

Kriisipalvelu at http://habbomatic.com/ Nerokala at http://nerokala.com/ and DJ Joneppe at http://www.kolumbus.fi/djjoneppe/index.htm

14

Proceedings of the Nordic Consumer Policy Research Conference 2007 The diversity of the popular Habbo activities is very large, however one way to compare activities is to look at the their characteristics over time. For instance, an activity can be open to everyone, or only for selected visitors (closed). It can be fleeting, or continue over a longer time period. The activity can be arranged in a hierarchical way or be more democratic, and it might be very visible or more underground. Below (Figure 8) are three examples highlighting these dimensions.

Figure 8. Three different kinds of social activities: orphanage (left), bingo (middle), gang (right).

To the left is an orphanage, where people come to be connected as child and parents, and continue to roleplay as a family. The room owner stands in the image to the right, the child is in the middle (yellow shirt), the child queue in the bottom left, and the wannabe parents in the upper left. The middle image shows bingo, using a set of interactive dice furni: the person who rolls the same dice number as the room owner wins. To the right is a gang’s recruitment room. Table 8 shows some differences in these group activities. Table 8. Some characteristics of group activities Orphanage

Bingo

Gang

Open-closed

Open

Open

Closed

Fleeting-durable

Fleeting

Fleeting

Durable

Hierarchical – democratic

Game master + democratic participants

Game master + democratic participants

Hierarchy with participant ranks

Visible – underground

Visible

Visible

Underground

The virtual world also has other properties that have inspired games. Furniture traps, telerunning, and sumo-wrestling is can be mentioned as examples. Figure 9 shows a furni trap game that plays with the spatiality of the hotel. The assistant to the room owner runs the game by moving six tables in a row-like way, one table one square forward at a time, thereby reducing the available floor space for the players. The number of players is reduced as they are trapped. In some games, the trapped players can buy themselves out of the trap to continue the game.

Figure 9. Furni trap game: 1) starting position (left), 2) decreasing available floorspace (middle), 3) one player is trapped (right).

15

Proceedings of the Nordic Consumer Policy Research Conference 2007 Yet another way of clustering Habbo activities is to look at the role of the furniture in the activity. In the previous examples furniture has played a secondary role in the room. It is also possible to consider furni as ends in themselves. For instance, furniture museums have appeared in Habbo: when a room is filled with all different furni items, which the owner has spent much time to collect. Or, when the room owners just want to decorate the room with the aim of making something beautiful. Figure 10 below shows three examples of Habbo furni laid out in a way that should make the room owner very cool.

Figure 10. Furniture that makes you look cool. TV-wall (upper left), lots of writing desks (lower left), icehockey theme (right).

The TV-wall is cool, because it’s an original idea in Habbo, since to place a tv-set above another, once has to use a special bug/feature that is not known to everyone. This shows technical competence, patience, and a sense of aesthetics. The room filled with writing desks is cool, since Habbo visitors know that one has to be member of Habbo Club for ten months to get just one desk of this type. This fortune shows that the room owner is a successful trader. The ice-hockey theme on the right is cool because of its thematic decoration, and because the owner show off with several furniture items that have only been for sale during two week periods in the hotel’s history. Table 9 summarises the role of furniture items (virtual assets) in Habbo activities. Furniture can be an end in itself, it can be a means for something else, or play no role at all. Table 9. The role of furniture in Habbo activities Role of furniture

Description

Furniture as an end in itself

to collect all different furni items furni as aesthetically pleasing

Furniture as a means for something else

to make friends to make others happy to become famous to become rich

to show status as props for role playing as props for chatting

Indifferent to furniture

socialisers / chatters that keep contact with friends made in and outside Habbo

hackers (against the system) grief players

Based on our player interviews, we have also looked at how the Habbo visitors talk about the other Habbo visitors. In addition to describing Habbo visitors based on what they do

16

Proceedings of the Nordic Consumer Policy Research Conference 2007 in Habbo (activities), Table 10 summarises the membership categories that were used during our interviews to explain to us what other kinds of Habbo visitors there can be. Table 10. Predefined and Emergent Visual and Nonvisual Membership Categories. Predefined visual categories avatar appearance: boy / girl purchasable badges: Habbo Club, Golden Habbo Club, Halloween smile special badges: Habbo staff, NGO workers, youth workers, mental support, VIP guests earnable badges: Habbo X (guides), fansite authors, Battle Ball gurus

Emergent visual categories

Non-visual categories

clothing styles: punk, gothic, teenie, wannabe (strictly dressed), personal style (independent of fashion) professions: journalist, nurse, TV show host, bartender / waitress, pharmacist, actor, police, doctor, nurse, fireman, postman, veterinarian, etc. TV show formats: Idols, the Bachelor(ette), Greed, Do You Want to be a Millionaire, Big Brother, Survivor, America’s Next Top Model, “Dating,” etc. categories formed from value judgments on others’ appearances: good looking, bad looking

visitors from other hotels (nationality / language region) speaking another language age: small children, “my age” (teen), older gender: combinations of nickname, avatar, real body (e.g., boy with girl avatar but masculine nickname) Habbo age: newbie, regular, guru relation to room: room owner, visitor, shared rights friends made in or outside of Habbo: Habbo friends, real friends trading furniture: little by little, skilled traders, cheaters time of day (is associated with distinct visitors): daytime (children with flu at home, mothers), after school (preteens), evening-night (older, best discussions)

5. Conclusion During our Habbo visitor interviews, we tried to find out whether the visitors would identify themselves with one particular visitor category or Habbo activity as presented above. None of the interviewees really accepted to be labelled as one particular category, as they all mentioned that they started out doing certain things in Habbo and then moved on to try other things. They claimed that they got bored of doing the playing the same game or keeping the same room activity after a few weeks. This brings us back to the original question, who are the Habbo users? During the project, we sought to generate some user groups or categories to explain that the people who visit Habbo can be divided into these and these categories. We started out by exploring Habbo for ourselves and making pilot interviews, ending up with a few preliminary categories: furni collectors, chatters, late-evening party people, mafioso, sheriffs (volunteer moderators), and cheaters. The survey gave us more background data to work with as it confirmed our assumption that time spent in Habbo really correlates with knowledge of and activity in Habbo. A cluster analysis of the data suggested six user clusters that we named Oldtimers, Playmakers, Silent majority, Gang-members, I don’t pay, and Older people. We also found two dimensions of Habbo activities that correlated more strongly than the others: a strive for publicity vs. privacy, and arranging events vs. not spending time in own room. The fansite studies and our interviews opened up the emergent types of use.

17

Proceedings of the Nordic Consumer Policy Research Conference 2007 We have shown examples of Habbo clothing styles, Habbo professions, TV-show formats, competitions, and other fun ways of being together in Habbo. We briefly analysed some characteristics of and the role of furniture in the Habbo activities. We also provided a listing of the categories our interviewed visitors used when describing Habbo players. When going through all these different ways of describing the Habbo users, the diversity of the Habbo uses is striking. Answering the question of who the Habbo users are with one particular categorisation does not seem fair. Neither is a simple two-dimensional model enough. What we can do, however, is to provide a summarising list of aspects that are important in Habbo. It is possible to distinguish different user categories based on all of the dimensions in the list. Depending on the motives behind the question, different categorisations can be made. The list can also function as a map for discussing a Habbo career with a Habbo user: which aspects drew the user into Habbo, which aspects made the user come back, and so on. Table 11. Summary of important aspects of Habbo Aspect

Description

one’s own avatar(s)

clothing styles, character description

one’s own rooms and furniture

collecting, trading, decorating, browsing the furniture catalogue

Habbo homepage

10

your avatar’s homepage that is visible to anyone on the web.

friends

school, hobbies, new friends, dating, distant friends

play

beauty contests (popularity), TV shows, games of chance, Habbo-sports, playing with the spatiality of the virtual world

Habbo career

celebrities, getting rich, popular room, in a game or gang, being a fansite author, being a Habbo guide

testing boundaries and rules

expressing self, treating others (e.g. cheating, bullying), finding and using glitches in the hotel architecture

While this study answers the empirical question of what users do in Habbo, the study also highlights a general aspect of many community software products: the users’ important roles in the innovation process. The users take part in the content production, some as active playmakers, and others as equally important participants and audience. What the users do in Habbo becomes both a fundamental part of the use experience and a source of inspiration to the Habbo developers. Our work shows that there are no self-evident ways to group all the users for all situations. For instance, it is not clear how stabile the user clusters based on the survey are over time. On the other hand, Table 11 supposedly presents the Habbo dimensions that won’t disappear over time (assuming the technology doesn’t change drastically). There are several reasons for why different organisational stakeholders need categorisations. Marketing for instance, needs some knowledge of the demographics of the Habbo users, but since demographics do not necessary correlate with action in the game, it is hard to design for demographics. Therefore categorisations of player activities are more useful for development. However, categorical knowledge is not the only type of knowledge useful in design. Previous research has shown that anecdotal knowledge (a.k.a. stories and narratives) plays an important role in design. Long-term interaction with the users 10

New feature of February 2007

18

Proceedings of the Nordic Consumer Policy Research Conference 2007 through various user feedback channels provide the developers a broad resource of anecdotes of what and how the users innovate with the Habbo furniture. So, the question is not how many personas is enough to represent the users of Habbo, since in the Habbo case there are two particular features that makes developer-user dialogue easier compared to many other product development contexts. First, the developers have easy access to what the users do with and in Habbo, anyone can log on and check it out. Second, the online discussions in the user-created fansite forums provide the developers with loads of user feedback and insights into what the users expect. Then it is possible both to meet the expectations as well as make surprises. Future work involves reflections on what kinds of user categories one can make. In this report, we have not taken a strong position in advance of what we were categorising but tried to be open towards different things: individuals vs. groups, ingame (rooms, activities) vs. demographics (age, gender, region, etc.). The empirical data could be analysed more to distinguish user careers and lifecycles.

References Aarseth, E. (2003). Playing Research: Methodological approaches to game analysis, Conference Paper for the Melbourne DAC 2003. Bartle, R. (1997). Hearts, Clubs, Diamonds, Spades: Players Who Suit MUDs. Journal of Virtual Environments, Vol 1, Issue 1, July 1997. Bartle, R. A. (2003). Designing Virtual Worlds. Indianapolis, Indiana: New Riders. Beyer, H., & Holtzblatt K. (1998). Contextual design: Defining customer-centered systems. San Francisco: Morgan Kaufmann. Bødker, S. (1998). Understanding representations in design. Human Computer Interaction, 13, 107–125. Book, B. (2004). Virtual Worlds Review. http://www.virtualworldsreview.com/ Brown, B. & Bell, M. (2004). CSCW at play: ‘There’ as a collaborative virtual environment. Proceedings of CSCW 2004. Cherny, L. (1999). Conversation and Community: Chat in a Virtual World. Stanford, CA: CSLI Publications. Ermi, L.; Heliö, S.; Mäyrä, F. (2004). Pelien voima ja pelaamisen hallinta. Lapset ja nuoret pelikulttuurien toimijoina. Hypermedia Laboratory Net Series 6. Hypermedia Laboratory, University of Tampere. Farmer, F., “Social Dimensions of Habitat's Citizenry”. In The Virtual Reality Casebook. C. E. Loeffler and T. Anderson, (eds.) New York: Van Nostrand Reinhold, 1994. Greenbaum, J., & Kyng, M. (1991) Design at work: cooperative design of computer systems. Hillsdale, NJ, USA: Lawrence Erlbaum Associates Publishers. Hackos, J. T., & Redish, J. C. (1998). User and task analysis for interface design. New York: Wiley. Hagel, J., III & Armstrong, A. G. (1997). Net gain: Expanding markets through virtual communities. Boston: Harvard Business School Press. Hedron (1997). "The Six Circles of the Adept Game Player." http://www.falseprophecies.com/sixcircles.htm Hyysalo, S., Johnson, M. & Heiskanen, E. (Eds.) (2007). “Design-use relationships in Sociotechnical Change”. Special Issue, Human Technology 3(2). Available online at http://www.humantechnology.jyu.fi/ ISO 9241-11 (1998): Ergonomic Requirements for Office Work With Visual Display Terminals (Vdts) - Guidance on Usability. International Standard. Geneva, Switzerland: ISO.

19

Proceedings of the Nordic Consumer Policy Research Conference 2007 Johnson, M., & Toiskallio, K. (2005, August). Fansites as Sources for User Research: Case Habbo Hotel. Paper presented at the 28th Information Systems Research Seminar in Scandinavia (IRIS 28), Kristiansand, Norway. Johnson, M. (2007). Unscrambling the “Average User” of Habbo Hotel. Human Technology 3(2), 127-153. Johnson, M., & Toiskallio, K. (2007) “Who are the Habbo Hotel users?” In M. Turpeinen (Ed.), Mobile content communities final report. Helsinki: Helsinki Institute for Information Technology. Kim, A. J. (2000). Community building on the web: Secret strategies for successful online communities. Berkeley, CA, USA: Peachpit Press. Kling, R., & Gerson, E. M. (1977). The social dynamics of technical innovation in the computing world. Symbolic Interaction, 1(1), 132–146. Konzack, L. (2002). Computer Game Criticism: A Method for Computer Game Analysis, in CGDC Conference Proceedings, Frans Mayra (ed.), Tampere University Press 2002, pp 89100. Also available at http://imv.au.dk/~konzack/tampere2002.pdf Kuniavsky, M. (2005) Observing the user experience: A practitioner's guide to user research. San Francisco: Morgan Kaufmann. Laukkanen, T. (2005). Modding scenes: Introduction to user-created content in co mputer gaming (Hypermedia Laboratory Net Series, 9). Tampere, Finland: Tampere University Press. Maguire, M. (2001). Context of use within usability activities. International Journal of Human-Computer Studies, 55, 453–483. Mulligan, J. & Patrovsky, B. (2003). Developing Online Games. Indianapolis, Indiana: New Riders. Nardi, B. (1993). A small matter of programming: Perspectives on end user computing. Cambridge, MA, USA: MIT Press. Nielsen, J. (1993). Usability engineering. Boston: Morgan Kaufman. Nordic Council of Ministers (2006). Innovation og forbrugerkvalitet. TemaNord 2006: 541. Copenhagen, Denmark: Nordic Council. Norman, D., & Draper, S. (1986). User centered system design: New perspectives on humancomputer interaction. Hillsdale, NJ, USA: Lawrence Erlbaum Associates. Nurmela J., Parjo, L. & Ylitalo, M. (2002). Suuri muutto tietoyhteiskuntaan. Tilastokeskus Katsauksia. 2002/4. Pietiläinen, R. (2004). Pyydä mut kamuks ni kerron! Kommunikointi ja toiminta Habbo Hotel Kultakalassa. Pro Gradu, Suomen kielen laitos, Helsingin yliopisto. Preece, J. (2000). Online communities: Designing usability, supporting sociability. New York: Wiley. Salovaara, A., Johnson M., Toiskallio, K., Tiitta, S., & Turpeinen, M. (2005): Creative Playmakers and Their Motivations in Four Multiplayer Game Communities. Proceedings of the ACM SIGCHI International Conference on Advances in Computer Entertainment Technology 2005 (ACE'05), Valencia, Spain. Schuler, D., & Namioka, A. (1993). Participatory design: Principles and practices. Mahwah, NJ, USA: Lawrence Erlbaum Associates. Sihvola, Sebastian (2005): Netari.fi. Nuorisotyö Habbo Hotellissa. Nuorisoasiainkeskusen julkaisuja 2005 (1). Helsinki: Nuorisoasiankeskus. Svahn, M. (1999). Den liderliga kvinnan och den omanlige mannen. Skällsord, stereotyper och könskonstruktioner. Stockholm: Carlsson. Sveningsson Elm, M. (2007). “Doing and Undoing Gender in a Swedish Internet Community” in Sveningsson Elm, M. & Sundén, J. (Ed.) Cyberfeminism in Northern lights: Gender and digital media in a Nordic context. Cambridge: Cambridge Scholars Press. Turpeinen, M. & Kuikkaniemi, K. (Eds.). (2007). Mobile content co mmunities final report. Helsinki, Finland: Helsinki Institute for Information Technology. von Hippel, E. (1988). The sources of innovation. New York: Oxford University Press. Yee, N. (2002). Facets: 5 Motivation Factors for Why People Play Mmorpg's. http://www.nickyee.com/facets/home.html

20

ARTICLE

III


Mikael Johnson (2007). "Unscrambling the ’Average User’ of Habbo Hotel". Human Technology 3 (2), 127–153. Reprinted with permission.

209

ISSN: 1795-6889

An Interdisciplinary Journal on Humans in ICT Environments

www.humantechnology.jyu.fi

Volume 3 (2), May 2007, 127–153

UNSCRAMBLING THE “AVERAGE USER” OF HABBO HOTEL Mikael Johnson Helsinki Institute for Information Technology Helsinki University of Technology and University of Helsinki Finland

Abstract: The “user” is an ambiguous concept in human-computer interaction and information systems. Analyses of users as social actors, participants, or configured users delineate approaches to studying design-use relationships. Here, a developer’s reference to a figure of speech, termed the “average user,” is contrasted with design guidelines. The aim is to create an understanding about categorization practices in design through a case study about the virtual community, Habbo Hotel. A qualitative analysis highlighted not only the meaning of the “average user,” but also the work that both the developer and the category contribute to this meaning. The average user a) represents the unknown, b) influences the boundaries of the target user groups, c) legitimizes the designer to disregard marginal user feedback, and d) keeps the design space open, thus allowing for creativity. The analysis shows how design and use are intertwined and highlights the developers’ role in governing different users’ interests. Keywords: design-use relationships, user, designer, categorization practices.

INTRODUCTION Designers and users are commonly treated as trivial roles in research on design processes, but the question, “Who are the users?” is far from trivial. Take, for instance, the case in this paper, the on-line chat and game environment Habbo Hotel. Should the users be defined as the Internet surfers who visit the hotel at least once a month? This is the designation of users when discussing popular Web site statistics. Considering Habbo’s role in the everyday life of some teenagers, one could say that the Habbo users are mostly teenagers who—apart from going to school, practicing their hobbies, and spending time with their families—log on to Habbo for 30 minutes a day. Yet another way of describing the users, found on the Habbo Web site and in their press releases (Sulake 2006a, 2006b), is to say that they are creative habbos1 who decorate magnificent guest rooms in the on-line hotel and spend time with their © 2007 Mikael Johnson and the Agora Center, University of Jyväskylä URN:NBN:fi:jyu-2007277

127

Johnson

friends there. From a global perspective, one could perhaps say that the users are privileged children whose parents have an Internet connection at home. Or, considering all the different stakeholders of Habbo, the users could be defined as those whose income does not depend on Habbo’s commercial success. These brief examples show that, depending on who is asking, when and for what purpose, there is a wide range of alternative designations for the users2. This diversity of meanings for the term user highlights its elasticity, making it a blessing for some, but a curse for others. In addition, the term also is commonly used interchangeably with actors, participants, humans, knowledge workers, customers, interactors, citizens, and so on. It is no wonder that the usercentered design literature warns against designing for the “elastic user,” and recommends the use of techniques for personifying the user, such as personas (Cooper & Reimann, 2003). The ambiguousness of the user has been noted and problematized by a few researchers (Oudshoorn & Pinch 2003; Stewart & Williams, 2005; Westrup, 1997), and there are several different approaches to studying use-related processes of design and innovation in information and communication technologies. In this article, I distinguish between users as social actors (Lamb & Kling, 2003), users as participants (Hales, 1994; Kling & Gerson, 1977), and users as co-constructed and configured (Mackay, Carne, Beynon-Davies, & Tudhope, 2000; Woolgar, 1991). A focus on design-use relationships and an ambition to make sense of a developer’s reference to the “average user” guide the comparison of these approaches. The aim of this paper is to provide a means to understand and tackle the dilemma of user categories and their management in practice, in the design and the continued development of virtual world features for complex and heterogeneous communities. How users are defined and by whom greatly shapes the result of any design process, since undoubtedly the words that we use shape our outlook and affect our behavior. The increased number of people using computers at home and in leisure situations, as well as emerging software for virtual communities, has created a need for both researchers and practitioners to reassess Grudin’s question from 1993: “Have our terminology and habits of speech kept up with the changes, or do they perpetuate an outmoded perspective that holds us back?” (p. 112). This became one of the tasks for the Mobile Content Communities research project that started in 2003 (Turpeinen & Kuikkaniemi, 2007). One of the aims of the project was to understand virtual communities (Porter, 2004) by trying out design concepts, models, and prototypes. This author participated in a case study on Habbo Hotel, which contributed to one of the results of the project: a community model3 that aims to sensitize designers to ask the right questions about social and cultural issues of computer-mediated communities. During the 3 years of the study, Habbo grew from a popular on-line meeting place for 1 million regular visitors in five countries to more than 7 million visitors in 18 countries. The company developing Habbo, Sulake Corporation, grew from a small organization of about 50 employees to a medium-sized organization with 270 employees (Sulake 2006b). The fastest way to get insight into the Habbo communities is to read the Habbo-themed Web sites, also called fansites, created by active community members. These fan magazines on the Web are broadly known among Habbo visitors and many visit them as often as the hotel itself (Johnson & Toiskallio, 2005). The fansites can be seen as an amateur media world around the Habbo site that informs community members about Habbo news, gossip, opinions, hints, events, competitions, and so on. Because they complement the official Habbo Web site and influence the norms for behavior in Habbo, the fansites play an integral part in the communities.

128

Unscrambling the “Average User” of Habbo Hotel

In interviews I conducted with Habbo developers, I learned that the fansites have become an important source for user feedback. The fansites were a frequent topic during the interviews, and in this paper I reflect on one developer’s statement about the community members participating in the fansite discussions. The developer referred to the Habbo Hotel’s “average user,” which at first seemed like a strange category, especially since humancomputer interaction (HCI) literature discourages design for the average user. However, in this paper the developer’s reference to the average user is taken seriously. This concept of average user has been analyzed to understand which subcommunities are given voices through the fansites. Especially interesting is the developers’ agency, or their ways of shaping the user groups while they speak. First, this paper begins with a literature overview of different approaches to users. It continues with a brief introduction to the fansites, to contextualize the text extract from the interviews with Habbo developers. After the analysis of one developer’s reference to the average user, these findings are related to the other Sulake developers’ ways of speaking about the users.

WHO IS THE “USER”? User is a highly controversial and ambiguous term in computing. Recent user-centered design literature defines the user in two major ways. First, users are defined in terms of their knowledge of computers or of a particular computer program: as novice, intermittent, or expert users (Shneiderman, 1998). Second, a distinction is often made between those who actually operate the computer, the primary users, and the secondary users, who are indirectly affected by the computer system (Courage & Baxter, 2005; Hackos & Redish, 1998). In addition, Courage and Baxter (2005) recommend considering anti-users, who would not buy or use the product. Hackos and Redish (1998) warn against confusing users with buyers, against interacting with surrogate users only, and recommend studying users as members of communities. Table 1 summarizes these viewpoints. However, outside local design contexts and for research purposes, there is a need to define the user more carefully to achieve a shared understanding. This has been a challenge, since different fields of research discuss users differently. In HCI, user traditionally referred to a person sitting in front of a computer in an office, completing tasks by entering information and commands and using the output. In contrast, the information system (IS) user refers to a beneficiary of the computer output, who might be a person (or even an organization) not directly interacting with the computer. To distinguish the person sitting in front of the computer, IS researchers talk about “end users” (Grudin, 1993). Table 1. Users in a Sample of the User-Centered Design Literature.

Hackos & Redish (1998)

Shneiderman (1998)

Courage & Baxter (2005)

Primary users Secondary users User communities ¬Users as buyers ¬Surrogate users

Novice or first-time users Knowledgeable intermittent users Expert frequent users

Primary users Secondary users Tertiary users Anti-users

129

Johnson

Over time, these differences have changed, since the HCI researchers have followed the user outside the office, into mobile technologies, and the use of computers in leisure contexts. In particular, research on user experience and mobile HCI puts the user in mobile contexts of use (Maguire, 2001; Toiskallio, Tamminen, Korpilahti, Hari, & Nieminen, 2004) with a focus on fun and pleasure (Blythe, Overbeeke, Monk, & Wright, 2003), instead of the traditional effectivity and efficiency focus within work contexts. Also, some researchers report that the term user is understood differently in the USA as opposed to Scandinavia. Carmel, Whitaker, and George (1993) find an unambiguous definition of user impossible: The North American reader understands “user” to mean any non-IS/nontechnical individual in the organization who is affected by the system—this includes managers. The Scandinavian reader understands “user” to mean any operational worker who is affected by the system—this does not include managers. (p. 40) Friedman and Cornford (1989, p. 274) report similar differences, stating that Scandinavians seem to be more likely to consider power relations between users and system designers in their definitions, whereas American analyses tend to focus on personality conflicts and differences in cognitive styles. However, these generalizations have exceptions (Kling & Gerson, 1977; Lamb & Kling, 2003; discussed below), so it is not clear to what degree these cultural aspects have shaped or continue to shape research on the user. The globalization of research and the increase of the number of both publications and researchers have probably made the fields more heterogeneous. Another difficulty in defining the (computer) users is that the computer technology has changed a lot since its invention. From being room-sized and expensive equipment available only to a few, the hardware innovations and mass production of computers have made them smaller, mobile, and more affordable. This shift is noticeable in the change of the meaning of the term user. In the early days of computing, the user referred to those who used computer hardware, in other words, software developers (Friedman & Cornford, 1989; Westrup, 1997). In sum, user is a complex term. So far, one can only say that it is used in the context of computer systems development, and its meanings have changed over time and space (e.g., USA vs. Scandinavia). This encourages a further look into different approaches to users. Without aiming for completeness, I will explore three approaches in which users can be distinguished: as social actors, as participants, and as co-constructed and configured. Users as Social Actors The definition of users within the HCI and IS fields has been challenged because of its implicit technocentrism (Kuutti, 2001). Those opposing the use of the term argue that when a person is defined with respect to a technological system, as a computer user for instance, this represses other more relevant identities as well as the multiple and intertwining reasons for use. In contrast to many car owners, who identify themselves as owners, few computer users make a lifestyle (or identity) of being someone who uses computers (Grudin, 1993). More important identities are their professional identities, the relation to their family and special someone(s), hobbies, and so on. Most people who use computer applications utilize multiple

130


applications, in various roles, and as part of their work or leisure activities, while interacting with a variety of other people in multiple social contexts (Lamb & Kling, 2003). This critique is noted in some studies on designers’ conceptions of the users (Dagwell & Weber, 1983; Isomäki, 2002; Kuutti, 2001; Nurminen, 1988), arguing that these conceptions influence the design more fundamentally than do applying human-centered design methodologies. The fear is that the designers cannot contribute to the humanization of computer systems with socially thin concepts of the users. Lamb & Kling (2003) have responded to the critique of those user concepts that include only individualistic or cognitive dimensions, by reconceptualizing the user as a social actor. Their view of a social actor is based on four dimensions: affiliations, environments, interactions, and identities (see Table 2). Their concept has clear benefits: It is “1) predictive without being deterministic, 2) scalable, based on the multilevel explanatory power of institutional theory, and 3) extensible in multiple ways” (Lamb & Kling, 2003, p. 221). This critique intentionally leads away from the reference to systems development in defining users, which can be fruitful in studies of ICT use. Research in the field of information systems is influenced by and related to both the “social shaping of technology” approach (MacKenzie & Wajcman, 1985) and the “social construction of technology” program (Bijker, Hughes, & Pinch 1987)4. These approaches shared the aim of criticizing technological determinism by coupling technology to designers’ conceptions and values. In addition, political, economic, and cultural interests as well as established social categories, like class and gender, were seen as shaping technology (Sørensen, 2002). However, even in these more nuanced models of the user, the perceived “problem” with current design processes seems to be the designer. Seeing the designers as almighty heroes or demons, either praising or blaming only them, has been criticized as the “design fallacy” by Stewart & Williams (2005). They argue that a design-centered view describes technology as finished when leaving the hands of the designers, which ignores the innovation that takes place in use. If technology is seen as influenced by designers only, this also ignores user involvement in design. These shortcomings are partly resolved if the users are seen as participants in design, as in the next subsection. Table 2. Social Actor Dimensions (Lamb & Kling, 2003).

Affiliations

organizational and professional relationships that connect an organization member to industry, or national and international networks

Environments

stabilized, regulated, and/or institutionalized practices, associations, and locations that circumscribe organizational action

Interactions

information, resources, and media of exchange that organization members mobilize as they engage with members of affiliated organizations

Identities

avowed presentations of the self and ascribed profiles of organization members as individual and collective entities

Users as Participants The definitions of users can also be seen as reflecting design philosophies, as has been the case particularly in the computer-supported cooperative work (CSCW) field (Mackay et al., 2000). For instance, Mike Hales (1994, p. 155) discusses different conceptions of users based on

131

Johnson

different design styles: users as clients (the “specify and deliver” style), users as actorconstructors (the “enable and empower” style), and users as codesigners (the “reflect and reinterpret” style). These distinct design styles designate the different ways of users participating in the design. The multiplicity of actors in design is highlighted by Kling and Gerson (1977), who distinguish 14 major orientations that people may adopt within the computing world (Table 3). They are arranged by their closeness to the center of the computing world (Kling and Gerson’s concept for all those people and groups that collectively produce computers and computer-based services). The first eight orientations are insiders, while people who adopt the last six positions are more in the margins. From this perspective, the term user is not about identification but participation in the technology production. The user is seen as a very established category referring to a particular way of participating in technology production. It is not a very favorable position, since traditionally the innovation is seen to flow from the developers via the other stakeholders to the users, and not vice versa. The interests of the users traditionally are not served first, because business and technology production are primary, which leaves the user interests in a challenging position. In Hales’ (1994) view, where the users’ relation to production becomes visible, emancipating the users becomes less a question of finding other labels for this group and more a question of organizing the production differently. Friedman and Cornford (1989) have studied the history, organization, and implementation of computer systems development. They found a six-fold typology useful to explain the changes of the computing world over time: a) patrons, who initiate the system, b) clients, for whom the system is intended and designed, c) design interactors, who are involved in the systems design process, d) end users, who directly operate the man/machine interface, e) maintenance/enhancement interactors, and f) secondary users (both the system victims and those who benefit indirectly from it). According to Friedman & Cornford (1989), the user relation became the biggest critical factor constraining development in the early 1980s, after the hardware and software constraints were mitigated in earlier phases. To decrease the distance between programmers and end users, five major strategies emerged to reorganize technology production: a) user involvement, b) end-user computing, c) decentralization, d) prototyping, and e) job rotation (p. 271). Table 3. A Sample of Research on Users as Participants.

Kling & Gerson (1977)

Friedman & Cornford (1989)

Hales (1994) users as clients (the “specify and deliver” style)

technology stimulators

users

patrons

innovators

feeders

clients

diffusers

tenders

design interactors

vendors

sustainers

end users

service providers

hobbyists

educators

consumers

maintenance or enhancement interactors

system architects

secondary users

application architects

132

users as actor-constructors (the “enable and empower” style) users as co-designers (the ”reflect and reinterpret” style)


These and other movements strive to give the users greater influence in the technology production. Some aim to influence the process: involving the users early on, demanding a multidisciplinary development team, or applying iterative design cycles with prototypes and user evaluation in all phases. As Hales (1994) noted, the emphasis put on participation, democracy, and emancipation varies in different approaches. For instance, participatory design (Ehn, 1988; Greenbaum & Kyng, 1991; Schuler & Namioka 1993) gives the users the status of codesigners, whereas many user-centered design techniques (Beyer & Holtzblatt, 1998; International Organizations for Standardization [ISO], 1999; Norman & Draper, 1986) maintain that consulting the users is enough. Others initiatives try to influence the product by making sure that the division of labor between people and machines is appropriate (Mumford, 1983), or making it flexible to support customization (Andersen, 1999; Kay & Goldberg, 1977; Laukkanen, 2005; Mørch, 1997; Nardi, 1993). A third focal point is the methods and techniques to represent the work processes and their relations to the computer system (Bødker, 1998; Checkland & Scholes, 1981; Mumford, 1983; Suchman, 1995), as well as the models of the user and contexts (ISO, 1998, 1999; Maguire, 2001). However, despite the good intentions and the recognition of the multiplicity of actors involved, these participatory and user-centered design initiatives have been criticized due to lack of impact on system design overall. Stewart and Williams (2005) note many obstacles to the wider applicability and uptake of the initiatives, and strive to broaden the understanding of the design process. They argue against too simplistic models of the design process and point to studies where many aspects of it have been problematized. One important point is to rely not on armchair philosophy regarding the user, but rather to ground the term user empirically in order to understand design practices. Users as co-constructed and configured Since the early 1990s, some researchers have shaped an approach that studies the “configuring of the user,” which can be seen as an extension of the broader social shaping research community (Sørensen, 2002) mentioned above. It takes as its starting point that there are no users prior to the conception of a particular computer system. The argument is that qualitative research should not take a category such as the user as given, but instead acknowledge the considerable work that has gone into its constitution (Westrup, 1997). This implies a distinction between the users as imagined by the developers and the users who actually use the system, including an analysis of their interrelations. Woolgar (1991) coined the notion of configuring for the process of “defining the identity of putative users, and setting the constraints upon their likely future actions” (p. 59). His work is an important theoretical move for studying how users are imagined in computer systems development (Mackay et al., 2000). It is one alternative to study the interrelations of the social and the technical in design, especially regarding the construction of “affordances” (Norman, 1988) and “mental models” (Norman & Draper, 1986), both practical and widely used design concepts. Woolgar’s (1991) work on the configuring of the user in a microcomputer manufacturing company has been both extended and criticized. Among others, Westrup (1997) noted that not only are the users configured, but so are the developers. He used his extended version of Woolgar’s work to examine two approaches to requirements analysis in a novel way. He argued that the categories of users and designer are constituted by the techniques that seek to

133

Johnson

represent them, in this case Mumford’s (1983) influential ETHICS/QUICKethics methodology and the techniques of the participatory design project UTOPIA (Bødker, Ehn, Kyng, Kammersgaard, & Sundblad, 1987). Westrup (1997) also highlighted difficulties in the very vocabularies of systems development. Many authors (Hyysalo, 2004; Mackay et al., 2000; Williams, Stewart, & Slack, 2005) criticized Woolgar in that he stopped his analysis too soon, not giving room for the process of consumption work by the users. Woolgar’s (1991) approach, together with similar initiatives from Akrich and Latour (Akrich, 1992; Akrich & Latour, 1992; Latour, 1992), are termed material semiotics by Oudshoorn and Pinch (2003), as they outline influential approaches to the co-construction of users and technologies. Another approach has emerged from gender studies, where the focus is on the mutual shaping of gender and technology, as well as an inclusion of more invisible or implicated actors. Consumption and domestication studies de-emphasize the developers and focus on active consumers using products in ways that might not have been imagined by developers. Agre (1995) suggested that developers do not intentionally try to configure users, and Hyysalo (2004) has drawn attention to the imaginaries bound in the professional practices of developers. Williams et al. (2005) warn against the tendency of the early work (Akrich, 1992, 1995; Bijker et al., 1987; Woolgar, 1991) to demonize designers as omnipotent manipulators of users, which they see as a consequence of studying snapshots of design or use processes. They argue that technologies should be seen and studied as sums of many projects, configurations of previous technical frameworks, and never complete. All actors involved at multiple locations need to be considered, as well as their interrelations, while remembering that information about the users is typically incomplete and uncertain. To sum up, there is no general answer to the question “Who is the user?”: It is up to the researcher to make the relevant aspects of the user explicit, both theoretically and empirically. I have outlined three different approaches to the user, finding that especially the reconceptualization of the user as a social actor is appropriate if the focus is either on design or use alone. When users are seen as participators in computer systems development, the point is that not all stakeholders can participate on an equal basis. When users, designers, technology, organizations, and so on, are seen as co-constructed, the point is that the distribution of agency, power and actors are empirical questions.

METHOD AND DATA This paper is part of a larger research undertaking where I study the slowly paced dialogue between different developers and users through material software. Because the Habbo software has several design cycles in a year, it has been possible to study a dialogue that starts from the developers’ vision of future use. The vision is realized during game development into Habbo features. Next, the users appropriate the hotel features for their own purposes and invent their own ways of using it. Soon after, the developers learn a little about these more or less unexpected use practices through various forms of user feedback. This feedback modifies the designers’ original vision about future use and, as new features are developed, the cyclical dialogue continues with the next loop. In this paper, I freeze the moment when a developer talks about user feedback.

134


One developer’s reference to the “average user” is carefully scrutinized. It is studied as an example of a general dilemma for people engaged in creating products for mass consumption: There is no single “correct” way of categorizing users. Still, all of us act in worlds where we must use categories, and the question is how to manage this dilemma. In this study, the text extract around the “average user” is analyzed with the interview data analysis technique called membership categorization (Baker, 2004). The interviews are not treated as simple reports on some state of affairs (representationalism), but talk is considered as social action. This makes it possible to analyze not only the meanings of the categories, but the work the categories do and how they influence development. The categories referred to in the interview talk are treated as empty at first, giving them only the meanings emerging from the particular situation. During the interview the membership categories are elaborated on and refined; the established categories might even be internally conflicting. The point is to understand why conflicting categories are needed to communicate. In this case study, we (Johnson & Toiskallio, in press) had the opportunity to conduct both quantitative and qualitative research to understand the Finnish Habbo communities. We started with an explorative survey on the visitor profiles (N = 10,000), because no use statistics were available. In 2004 we identified 173 Finnish Habbo fansites, and we analyzed 23 that were written for a large Habbo audience (Johnson & Toiskallio, 2005). The survey findings provided background statistics, whereas the fansites and forum discussions allowed us to distinguish user groups and popular activities. During the spring of 2005, we conducted 10 themed interviews, lasting 2 to 3 hours each, with 10 Habbo developers, or about two thirds of the Habbo game development organization. Six of the interviewed developers (graphical designers, and client and server developers) had been in the organization since the beginnings, 5 years earlier, while four developers had about 1 year of Habbo experience. In addition, we conducted individual, pair, and group interviews, 2 to 3 hours in length, with a total of 12 Habbo community members (users) from different subcommunities. These interviews made further elaboration of membership categories possible. The interviews were conducted in Finnish, as all developers and the author are fluent speakers of Finnish. All of the interviews were recorded and content logs created. The interviews were transcribed in detail and excerpts translated to English by the author as needed. The different data sources have afforded triangulation (Yin, 1994) of the slowly paced user-developer dialogue. Preliminary results include conference talks on the design reasons behind the retro look of Habbo (Johnson, 2006a), user categorization practices (Johnson, 2006b), and stereotypical images in membership categorization practices (Johnson, 2006c). The approach adopted here is informed by qualitative research (Silverman, 2004) in HCI (Thomas, 1995) and IS (Lee, Liebenau, & DeGross 1997), as well as science and technology studies on the social shaping of information and communication technologies.

CASE HABBO HOTEL Habbo has its origins in the experiences drawn from two Internet chat rooms, Mobiles Disco and Lumisota. In terms of Internet technologies, Habbo is a graphical chat environment on the Web that can be accessed with a Web browser with the Shockwave plug-in. This chat

135

Johnson

environment is designed as a virtual world where people can hang out and make new friends. When checking in to the virtual hotel, one creates one’s own cartoon-like Habbo avatar (Figure 1) that can walk, dance, eat, drink, and chat in the cafés (Figure 2), restaurants, swimming pools, and games rooms. Besides experiencing these common rooms in the hotel, one can decorate and furnish a room of one’s own. In contrast to many on-line games, there is no entrance fee to the virtual world, which allows the majority of the users to chat for free. Instead, the profit model is based on micropayments in the hotel. Virtual furniture, minigames, and membership in the Habbo club are bought with so-called Habbo credits. These credits can be purchased with prepaid cards, bank transactions, or special text messages that add a specified amount of money to the customer’s mobile phone bill.

5

Figure 1. A Habbo avatar. The image © by Sulake Corporation. Used with permission .

Figure 2. The welcome lounge in Habbo. The image © by Sulake Corporation. Used with permission . 5

136


During the early Habbo days, the hotel was developed by a handful of game developers with their core competencies in graphic design, macromedia flash clients, and java server programming. At first they developed a hotel called Kultakala (goldfish in Finnish) for themselves and their friends but, already within a year of the launch, the site became popular among teenagers. According to our survey in 2004, 75% of the visitors were between 10 and 14 years old. Later, with the internationalization of the hotel, the organization grew and, in every country where the Habbo Hotel operates, a local office was started with a few employees working on the site moderation, community management, customer relations, and marketing. More administration and business people have joined the headquarters in Finland, and the game development division now includes more than a dozen game developers. Based on developer interviews, it became clear that the original group of Habbo developers had made use of their own retrogaming subculture. Video games from the 1980s are and have been a great source of inspiration (Johnson, 2006a). This is of course easily deducible by looking at the retro appearance of the Habbo lounge in Figure 2, but there is more to it. The benefits of making Habbo retro include positioning Habbo as original compared to other games that strive to be photorealistic. It also avoids pressure to follow fast-changing 3D graphics techniques, which in turn allows for Habbo to stay popular for a longer time. In addition, there are reduced performance requirements for both client computers and network bandwidth. Finally, the retro look encourages a simplistic design making Habbo easy to use. Learning from the Habbo Fansites We (Johnson & Toiskallio, 2005) studied 173 Finnish Habbo fansites in 2004. The focus was on what could be learned about the Habbo visitors and their Habbo practices from user research (Hackos & Redish, 1998; Kuniavsky, 2005), focusing on the membership categories visible through the fansites. Since the fansites are accessible without research intervention, the risk of distorting the data by the presence of the researchers is reduced. The most popular fansites are usually made by a small team of Habbo fans with different expertise and roles. For example, one designs the look and layout of the site, another one writes the stories, and a third has the technical skill to publish the site on the Web. The most active fansites have small updates (such as news and rumors about Habbo) several times per week, but publish reviews and articles once a week or bi-weekly. Some fansites group their articles together and publish them as an issue of a Web magazine. The Web magazines seem to follow a rhythm of one issue per one or two months. The fansite contents were classified and a list of common fansite elements was produced (Table 4). Based on the fansites, we clustered the hotel residents into eight groups: a) furniture traders and collectors, b) chatters (in public rooms), c) gang-members and VIPs (insider groups not open to everyone), d) supervisors with administration powers, e) cheaters, f) quiz-makers and players, g) the hotel manager (a Sulake employee), and h) celebrities. Similarly, 11 popular activities were identified: trading furniture, casinos, dating, beauty contests, competitions, dice games, team sports, formula tracks, talk shows, clubs & hotels, and orphanages. More important than the exact details of these listings are two observations about Habbo that they convey: the diverse and commonplace qualities of Habbo. First, there is not one particular Habbo activity that attracts all Habbo visitors, but many different ones. Second, the activities

137

Johnson

Table 4. Common Habbo Fansite Elements (based on Johnson & Toiskallio, 2005).

Fansite Elements Description News and rumors

Fansites are convenient for Habbo visitors who want to reach a large audience, a fast way of spreading information about Habbo happenings (e.g., competitions, pop idols visiting Habbo), new features, news about Sulake Corp.

Participation

The fansite audience is provided ways to comment on the fansite through discussion forums, guest books, polls, etc.

Links

The fansites link to relevant Habbo places: other fansites, and to the hotels in other countries.

Hints, secrets, guidelines

Fansites teach newcomers both basic and advanced tricks with which to impress others. Guidelines on acceptable behavior are frequent.

Reviews and lists

The fansites keep track of the features and possibilities in Habbo: public spaces, different furniture items, pets, etc.

Histories

Two major histories are told on the fansites: the history of Habbo and the history of that particular fansite

Fashion and celebrities

Habbo “journalists” interview Habbo celebrities, avatars who have become famous in Habbo, and report on fashionable clothing and activities.

Graphics

Edited screenshot pictures are an integral part of many fansites, some even provide pixel graphics drawing schools.

Habbo fiction

A few fansites write fictional stories about characters in Habbo.

About

Who comprises the fansites staff, number of visitors, updates, banners, etc.

Real life

Habbo meetings “in real life,” stuff not about Habbo that is important to teenagers, as well as blogs, e-cards, etc.

going on in Habbo resemble games with rules and pretend play familiar from schoolyards, playgrounds, youth clubs, and so on. The Habbo visitors and their practices seem to be strongly influenced by the fansites. They complement Sulake’s official Web site by providing more detailed information about the hotel from an experienced visitor’s point of view. Hints, secrets, and guidelines, and stories about Habbo fashion influence the boundaries for acceptable behavior in Habbo. The fansites improve the Habbo visitors’ awareness of the fan cultures around Habbo, and also reproduce and reinforce social positions (like potential Habbo career paths or legitimized visitor groups). Membership Categories The membership categories from the fansites were further developed, based on interviews and artifacts analyses (Johnson, 2006c). In a discussion on the influence of images on membership categorization, Johnson discussed visual and nonvisual categories (Table 5).

138


Table 5. Predefined and Emergent Visual and Nonvisual Membership Categories based on Johnson, 2006c).

Predefined visual categories Emergent visual categories Non-visual categories avatar appearance: boy / girl purchasable badges: Habbo Club, Golden Habbo Club, Halloween smile special badges: Habbo staff, NGO workers, youth workers, mental support, VIP guests earnable badges: Habbo X (guides), fansite authors, Battle Ball gurus

clothing styles: punk, gothic, teenie, wannabe (strictly dressed), personal style (independent of fashion) professions: journalist, nurse, TV show host, bartender / waitress, pharmacist, actor, police, doctor, nurse, fireman, postman, veterinarian, etc. TV show formats: Idols, the Bachelor(ette), Greed, Do You Want to be a Millionaire, Big Brother, Survivor, America’s Next Top Model, “Dating,” etc.

visitors from other hotels (nationality / language region) speaking another language age: small children, “my age” (teen), older gender: combinations of nickname, avatar, real body (e.g., boy with girl avatar but masculine nickname) Habbo age: newbie, regular, guru relation to room: room owner, visitor, shared rights friends made in or outside of Habbo: Habbo friends, real friends

trading furniture: little by little, skilled traders, cheaters categories formed from value judgments on others’ time of day (is associated with appearances: good looking, distinct visitors): daytime (children bad looking with flu at home, mothers), after school (preteens), evening-night (older, best discussions)

To sum up the Habbo experience, the following dimensions can be found to delineate what is important in Habbo: x one’s own avatar: clothing styles, character description x one’s own room and furniture: collecting, trading, decorating x friends: school, hobbies, new friends, dating, distant friends x play: beauty contests (popularity), TV shows, games of chance, Habbo-sports, insider clubs, roleplay, playing with the spatiality of the virtual world x Habbo Career: celebrities, getting rich, popular room, in a game or gang, being a fansite author, being a Habbo guide x testing boundaries and rules: expressing self, treating others (e.g., cheating, bullying), finding and using glitches in the hotel architecture. The Role of Fansites in Development The fansites are not only important to the visitors, but they also play a big role in the game development. The developers visit them regularly to follow what’s going on, reading both articles and forum discussions. Otherwise much of it [user feedback] doesn’t come all the way to me; that’s why I try to look at those ... Habbo fansites, what they discuss there, because if I have to

139

Johnson

develop something for Habbo, then I should know what’s going on there. (Game Developer61) Game Developer [GD] 2: We get feedback from the users all the time; it’s something that is really important. Author [A]: In what ways do you get feedback? GD2: If we publish, for instance, a red chair then, as soon as it’s out there, then the users tell us if it’s good or not. A: Where, in Habbo or somewhere else? GD2: Well, in Habbo and on the users’, those, fansites. Well, those fansites are really important, of course. There things are said quite frankly. Then there are some e-mails directly from the users, though I never reply to them. Sometimes I followed the feedback e-mails and the bug e-mails some many years ago, and then directly from friend-users I get [feedback]. But if I want to know how some new feature is liked, then I go and look at the fansite forums. Of course, they are always a little over-critical. (Game Developer 6) The first quote (GD1) shows how the developer feels a responsibility to check out the fansites: it is as if they are part of the job. The second quote (GD2) gives an example of the immediacy of the fansite feedback, while the third quote (GD6) shows a brief comparison of feedback sources in favor of the fansites. Some more active fansite readers among the developers send e-mails to the other developers about interesting articles. Well, actually, to be totally honest, what always gets me to surf there is usually that [one developer], or somebody else who follows them actively, sends a link that “Here is something funny.” Then I usually go and look, but I don’t remember when I last would have gone looking just myself. (Game Developer 4) This quote (GD4) shows how one of the developers lets another developer act as a mediator to the fansites. Some developers, who are eager to get responses on what they have created, visit the fansites regularly, especially after new Habbo releases, when the users discuss the new features. Well, it is probably a bias towards the final stage of the release, when something comes out, one can see if there is something new about it. But yes, every now and then when, when there is a pause or moment that I don’t have anything terribly urgent, then I can go and peek at them. (Game Developer 9) One developer describes how the fansites can influence the development of new features: “The fansites are worth gold. From them, one can see what they expect and what they, on the other hand, what they don't expect, and then one can do that too” (Game Developer 3). This is worth underscoring. Because the fansites are “out there” without developer intervention, the discussion topics are not limited to specific questions from the developers’

140


side: They are rather fairly open-ended. The above quote (GD3) shows how this enables the developers to learn about what the fansite writers expect, but also what they do not expect. Not all developers have equal interests in the fansites, often because of their different work tasks. Server developers are furthest away from the users, whereas client developers and graphic designers are closer to the users. This can be seen as shaping their interests in fansites as well, allowing, of course, for individual differences. One developer described the graphic designers’ interest in fansites: “When a graphic designer designs a new [Habbo] space, s/he7 is extremely interested in how the users receive it. And because of that s/he surfs on the fansites and discussion forums to find out what’s said about it” (Game Developer 4). However, another developer stated that the fansites do not provide reliable user knowledge. “From there [the fansites] one doesn’t get real knowledge [about user opinions]. But [by] going into the [guest] rooms, one gets better [knowledge]” (Game Developer 8). This developer was concerned that the fansites do not provide representative user feedback. He continues by discussing the user groups visible through the fansite, and coins a category “average user.” This is interesting because the average user seems to have an effect on the boundaries of the other target user groups.

THE “AVERAGE USER” In design guidelines, the category “average user” is mainly used in two ways. On the one hand, it is contrasted with more technically skilled developers, suggesting that developers should remember not to assume too much technical competence among most of the users (Spillers, 2006). On the other hand, in HCI literature a common phrase is, “There’s no such thing as an average user” (Budde, 2004, p. 54), which can be seen as a warning against reducing identities, practices, and tastes into too abstract user needs. Abstract user needs that are not grounded in particular settings might harmfully shape a design into something that nobody can identify with (Cooper & Reimann, 2003). However, in my empirical data, yet another aspect of the average user has emerged, which is not about the technical skill of developers or average users, nor about statistical methods to advocate a representative user (Muller, Millen, & Strohecker, 2001). In my analysis, the average user is used in relation to other user groups, not developers. By analyzing the following quote, one can better understand the complexities of categorization practices in design for complex and heterogeneous communities: If one goes to the [guest] rooms ... then one gets feedback from the average user. But in the forums, the users have used Habbo Hotel for a longer time and slightly grown out of it [Habbo] and they have moved on to the forums to discuss it. There are the [furniture] collectors, the older ones, and the other HC [Habbo Club] users.... I have the impression that the opinions are not that black and white among normal users. In the forums everything is either extremely great or then it really sucks. (Game Developer 8). I started the analysis by reading what was stated by the developers about the average user category, and then what this category’s relation was to the other mentioned categories. In the

141

Johnson

above text fragment, the average user concept is used twice8, first to state that the average user rarely writes in the fansite forums, and then that they have more nuanced opinions than those expressed in the fansite forums. However, the reason for talking about the average user emerges by reading the whole paragraph. The function of the average user in this text is to contrast them with other mentioned user categories: furniture collectors, older users, and Habbo Club users. So, actually the average user cannot be taken literally, since the developer is not talking strictly about the average user of the whole Habbo population, but about the users who are not opinionated furniture collectors, older, or Habbo Club users. This interpretation is supported by the change to talk about “normal” users. Based on his experience, this game developer feels there is a large group of users whose opinions do not get voiced in the fansite forums. Still, he feels it is important to include them in the design considerations. The problem is that since they are not so opinionated, they are hard to reach easily, and not much is known about them. Even though the group is probably as heterogeneous as any other user group, it is difficult to distinguish the subgroups. Therefore they are grouped together as the average user. Why the name “average user”? One interpretation is that it implies large masses of users; as the developers want to please as many users as they can, the “average users” get more emphasis than other more marginal (although well seen and heard) user groups. The designer not only describes the user groups, but also actively constructs and configures the user groups in his speech, while reflecting on the constructed user groups that have influenced his earlier actions. He actively speaks for some users, and devalues other users, who in his perspective can and do speak for themselves. User Feedback Seen as Reflected and Refracted By Fansites In this section, the users’ voices and the developer from the previous quote will be visualized through three simple sketches. Although the first drawing is problematic, it assists the understanding of the second, as it makes it possible to contrast and compare the two drawings. This comparison brings forth qualities of the average user that elaborates on the previous analysis. Physics is a source for inspiration to understand social relations: think of attraction and repulsion, for instance. When retheorizing relationality, gender researchers have criticized reflection and refraction as useful optical metaphors, and turned to diffraction and interference as more promising candidates (Barad, 2003). In this article, to illustrate the user– developer dialogue described above, I will follow the same reasoning. Inspired by reflection and refraction, I draw as Figure 3 an image of the user–developer dialogue through the fansites, criticize it, and come up with a more useful visualization. Figure 3 shows how voices (the waves in the image) of different user groups become either reflected or refracted by the fansites. The image succeeds in showing that not all users get their voices heard on the fansites.

142


Figure 3. Drawing displaying user feedback that is reflected (average users) and refracted (furniture collectors, old-timers, Habbo Club users). The drawing illustrates that not all users get their voices heard on the fansites, but its lack of dynamics is problematic.

However, the metaphors of reflection and refraction have their limitations. First, the average user gets the same ontological status as the other user groups. This is wrong, because it is not possible to observe an average user: One can only observe particulars, whereas the “average” is an abstraction. Also, this drawing is not dynamic; there is more going on in the text fragment. Both the game developer and the average user are more active, yet their agencies are rather absent in this drawing. Of course the furniture collectors are a constructed and heterogeneous group, since there are probably many different reasons for collecting furniture. The point is that, except for the average user, the defined groups have a fairly clear “shape.” They can be located in user practices. But the average user is somehow more open, and not so stable or well defined as the other groups. It is as if the term average user becomes a placeholder for the unknown users. “Configuring” the User Groups The second visualization is based on the form of a comic strip. I tried to make one drawing of diffraction, but it was difficult to get a single image dynamic enough. As I started drawing several images in a row, the whole changed, and the optical metaphors became deemphasized. However, the qualities associated with diffraction (interaction, interference, reinforcement, difference) are present in Figure 4, which presents a more dynamic illustration of the user–developer dialogue through the fansites. The first panel shows some of the user groups on the fansites. In the second panel, the developer discusses the user groups that get voiced on the fansites. The next panel shows the developer creating the “average user,” which in the final panel dominates the other user groups. In this comic strip, the active role (agency) of both the game developer and the average user become visible, and the rigidity of Figure 3 is swept away. Making the average user bold and larger than the other user groups in the fourth panel symbolizes the work the category does. Even though the average user is fairly shapeless and unknown, it seems to direct the design. The developer wants to design for the average users rather than pay much attention to what the opinionated user groups write in the fansite forums. Creating the average user concept is a way of legitimizing this. Using the average user

143

Johnson

Figure 4. Configuring the users. The first panel shows the user groups on the fansites that the developer is very familiar with. Panels 2-4 show how the developer, in his discussion of the user groups that get voiced on the fansite forums, creates the concept of the “average user,” which then dominates the user groups in design talk/practices. The developer interacts with the user groups and reinforces a group in favor of others. The resulting target user group constellation is not a mere displacement of panel 2, but show the difference that matters in the subsequent design considerations.

in connection to the fansite forums (referring to the users not writing there) also leaves room for more representative user feedback from other sources. Yet another reason for leaving the average user shapeless is perhaps to allow and give room for new designs. Had the average user been completely defined, there would not be room for creativity. To sum up, the category “average user” cannot be taken literally. It gets its meanings in relation to the other categories defined and mentioned. These categories are different, as they are locatable in user practices, while the average user is not. Furthermore, the categories are shaped or configured by the developer, to fit his aims. At the same time, these configured categories shape the developer, as they keep the design space open. This second element, however, is still missing from Figure 4. To change Figure 4 from a rhetorical drawing towards a description of the co-construction of the developers as well, the transformed developer needs to be visually marked as well. Figure 5 is an attempt at visualizing this by fading out bars in front of the developer. The bars represent design obstacles that are mitigated as the designer speaks.

Figure 5. Visualizing the “configured” developer as well. The fading bars in front of the developer represent design obstacles that are mitigated as the designer speaks. The numbering of the panels start 2, as it refers to the first panel where the developer is visible in the panel series of Figure 4.

144


Now, that the “average user” has been unscrambled, this paper will conclude by zooming out from this particular situation and this particular developer. This analysis is related to the other Sulake developers’ ways of speaking about the users.

HOW FAR DOES THE “AVERAGE USER” REACH? How far does the agency of the average user reach? Do all designers at Sulake talk about the average user? Does it shape other different representations of the user? When reviewing my interviews with the designers other than GD8, I did not find exactly the same configured actor of “average user”: The others did not talk about the average user as such. However, when considering the different aspects of the average user, I found similar configurations, such as the “typical user” and the “normal user.” One developer discussed the importance of fixing problems for certain users, favoring the normal user over trouble-causing users. “It is perhaps less important... if they [the troublemakers] get some problems from causing problems to other users. Then it is not necessarily as serious as when the normal user gets problems from using Habbo normally” (Game Developer 7). Here, normal again designates everybody else, except for the troublemakers. The category “normal users” is not locatable in user practices, while the trouble-causers are. This developer also shapes the categories, giving the normal users a higher priority than the troublemakers. At the same time, these categories shape the developers’ actions, freeing the developer to design for the normal use, instead of the exceptional and generally not desired use. Another developer discussed a “typical user” when asked whether enough is known or not about the users. GD9: ...but, when we think about a new big feature... we should know how large a group, or how large the number is, absolutely, who would use it then. Another thing is that we don’t really know how lasting they [the groups] are, that for how long-term one can design something.... So the user, the typical user, is still a little bit hazy at this time. A: The typical user isn’t visible in the forums or the statistics? GD9: It isn’t visible anywhere. Some kind of an age group exists and the boy-girl distribution is fairly clear, but [these] are limited to terribly high-level things. On this occasion, typical user became a category for the unknown, but perhaps also a wish for regular use patterns. In contrast to the previous quotes, this typical user did not free the developer to do something else, but stayed as an obstacle for design. Creativity in the Developer–User Dialogue While the shapeless average user leaves room for the creativity of the designers, the designers are also careful not to restrict the users in their design. Designing for open-ended use leaves room for creativity among the users. This has been possible due to the kind of product Habbo is: a software product that can be frequently updated without requiring the user to do

145

Johnson

anything. The Habbo development process started out with very frequent release cycles. Even though the product was on-line, during the first year it was continuously updated, sometimes several times a week. Gradually the development process became more complex, as the product grew and more people were involved. A developer describes the transformation from on-line prototyping to robust release processes as follows: Kultakala was done very, very quickly, in a fairly, so to speak, feeling-based way, or, organically. It was up; then we started to fix it, or develop it.... Today we talk about totally different numbers of users, amounts of money. We simply can’t afford to have [a situation where], in tens of countries around the world, everything would stop. (Game Developer 1) This on-line prototyping stage gave the users lots of space to explore the design and take advantages of things not yet considered by the developers. One famous bug that turned into a feature was a way to lift things up in the room and leave them floating. It was possible because the world model was not that well thought through in the early days. Lifting up furniture can seem harmless, but as a piece of furniture is moved out of reach, it becomes a way of destroying furniture. That is a serious threat to the world model and also the business of selling virtual furniture. One developer explained how this user creativity influenced the world model: “You can put things on a table, and tables have a height. But then, we hadn’t programmed that kind of a model; so that when you took the table away, the thing [on the table] stayed in the air. From this followed that the users lifted up furniture with the aid of two stackable objects, by rotating them in peculiar ways. Other people’s things [were lifted] over that box, so high in that room that [they] moved up over the upper edge of the room, which made them gone in practice; they were destroyed. Then we had to make a limit, so no thing could rise up more than to the height of three or something. (Game Developer 1) Some features were developed for a certain use, but the users were able to use the features in their own way to create unexpected phenomena. The teleport is perhaps the prime example, as one developer described it: We had the idea that, we thought the users would start creating homes: There would be a bedroom and a kitchen and that they would jump between these with the teleports, a kind of a door metaphor to the rooms. But they [the users] are much more inventive than we: They made up all those teleport racing games. ... There is this room, two teleports, and two users start running and who comes back first from the other teleport, they must run a race following a route.... And perhaps the first thing that emerged with teleports was the teleport centers, a whole room filled with teleports leading somewhere else, so it was quite a “wow” effect when I first saw it, so “Oh, they did it like that!” We hadn’t thought about someone building link-places for everyone else. (Game Developer 3)

146


On several occasions, many developers tried not to restrict the user too much. GD3 stated that he expects the users to do “unexpected things” with the features he develops, while GD4 noted that, “We create the environment and the building blocks, letting the users exercise their own creativity.” Clearly there is an affinity between the average user of one developer and the user in other developers’ speech. In addition, two particular configurations seemed to be widely known, stabilized, and discussed among the original developers. First, there is the “Easy access, easy play” maxim: The users should be guided step by step into the game environment, without letting any step be too large. One developer explained it like this:“[The] easiest possible login, you can create an character, basic navigation... the first contact to another user, first friend on the friends-list, all sorts of playing, then the own room and decoration, then various groups that you gradually join” (Game Developer 2). Second, there is the “Where else?” maxim: New feature developments in Habbo should be unique and personal. If something exists on some other site already, copying it directly to Habbo is not an option. We have always regarded those [simple feature requests] critically. We want an intentional pursuit for something special in these cases. We don’t make the obvious choices but rather something personal that gives the Habbo world oddity and its own thing, creates its own such persona. (Game Developer 2) This maxim was also called the “Habbo Way” by some developers. However, the “Habbo Way” also denotes the rules listed on the official Web site that new users should approve when becoming habbos. Mixing “I-methodology” and Feedback on Experience Akrich (1992, 1995) argued that successful artifacts depend on the ability of developers to generate user representations and integrate them into their design. She observed many different techniques for creating user representations, both explicit and implicit. The explicit techniques included market surveys and consumer testing, whereas the implicit were the I-methodology, experts, and other products. The term I-methodology exists when the designer puts him/herself in the position of the user and uses his/her own knowledge as a base for design. On the one hand, the concept of the average user could be seen as manifesting Akrich’s I-methodology: When things are uncertain, the designers go with their intuition or feelings based on their own experiences. On the other hand, their knowledge seems to be grounded in their proximity to the users: They can go and look at what’s happening in the hotel and they can read the fansites. The fansites give the developers a way of getting a feeling for what the users want by reading between the lines, as a developer said when reflecting on how often he visits the fansites and which fansites he follows: It depends on the situation. It might be a week or two that you don’t have the time because you know that you’d become immersed totally in the wrong way. It’s HabboForum ... and of course these Finnish ones, from them it is even easier to

147

Johnson

see between the lines the feeling, as they are domestic users: There’s no language [problems] .... (Game Developer 5) The developers draw on their cumulative experience, as they were the users themselves in the early Habbo days, and they have been there since. One developer reflected on the developers’ way of knowing the design history, compared to most users who have not been there as long: An active user probably thinks about it [a new feature: rollers]—this is totally speculation—probably thinks about it in the context of his/her own rooms, what s/he has done so far and how could these [rollers] fit into that, reflecting on when you add a new thing, what more can it give. Whereas we think about it from a little broader perspective—where these and these phenomena have emerged previously and they were based on the teleport, and the dice and these, then reflecting on what if the roller is brought in, what could one do with it, from a different perspective but still.... (Game Developer 4) As a consequence, here in the Habbo case, it becomes hard to tell the difference between I-methodology and feedback from use, since the designers’ reflections repeatedly refer to feedback from experience of use over time. Those who follow the fansites and the activities in the hotel appear as legitimate representatives for the users. The designers balance being representatives for the users, for the business, and for their own interests.

DISCUSSION As a usability designer and researcher, I have been taught that one should never design for the average user. Doing so would probably result in a design that does not fit anyone, because the users are always a heterogeneous group. My first reaction when the game developer started discussing the average user was to shut my ears because it did not make sense. The use of the term average user seemed to contradict HCI guidelines, such as “know the user” or “define the user groups.” It seemed like the user opinions at hand were not followed, and the developer designed for the unknown average user. However, by using a membership categories analysis and extending the timeframe of the analysis, I realized that my initial reaction was wrong. When considering the development history and the developer’s experience of feedback from use, it made more sense to see the developer as a representative of the silent majority of the users. The developer brought in the users who did not voice themselves and denoted them by the category “average user.” In this way, he accomplished his goal of not having to strictly follow the immediate user feedback, which may not reflect the opinions of the majority of silent users, but he also accomplished his goal of considering the previous user feedback based on his experience. The rhetoric about the silent majority and the user is familiar from other contexts, for instance, in politics. The politicians often argue that they themselves represent the (mythical) citizen and give voice to the silent majority of voters. This might, or might not, be the case; it depends on the politician’s own agenda and its proximity to that of the voters. While the citizen

148


as a concept decontextualizes individual human beings, it also makes it possible to talk about the broader concepts of the rights of citizens and human rights. It appears that the concept of the user works in a similar way, giving developers a way of discussing possibilities and restrictions for all users (e.g., in terms of user access rights, user profiles, and groups visible to the computer system). In addition, as some citizens need more empowerment than others, so do some user groups. In this case study, the “average user” was not literally the average user of the Habbo population, but denoted those who needed to be voiced. This study on the developer–user dialogue highlights the importance of the fansites as one user proxy for the Habbo developers. In this case, rather than describing developers as “malevolent manipulators of users,” the case points out their role in balancing and governing different user interests. Digging into the practices of design brings forth its complexity and shows how design and use are intertwined. Over time, the knowledge of the designers is not easily separable into “own” knowledge versus feedback from use, as design and use knowledge becomes mixed. The results of this study can be generalized to user categorization practices in the design of other heterogeneous and complex virtual worlds and communities. If the product or service becomes a commercial success, the sheer number of users makes the user categorization and control more difficult. Designers have to invent categorization strategies of coping with complexity: Fuzzy user configurations can work if there is frequent interaction between designers and users. The virtual world itself and the discussion arenas around the product are important sites for designer–user interaction and sources of user knowledge. Besides, as a company grows, the role of the user representations in the internal communication increases, since not everyone in the company can be tightly involved with the user communities. Implicit user representations, such as gut feelings and design intuition based on long-time experience with users, might need to be made more explicit to function as convincing arguments. Finally, a small comment to the game developers who might ask what they can do about the situation where some user groups are not visible through the monitored fansites. Of course, keeping up with other user feedback channels, like spending time in the virtual world, is important. Another option is to find fansites that attract different user groups. In the Finnish Habbo fansite scene, for instance, the former official fansite Nefekala used to attract a younger age group and many girls. Even though this particular fansite does not exist anymore, new fansites seem to emerge all the time. Following a variety of fansites allows for more diverse user feedback, which can ground the gut feelings of the designers.

ENDNOTES 1. 2.

Habbos are what the Habbo Hotel visitors are commonly called. One becomes a habbo when checking in at the virtual hotel for the first time, then creates one’s own cartoon-like on-line character (avatar). The standard definition of the user as a “person who interacts with the interactive system” (ISO, 2006, p. 3) leaves it open for the researchers and practitioners to contextually agree on whether the user is understood as particular human beings in particular settings, or target groups imagined by the developers. Standard context of use models (ISO, 1998) on the other hand, provide listings of potentially relevant relations: goals, tasks, equipment, and the physical and social environments.

149

Johnson

3. 4. 5.

6. 7. 8.

Details available at http://mc2.soberit.hut.fi For an introduction to the initial differences, and later convergence and extensions to the social shaping versus the social construction of technology approaches, see Sørensen (2002). Images used with permission. ©2006 Sulake Corporation Ltd. HABBO, HABBO HOTEL, SULAKE and associated logos are trademarks or registered trademarks of Sulake Corporation Ltd. in the U.S., the European Union, China and various other jurisdictions. All rights reserved. I left the more exact job description out for anonymity reasons. In Finnish, the hän pronoun is not gender specific and stands for he or she. Actually once, but I interpret normal as a qualitative version of the more statistical average and analyze them together.

REFERENCES Agre, P. (1995). Conceptions of the user in computer systems design. In P. J. Thomas (Ed.), The social and interactional dimensions of human-computer interfaces (67–106). Cambridge, MA, USA: Cambridge University Press. Akrich, M. (1992). The de-scription of technical objects. In W. E. Bijker & J. Law (Eds.), Shaping technology/building society (pp. 205–224). Cambridge, MA, USA: MIT Press. Akrich, M. (1995). User representations: Practices, methods and sociology. In A. Rip, T. J. Misa, & J. Schot (Eds.), Managing technology in society (pp. 167–184). London: Pinter Publishers. Akrich, M., & Latour, B. (1992). A summary of a convenient vocabulary for the semiotics of human and nonhuman assemblies. In W. E. Bijker & J. Law (Eds.), Shaping technology/building society (pp. 259– 264). Cambridge, MA, USA: MIT Press. Andersen, P. B. (1999). Elastic interfaces: Maritime instrumentation as an example. In J. M. Hoc, P. Millot, E. Hollnagel, & P. C. Cacciabue (Eds.), Proceedings of the 7th Conference on Cognitive Science Approaches to Process Control (CSAPC’99; pp. 35-41). Valenciennes, France: Presses Universitaires de Valenciennes. Baker, C. (2004). Membership categorization and interview accounts. In D. Silverman (Ed.), Qualitative research: Theory, methods and practice (pp. 162 –176). London: SAGE. Barad, K. (2003). Posthumanist performativity: Toward an understanding of how matter comes to matter. Signs: Journal of Women in Culture and Society, 28, 801–831. Beyer, H., & Holtzblatt K. (1998). Contextual design: Defining customer-centered systems. San Francisco: Morgan Kaufmann. Bijker, W. E., Hughes, T. P., & Pinch, T. (1987). The social construction of technological systems: New directions in the sociology and history of technology. Cambridge, MA, USA: MIT Press. Blythe, M., Overbeeke, K., Monk, A., & Wright, P. (2003). Funology: From usability to enjoyment. Amsterdam: Kluwer. Bødker, S. (1998). Understanding representations in design. Human Computer Interaction, 13, 107–125. Bødker, S., Ehn, P., Kyng, M., Kammersgaard, J., & Sundblad, Y. (1987). A utopian experience: On design of powerful computer-based tools for skilled graphic workers. In G. Bjerknes, P. Ehn, & M. Kyng (Eds.), Computers and democracy (pp. 251–278). Aldershot, UK: Avebury. Budde, N. F. (2004). There’s no such thing as an “average” user. Interactions, 11(2), 54. Carmel, E., Whitaker, R. D., & George, J. F. (1993). PD and joint application design: A transatlantic comparison. Communications of the ACM, 36(6), 40–48. Checkland, P., & Scholes, J. (1981). Soft-systems methodology in action. Chichester, UK: John Wiley & Sons. Cooper, A., & Reimann, R. (2003). About Face 2.0: The essentials of interaction design. Indianapolis, IN, USA: Wiley Publishing, Inc.

150


Courage, C., & Baxter, K. (2005). Understanding your users: A practical guide to user requirements methods, tools, and techniques. Amsterdam: Morgan Kaufmann. Dagwell, R., & Weber, R. (1983). System designers’ user models: A comparative study and methodological critique. Communications of the ACM, 26, 987–997. Ehn, P. (1988). Work-oriented design of computer artifacts. Stockholm, Sweden: Arbetslivscentrum. Friedman, A. L., & Cornford. D. S. (1989). Computer systems development: History, organization and implementation. Chichester, UK: John Wiley & Sons. Greenbaum, J., & Kyng, M. (1991) Design at work: cooperative design of computer systems. Hillsdale, NJ, USA: Lawrence Erlbaum Associates Publishers. Grudin, J. (1993). Interface: An evolving concept. Communications of the ACM, 36(4), 110–119. Hackos, J. T., & Redish, J. C. (1998). User and task analysis for interface design. New York: Wiley. Hales, M. (1994). Where are designers? Styles of design practice, objects of design and views of users in CSCW. In D. Rosenberg & C. Hutchinson (Eds.), Design issues in CSCW (pp. 151–178). London: Springer-Verlag. Hyysalo, S. (2004). Uses of innovation: Wristcare in the practices of engineers and elderly. Helsinki, Finland: Helsinki University Press. International Organization for Standardization [ISO]. (1998, March). Ergonomic requirements for office work with visual display terminals, Part 11: Guidance on Usability. (Standards No. 9241-11). Geneva, Switzerland: ISO. International Organization for Standardization [ISO]. (1999, June). Human-Centred Design Processes for Interactive Systems. (Standards No. 13407). Geneva, Switzerland: ISO. International Organization for Standardization [ISO]. (2006, April). Ergonomic requirements for office work with visual display terminals, Part 110: Dialogue principles. (Standards No. 9241-110). Geneva, Switzerland: ISO. Isomäki, H. (2002). The prevailing conceptions of the human being in information systems development: System designers’ reflections. (Doctoral thesis, University of Tampere). Tampere, Finland: University of Tampere. Johnson, M. (2006a, March). Uses of history in the design of Habbo Hotel. Paper presented at the Proact2006 Conference: Innovation pressure; Rethinking competitiveness, policy and the society in a globalised economy, Tampere, Finland. Johnson, M. (2006b, March). Categorising users in the design of Habbo Hotel. Paper presented at the Proact2006 Conference: Innovation pressure; Rethinking competitiveness, policy and the society in a globalised economy, Tampere, Finland. Johnson, M. (2006c, May). Visiotypes in the membership categorisation practices of Habbo Hotel. Presentation made at the Public Imageries Today conference, co-organized by the Finnish Museum for Photography and the New Media School, Karlsruhe, Germany. Johnson, M., & Toiskallio, K. (2005, August). Fansites as sources for user research: Case Habbo Hotel. Paper presented at the 28th Information Systems Research Seminar in Scandinavia (IRIS´28), Kristiansand, Norway. Johnson, M., & Toiskallio, K. (in press) “Who are the Habbo Hotel users?” In M. Turpeinen (Ed.), Mobile content communities final report. Helsinki, Finland: Helsinki Institute for Information Technology. Kay, A., & Goldberg. A. (1977). Personal dynamic media. Computer, 10(3), 31–41. Kling, R., & Gerson, E. M. (1977). The social dynamics of technical innovation in the computing world. Symbolic Interaction, 1(1), 132–146. Kuniavsky, M. (2005). Observing the user experience: A practitioner’s guide to user research. San Francisco: Morgan Kaufmann. Kuutti, K. (2001). Hunting for the lost user: From sources of errors to active actors and beyond. Cultural Usability seminar, Media Lab, University of Art and Design Helsinki. Retrieved December 31, 2006, from http://www.mlab.uiah.fi/culturalusability/papers/Kuutti_paper.html

151

Johnson

Lamb, R., & Kling, R. (2003). Reconceptualizing users as social actors in information systems research. MIS Quarterly, 27, 197–235. Latour, B. (1992). Where are the missing masses? The sociology of a few mundane artifacts. In W. E. Bijker & J. Law (Eds.), Shaping technology/building society (pp. 225–258). Cambridge, MA, USA: MIT Press. Laukkanen, T. (2005). Modding scenes: Introduction to user-created content in computer gaming (Hypermedia Laboratory Net Series, 9). Tampere, Finland: Tampere University Press. Lee, A. S., Liebenau, J., & DeGross, J. (1997). Information systems and qualitative research. London: Chapman & Hall. Mackay, H., Carne, C., Beynon-Davies, P., & Tudhope, D. (2000). Reconfiguring the user: Using rapid application development. Social Studies of Science, 30, 737–757. MacKenzie, D., & Wajcman, J. (Eds.). (1985). The social shaping of technology: How the refrigerator got its hum. Milton Keynes, UK: Open University Press. Maguire, M. (2001). Context of use within usability activities. International Journal of Human-Computer Studies, 55, 453–483. Mørch, A. I. (1997). Method and tools for tailoring of object-oriented applications: An evolving artifacts approach. Unpublished doctoral dissertation, University of Oslo, Norway. Muller, M., Millen, D. R., & Strohecker, C. (2001). What makes a representative user representative? A Participatory Poster. In M. (Mantei) Tremaine (Ed.), CHI '01 Extended abstracts on human factors in computing systems (pp. 101–102). New York: ACM. Mumford, E. (1983). Designing human systems for new technology: The ETHICS approach. Manchester, UK: Manchester Business School. Nardi, B. (1993). A small matter of programming: Perspectives on end user computing. Cambridge, MA, USA: MIT Press. Norman, D. (1988). The psychology of everyday things. New York: Basic Books. Norman, D., & Draper, S. (1986). User centered system design: New perspectives on human-computer interaction. Hillsdale, NJ, USA: Lawrence Erlbaum Associates. Nurminen, M. I. (1988). People or computers: Three ways of looking at information systems. Lund, Sweden: Studentlitteratur & Chartwell-Bratt. Oudshoorn, N., & Pinch, T. (2003). How users matter: The co-construction of users and technologies. Cambridge, MA, USA: MIT Press. Porter, C. E. (2004). A typology of virtual communities: A multi-disciplinary foundation for future research. Journal of Computer-Mediated Communication, 10(1). Retrieved on December 31, 2006 from http://jcmc.indiana.edu/vol10/issue1/porter.html Schuler, D., & Namioka, A. (1993). Participatory design: Principles and practices. Mahwah, NJ, USA: Lawrence Erlbaum Associates. Shneiderman, B. (1998). Designing the user interface. Reading, MA, USA: Addison-Wesley. Silverman, D. (2004). Qualitative research: Theory, method and practice. London: Sage. Sørensen, K. H. (2002). Social shaping on the move. In K. H. Sørensen & R. Williams (Eds.), Shaping technology, guiding policy: Concepts, spaces and tools (pp. 19–35). Cheltenham, UK: Edward Elgar. Spillers, F. (2006). Designing for the “average user”. Retrieved from the April 21, 2006 blog Demystifying Usability (www.usabilitydiary.com) on December 31, 2006, http://experiencedynamics.blogs.com/site_search_usability/2006/04/designing_for_t.html Stewart, J., & Williams, R. (2005). The wrong trousers? Beyond the design fallacy: Social learning and the user. Reprinted in D. Howcroft & E. Trauth (Eds.), Handbook of critical information systems research: Theory and application (pp. 195–221). Cheltenham, UK: Edward Elgar. Suchman, L. (1995). Making work visible. Communication of the ACM, 38(9), 56–61, 63-64.

152


Sulake (2006a). Habbo Hotel. Retrieved December 31, 2006, from http://www.habbohotel.co.uk/ Sulake (2006b). Press Room, Press Releases. Retrieved December 31, 2006, from http://www.sulake.fi/pressroom.html Thomas, P. J. (Ed.). (1995). The social and interactional dimensions of human-computer interfaces. Cambridge, UK: Cambridge University Press. Toiskallio, K., Tamminen, S., Korpilahti, H., Hari, S., & Nieminen, M. (2004). Mobiilit Käyttökontekstit: Mobix Loppuraportti [Mobile contexts of use: Mobix final report]. (Technical report HUT-SoberIT-C8). Helsinki, Finland: Software Business and Engineering Institute. Turpeinen, M. & Kuikkaniemi, K. (Eds.). (2007). Mobile content communities final report. (HIIT Publications 2007-1). Helsinki, Finland: Helsinki Institute for Information Technology. Westrup, C. (1997). Constituting users in requirements techniques. In A. S. Lee, J. Liebenau, & J. I. DeGross (Eds.), Information systems and qualitative research (pp. 183–204). London: Chapman & Hall. Williams, R., Stewart, J., & Slack, R. (2005). Social learning in technological innovation: Experimenting with information and communication technologies. Cheltenham, UK: Edvard Elgar. Woolgar, S. (1991). Configuring the user: The case of usability trials. In J. Law (Ed.), A sociology of monsters: Essays on power, technology and domination (pp. 57–102). London: Routledge. Yin, R. K. (1994). Case study research, design and methods (2nd ed.). Newbury Park, CA, USA: Sage Publications.

Author’s Note All correspondence should be addressed to: Mikael Johnson Helsinki Institute for Information Technology PO Box 9800 02015 TKK [email protected] Human Technology: An Interdisciplinary Journal on Humans in ICT Environments ISSN 1795-6889 www.humantechnology.jyu.fi

153

ARTICLE

IV


Vili Lehdonvirta, Terhi-Anna Wilska, and Mikael Johnson (2009). "Virtual Consumerism: Case Habbo Hotel". Information, Communication and Society 12 (7), 1059–1079. Reprinted with permission.

239

Vili Lehdonvirta, Terhi-Anna Wilska & Mikael Johnson VIRTUAL CONSUMERISM

Downloaded By: [Teknillinen Korkeakoulu] At: 02:26 14 May 2010

Case Habbo Hotel

Selling virtual items for real money is increasingly being used as a revenue model in games and other online services. To some parents and authorities, this has been a shock: previously innocuous ‘consumption games’ suddenly seem to be enticing players into giving away their money for nothing. In this article, we examine the phenomenon from a sociological perspective, aiming to understand how some media representations come to be perceived as ‘virtual commodities’, what motivations individuals have for spending money on these commodities, and how the resulting ‘virtual consumerism’ relates to consumer culture at large. The discussion is based on a study of everyday practices and culture in Habbo Hotel, a popular massively-multiuser online environment permeated with virtual items. Our results suggest that virtual commodities can act in essentially the same social roles as material goods, leading us to ask whether ecologically sustainable virtual consumption could be a substitute to material consumerism in the future. Keywords virtual property; consumer behaviour; commodification; global culture industry; massively-multiplayer online game (MMO); realmoney trading (RMT)

Introduction For at least two decades, simulated shopping and commodity consumption have been central elements of play in many digital games and online hangouts. Elements of this central part of contemporary Western culture are incorporated into all sorts of settings, including even fantasy worlds, historical milieus, and futuristic visions. Players or participants, often young, are familiar with the logic of consumption and ownership and gladly engage in the simulated consumption games offered by the systems. What controversy there has been over activities portrayed in video games has centred on simulated sex and violence, rarely shopping. In recent years, however, it has become increasingly common for virtual goods circulated in consumption games to be exchangeable for real money. Information, Communication & Society Vol. 12, No. 7, October 2009, pp. 1059 –1079 ISSN 1369-118X print/ISSN 1468-4462 online # 2009 Taylor & Francis http://www.tandf.co.uk/journals DOI: 10.1080/13691180802587813


1060

INFORMATION, COMMUNICATION & SOCIETY

Using a credit card or mobile phone, players are now able to purchase virtual items, clothes and characters like any commodities in an online store, except that the goods are never delivered to the physical doorstep. This ‘virtual item trade’ or ‘real-money trade of virtual property’, as it is variously known, has forced a re-evaluation of the status of fantastical consumption play. Economists (e.g. Castronova 2006; Huhh 2008) have observed that what were previously considered fiction can actually be analysed as goods in the economic sense. Legal scholars (e.g. Fairfield 2005; Lastowka & Hunter 2004) have put forward questions regarding the ownership and legal status of virtual assets. In popular media, the phenomenon has been welcomed with a sort of bemused wonder, at times also with great controversy. From the point of view of parents of young players, it may seem as if the previously innocuous digital hangouts are suddenly enticing gullible children, not yet able to distinguish between real and make-believe, into giving away their money for nothing. In Finland and Sweden, consumer protection authorities have been called on more than once. For authorities and regulators, there is a distinct lack of understanding on how to categorise, deal with or even approach the phenomenon. The purpose of this article is to offer a re-evaluation of the status of online consumption games from a sociological perspective. Using a case study of Habbo Hotel, a popular ‘virtual world’ for teenagers, we descend into the everyday practices and culture of a social world permeated with virtual goods. We drop assumptions regarding the nature of the artifact as a digital game, and instead adopt angles of interpretation from the sociology of consumption, treating the phenomenon as virtual consumption. This way, we aim to clarify the motivations of those who participate in the system and spend money on it, explain some of the social dynamics that shape virtual consumer behaviour, and offer a perspective on how virtual consumerism fits in the bigger picture of consumer society. We begin by providing some context on consumption games and virtual asset trade. This is followed by a theoretical discussion of the notion of virtual commodities, where we compare them to digital information. Next, we turn to sociological literature on consumption, reviewing three main perspectives on consumer culture. Armed with these perspectives, we reach into the case material. In the last section, we summarise our findings, present conclusions and briefly consider the future of virtual consumerism: to what extent could it substitute material consumption, and what would be the ecological implications?

Consumption games Habitat, opened in 1985, was an early pioneer of virtual environments. It used two-dimensional graphics to represent spaces such as home, hotel and arcade, where cartoon-style characters controlled by users could talk and interact. The


VIRTUAL CONSUMERISM

landscape was scattered with vending machines from which users could purchase virtual items ranging from weapons to furniture. Purchases were paid with a currency called Tokens, which was distributed to the users for free. The most desirable items were spare heads that could be used to customise one’s character (Dibbell 1998, p. 172). More recently, shopping was placed in a central role in a series of single-player video games called The Sims. Described as ‘life simulators’, they became one of the most successful game series ever. In both Habitat and The Sims, shopping is conceived of as simulation: ‘consumption play’. In so-called massively-multiplayer online role-playing games (MMORPG) launched in the late 1990s, consumption play began to be mixed up with real money (Lehdonvirta 2005). In Ultima Online and EverQuest, hundreds of thousands of players ‘traded’ with other players to exchange game assets accumulated during months of play for other game assets. As with the previous systems, the economy was intended to be like Monopoly: no real money would change hands. But in 1999, some players put their game assets on auction at eBay. Perhaps surprisingly, they received bids from other players. When an auction was completed, payment was carried out using ordinary means such as cheque or money order. The two players then met up in the game and the seller handed the auctioned object to the buyer. This way, an exchange value measured in US dollars could soon be observed for virtual goods ranging from castles to gold nuggets (Lehdonvirta 2005). In 2002, a massively multiplayer online version was created of The Sims, and real-money trading followed. What began as trading between individual players quickly became a greymarket commercial activity. Jin (2006) describes how hundreds of ‘gaming sweatshops’ in China hire young men to play online games in 12 hour shifts to harvest game assets. Virtual property dealer companies such as IGE buy these assets and sell them to players, often Western, for significant profit. In the MMO Second Life, real-money commerce also involves services such as design and construction consultancy for virtual landowners. Today, the idea of selling virtual commodities for real money has spread beyond its origins in gaming and found its way to other online services, particularly those where social interaction is a key element. Finnish online image gallery IRC-Galleria, Korean social networking site Cyworld, Chinese instant messaging service Tencent QQ and US social networking site Facebook are examples of extremely popular services that earn revenues by selling virtual commodities to their users. The commodities are often small graphical objects that can be used to decorate one’s online profile or given to other users as gifts. Common objects are priced at a dollar or less, while notable ones can fetch tens of Euros. This spending opportunity hardly existed five years ago, but is now increasingly pervasive. Virtual asset trade has begun to make occasional appearances in sociological and media studies texts that deal with sites where virtual assets are found. Game

1061


1062


scholar Taylor (2006) discusses real-money trading in her book on the culture of the MMORPG EverQuest, while Fung (2006) mentions them in a study of online gaming communities in Hong Kong. Grimes (2007) and Herman et al. (2006) discuss virtual commodities in the context of debates over intellectual property rights, while Malaby (2006) considers virtual assets as part of a Bourdieuan capital framework. Virtual asset sales are frequently referenced in discussions concerning demarcations between ‘real’ and ‘virtual’ worlds, as they appear to transcend this (artificial) dichotomy. In the works referenced above, the authors also make valuable starts into trying to understand the role of virtual commodities in the users’ everyday life. Yamaguchi (2004) has compared game economies to so-called local exchange trading systems (LETS), small alternative economies that exist alongside the formal economy. What they have in common is the limited scope of possible transactions. But while LETS and similar systems focus on using new mechanisms to exchange familiar goods, the most interesting aspect of the present phenomenon in our opinion is the use of established markets like eBay to exchange a completely new category of goods.

From digital representations to commodities Our intention is to go deeper into understanding virtual goods by examining the phenomenon as a type of consumer behaviour. But can real-money trade of virtual assets be considered consumption in the first place? The literal meaning of ‘consumption’ is using up, destroying or eating something, which indicates that in economic terms, objects have life cycles: first they are produced, then exchanged with money, and finally consumed, until they disappear or fade and lose their value (Wilk 2004, pp. 15– 17). In the present phenomenon, nothing is consumed in the sense of something tangible being destroyed, expended, used up, worn down or eaten.1 But even with regard to traditional consumption, economic metaphors are not always appropriate. Objects do not necessarily lose their value when used (e.g. antique, jewelry or collectibles), and they can be used several times. The value of goods may be based on nonexisting properties (e.g. stock markets). Many objects also have different ‘social lives’, which means that their use may change over time (Appadurai 1986; Douglas & Isherwood 1978; Wilk 2004). As practical quantitative measures of consumption, researchers observe the spending of time and money. However, time and money can be spent on media consumption as well as on commodities, and the sociology of consumption has developed very different approaches to these two. Media is intangible, superstructural (Lash & Lury 2007), while commodities are used to create distinctions through ownership. Computer screens showing digital representations of goods would normally be analysed as media surfaces, but our contention is that virtual assets should


VIRTUAL CONSUMERISM

rather be analysed as commodities. This is due to a key ‘innovation’ that enables digital representations to take up roles commonly associated with luxury goods and other tangible commodities: artificial scarcity (Castronova 2006; see also rivalrousness in Fairfield 2005). Since the cost of reproducing digital objects is close to zero, there is little technical reason why every participant in a virtual economy could not be given a copy of each and every commodity. Indeed, armchair theorists of the 1990s predicted that cyberspace would liberate us from the restrictions of the physical world in just this way. Yet in some circumstances people seem to prefer virtual scarcity over digital abundance. Rivalrousness and scarcity can transform digital representations from media messages into objects with a ‘thing-like’ nature, in a process that resembles Marxist theories of commodification (see Lury 1996, p. 41; Miller 1987, p. 13). This allows attributes that in reality exist by virtue of social relationships between people to be perceived as inherent characteristics of digital objects. One such attribute is economic value. This thingification of representations into virtual commodities meshes particularly well with Scott Lash and Celia Lury’s recent notion of a ‘global culture industry’ (Lash & Lury 2007). Though critics of mass culture such as Horkheimer and Adorno (1976) have long seen culture as ‘industrialised’ and assimilated into market capitalism, culture used to remain in the superstructure. Culture and media were a utility, a means to promote the consumption of commodities, which was the end purpose. But ‘[i]n global culture industry, what were previously media become things’ (Lash & Lury 2007, p. 8). Media representations materialise as tangible commodities and commodities simultaneously become media. And ‘When media become things [. . .] they no longer exclusively have cultural value. They come very importantly to have use-value and exchange-value’ (Lash & Lury 2007, p. 8), completing the true industrialisation of culture (and also the culturisation of industry).2 The commodification outlined above is a clear point of difference between services based on virtual commodities and other ‘social web’ services that emphasise user involvement and social interaction. For example, by publishing a prominent video clip on YouTube, a user might gain a reputation among fellow users, but that reputation is not something that can subsequently be detached and sold as an attribute of the video to another user. A commodified object, in contrast, can obtain its own history and reputation in its life-course, independent of its maker and independent of other objects (copies) of the same type. In this aspect, services based on virtual commodities actually bear more similarity to old-fashioned online marketplaces like eBay that sell ‘naturally scarce’ consumer commodities than to cutting-edge social web platforms that distribute media content. Ideologically, virtual consumption represents a return to twentieth century material consumerism, which the ‘new media’ and the ‘Web 2.0’ paradigm with its emphasis on the amateur and the ethos of sharing were trying to subvert.3

1063

1064



Understanding virtual consumption How do digital representations that have transformed into quasi-tangible commodities become so attractive that users are willing to spend money on them? It is clear that materialistic explanations are difficult to reconcile with the digital nature of the objects in question. In the sociology of consumption, there has thankfully long been a ‘movement away from regarding goods merely as utilities having a use-value and an exchange-value which can be related to some fixed system of human needs’ (Featherstone 1991, p. 85). In his book Consumer Culture & Postmodernism, Featherstone (1991) identifies three main perspectives on consumer culture. First is the Marxist ‘production of consumption’ perspective, which posits that in a capitalist society, mass media and other means are used to ‘educate’ people into becoming consumers in order to open up new markets for the ever-expanding production sector (Miller 1987). As mentioned above, this is the way in which virtual consumption is currently perceived by, for instance, concerned parents. It entails a view of consumers as victims that the manipulative capitalist is luring away from some better, ‘more original’ set of values and social relations. For instance, Horkheimer and Adorno (1976) express concern over the replacement of high culture with a homogenous ‘mass culture’ produced by a ‘culture industry’. However, the ‘production of consumption’ view has been criticised as elitist and for the fact that it fails to examine actual consumption processes, which reveal diverse audience responses and complex, unexpected uses of goods. The second perspective goes deeper into practice, focusing on the use of goods as tools for building social bonds or distinctions. Generally speaking, consumers are seen as communicators who use symbolic meanings embedded in commodities to express status, class, group membership, difference or selfidentity (e.g. Bourdieu 1984; McCracken 1988; Simmel 1904/1957; Veblen 1899/1955). The satisfaction derived from goods is primarily linked to their use as markers, and only secondarily related to their physical consumption (Douglas & Isherwood 1978). An extreme example of this is the accumulation of collectible objects, which can be completely ‘useless’ and non-functional (Baudrillard 1994, 1996). According to Belk (1995, 2004), even if the collected objects were once useful in some way, when they enter the collection they are no longer used in their original purpose. Although collectors frequently describe the thrill of the hunt for collectible objects, Belk (1995) argues that the hunt is, in the end, usually a highly competitive game of status seeking. On the other hand, Belk (2004) acknowledges that the collector’s singleminded pursuit may also be seen as highly pleasurable romanticism; noble saving of objects that few others appreciate. This position echoes a third perspective on consumer culture, described by, e.g. Campbell (1987, 2004) and Featherstone (1991), which focuses on the individual emotional pleasures of consumption. Campbell sees consumption as an individual, hedonistic


VIRTUAL CONSUMERISM

process. Featherstone argues that urban everyday life in particular is aestheticised: overflowing with imagery (though not necessarily overcome by it as for Baudrillard 1971/1988) that can evoke dream-like and pleasurable aesthetic sensations. For an individual, contemporary consumer culture also entails the creative mixing of consumption styles in a project that resembles artistic expression. Does virtual consumption involve the same motives as those detected in other modes of consumption? Perhaps the only study so far to approach the question is a recent article by Yue Guo and Stuart Barnes, titled ‘Why People Buy Virtual Items in Virtual Worlds with Real Money’ (Guo & Barnes 2007). The goal of the article is to develop a preliminary model of ‘individual determinants for the decision to purchase virtual items within virtual game communities.’ The model is largely based on models intended to describe user acceptance of new technologies in management and information systems sciences. Thus, the determinants it identifies include ease of use, ‘performance expectancy’ and ‘perceived enjoyment’. The only social determinant is ‘the degree to which an individual perceives that important others believe he or she should use community transaction mechanisms to gain high-level virtual items for enhancing character competence’ (Guo & Barnes 2007, p. 72). The result is unfortunately a rather limited utilitarian view of user motivations, which furthermore makes strong assumptions regarding the mechanics of the underlying service that seem to confine its application to certain MMORPGs only. In this article, we attempt to start towards a richer, more inclusive and theoretically better motivated understanding of virtual consumption by taking the established perspectives from consumer sociology outlined above as our interpretative frame and descending into the everyday practices and culture of a social world permeated with virtual items.

Case Habbo Hotel One prominent site of virtual consumption is Habbo Hotel, a popular massively multi-user online environment that is not quite a game, yet not simply a graphical chat program. The environment resembles a giant contemporary Western indoor space, presented in isometric ‘retro style’ three-dimensional graphics and populated by blocky avatars, each controlled by a user. The service was launched in 2000 by a Finnish company called Sulake, and a localised version is available in 32 countries. Habbo is targeted at teenagers: the average age of users is 15.5 years and their gender distribution is close to equal (Sulake 2008a). According to Sulake (2008b), Habbo currently has around 9.5 million monthly unique visitors globally, which makes it one of the most popular ‘virtual world’ type services, about ten times as popular as Second Life. Based on Sulake’s figures, it is possible to estimate that more than a quarter of

1065


1066


13– 18-year-olds in Finland use Habbo Hotel.4 In 2006, Habbo earned approximately 55 million US$, mostly from microtransactions (Hyatt 2007). Due to its maturity as an internet service, its popularity, and its strong focus on virtual commodities and real-money purchases, we believe Habbo Hotel is an excellent subject for an explorative study into virtual consumerism. Teenagers are often in the forefront of material consumer culture, and can be expected to provide a similarly rich view of virtual consumerism. In the following sections, we provide descriptions that are based on a case data that was collected in 2004 –2007 and consists of user interviews (N ¼ 12), studies of Habbo fansites (N ¼ 24), and first-hand use experience (‘online fieldwork’ similar to Slater & Miller 2000). Unless otherwise noted, the data described here pertains to the Finnish version of the service. Quotes were translated from Finnish into English by the authors. By Habbo fansites, we refer to websites maintained by users independently of Sulake that contain articles and discussion on Habbo Hotel and activities taking place inside it. Johnson and Toiskallio (2005) argue that fansites are a valuable source for user research, especially as a complement to other fieldwork methods, as they reveal lore and practices that could otherwise take months of participation to discover. It should be noted that since the data was collected, Sulake has extended the web aspect of the Habbo platform by introducing individual user homepages, group homepages and group discussion forums, as well as social networking style features. These new features have taken over some of the functions previously carried out by fansites. The combination of Habbo Hotel with the web-based features has been rebranded as simply Habbo.

Aesthetics, self-expression and identity Upon first entering Habbo Hotel with a web browser, a new user is requested to create an avatar, one’s digital representation inside the hotel. The avatar can resemble either a girl or a boy and is customised by choosing a set of clothes, shoes, a hairstyle, a face and a skin colour. The wide range of options includes T-shirts, sweaters, jackets, skirts, sneakers, baseball caps, as well as quirkier items such as rabbit costumes and Afro haircuts. After creating an avatar and walking around in the hotel’s corridors and cafe´s, the next thing a new user often does is creating his or her own ‘guest room’. Guest rooms can be private or open to visitors. They are initially empty, but can be decorated with a wide range of virtual items and furniture, from sofa sets and lava lamps to DJ decks and rubber ducks. This selection of appearance, clothing, items and furniture sets the landscape for virtual consumption choices in Habbo Hotel. Given the practically limitless number of possible combinations, how do users choose what to consume? When discussing consumption choices, users commonly point to selfexpression and aesthetic considerations as the motivating factors for their choices. User ‘Kanahaukka’ writes on fansite Habbomatic.com:

VIRTUAL CONSUMERISM


Regarding my current style, the reason I wear those colors is that the yellow hoodie is close to my favorite color, which is golden yellow. White goes with all clothes, that’s why the bottom part is white. Hair also yellow, same reason as the hoodie. IRL [in real life] I don’t like to wear shorts and a hoodie at the same time, but in Habbo it somehow pleases my eye. In room decoration, prevalent aesthetics are reflected in monthly decoration competitions run by Sulake, in which users vote for a favourite room from a set of candidates. Such rooms are popular destinations for visitors who enjoy the view (Figure 1). Even items that have a functional purpose are appropriated for aesthetic purposes. For example, Sulake has created the Habbo equivalent of Post-it notes for attaching notes to walls. Instead of using them to relay information, users sometimes arrange them into decorative patterns. Functional considerations are important in rooms that seek to cater for activities such as games or to accommodate a large number of characters, but even then, users simultaneously strive to make the arrangement aesthetically pleasing or even ‘artistic’. Sometimes consumption choices are used to express a specific role. ‘Myself, I am a Gothic witch,’ writes user ‘Scekti’ in the same discussion thread as Kanahaukka above. Scekti has a white face, black hair and black clothes. Others pretend to have a ‘profession’. This involves not only clothing one’s avatar accordingly and behaving in an appropriate manner, but also decorating a room or a set of rooms in a corresponding fashion. For example, one user has taken the role of a doctor. He dresses up in a white coat and has built a hospital ward by decorating a room with white tiles, hospital beds and a Swiss flag – the flag resembling an

FIGURE 1

Winning entry of the April 2007 decoration competition.

1067


1068


inverted red cross. This type of creative repurposing is common, as it widens the range of expression (see also Johnson & Toiskallio 2007). While many develop individualistic styles, virtual consumption behaviour is also a way to identify with a subculture. Habbo’s ‘Goths’ are a good example, with their preference for black clothes, tall candles and gloomy furniture. Some of those who identify with Goths believe that being one simply consists of making the appropriate consumption choices, while others contend that it also entails various beliefs or values. In any case, Goths are not a group or organisation with a defined membership, nor is the Goth identity based on traditional socio-economic or geographical divisions. The one thing that most Goths have in common is their shopping behaviour. Table 1 shows a selection of Habbo clothing styles, as identified by a fansite author. Other categorisations are constantly put forward and debated. Users also form groups of varying degrees of organisation and longevity, styling themselves as, e.g. ‘gangs’, ‘militaries’ and ‘model agencies’. The groups often have a set of rules that specify member ranks and membership criteria, such as required items and appearance. For example, one ‘military’ organisation requires its members to wear green tops, other pieces of clothing indicating rank. A gang calling itself ‘Egyptian Mafia’ instructs new recruits to label themselves with the text ‘E.M. rank: slave’ and wear brown clothes. Failing to conform to the dress code may result in expulsion from the organisation. For a price of one Plastyk, members may purchase a right to be exempt from the dress code.

Luxury and social status Inside each Habbo Hotel there is a ‘virtual economy’ where items are traded on markets and prices determined as a function of supply and demand. The original source of all the commodities circulating in the economy (the ‘producer’) is TABLE 1

Common clothing styles according to fansite Kriisipalvelu.net (2005).

Punk Bright colours, irokese

Gothic White skin, dark clothes

Teinix (teenie) Pastel colours, round cheeks

Wannabe Smart dress, stiff behaviour

Own style Does not care about fashions


VIRTUAL CONSUMERISM

Sulake. Sulake provides an online catalogue where new virtual items can be bought at a fixed price. The items are paid for using the official currency known as credits, which must first be purchased from Sulake. In Finland, each credit costs approximately 17 eurocents, depending on payment method. As an example of the catalogue’s price level, the aforementioned Plastyk chair costs five credits (E0.85), while a wood-panelled ‘Tubmaster’ jacuzzi goes for 50 credits (E8.50). Although Sulake’s catalogue provides a perfectly elastic supply of many items, there are many items for which the supply is limited. Seasonal items such as Christmas trees are available in the catalogue only at specific times of the year. ‘Collectibles’ are available only for a limited period of time after their introduction, usually for two weeks but sometimes for only two hours. A special item is sometimes made available in a strictly limited quantity as part of a promotion or a competition. In Habbo slang, such items are known as ‘super rares’. Items that are not available in the catalogue can always be bought or bartered from other users, but their prices tend to go up as the supply grows thin. For example, a limited number of DJ style record players were distributed by Sulake in 2002 as part of an advertising campaign for a mobile operator. In 2006, users were trading them for around 250 Plastyk, which equals a re-purchase cost of approximately E200. Even though the purpose of virtual items in Habbo Hotel is ostensibly aesthetic, something to decorate rooms with, in practice they are simultaneously, perhaps even primarily, pieces in a status game. User ‘Roundact’ on fansite Habbomatic.com writes: What stuff do you have in Habbo? Are you rich? I have 12 moccas, 2 typewriters (genuine), 3 DJ Esko posters, 4 dice, 6 typewriter-sets and some random stuff. Possessions are constantly discussed and compared both inside the hotel and on fansites. Those with lots of rare and valuable items are labelled ‘rich’. Rich users do not necessarily use their possessions for anything; they may simply line them up on the floor, as in Figure 2. The expensive record player is actually not even capable of playing music. However, changes in market conditions can affect the value and status of items. Since mid-2006, the record player was again distributed as part of a new promotion. This multiplied its supply on the user-to-user marketplaces, leading to a drastic decrease in its value and the prestige associated with owning one. Besides scarcity, other factors that can contribute to an item’s value are age, previous owners and how the item was originally acquired. For example, a decorative trophy that was originally given as a prize to a well-known ‘celebrity’ avatar (and still has that avatar’s name inscribed on it) is valued many times higher than a visually identical new trophy that lacks such history. What difference does it make whether one is rich or not? For example, some popular venues

1069


1070


FIGURE 2

A room with valuable items, including a record player.

(e.g. casinos, dating games, trading houses) only welcome avatars of sufficient wealth. After losing his property to a scammer, user ‘Hpsaea’ wrote in HabboSanomat: [Losing the items] was not the worst of it, though. I was broke. Literally in the blink of an eye, I was demoted to the lowest value class in the eyes of others. Besides property ownership, there is another method sanctioned by Sulake for establishing status group differences. ‘Habbo Club is a club for all active habbos who wish to gain access to many different wonderful privileges and rights. One month’s membership costs 25 credits [E4.17],’ reads the service’s description. Benefits include skipping in lines when entering crowded areas and having a larger and more varied selection of clothes to choose from. An avatar’s status as a member of Habbo Club is indicated by a badge with the letters ‘HC’, and is also often evident from the avatar’s clothing style. Some HC members routinely discriminate against non-members. For example, they might not speak to nonmembers and may label their rooms as ‘HC only’, indicating that non-members are not welcome and may be thrown out. One member explained that this is because non-members can potentially be ‘throw-away avatars’ used in fraud or harassment, unlike HC avatars, which are financially committed. But those who remain non-members for financial or other reasons complain about the discrimination. HC member ‘Simpucca’ writes at Habbomatic.com:

VIRTUAL CONSUMERISM

Sometimes in a beauty pageant you may have encountered a situation where a non-HC says to a HC, ‘you are sure to win because you are a HC’ or something similar. This is not nice. Winning does not depend on HC only. Personally, I have seen ordinary characters win many times.


Goods as vehicles of arbitrary meaning Being extremely rich, hosting an exceptionally popular venue or activity, producing a popular fansite, or being a Sulake employee may result in one becoming a ‘Habbo celebrity’. Habbo celebrities are users or avatars with high social status that receive a lot of attention from other users and are frequently interviewed in ‘talk shows’ and at fansites.5 Other users sometimes imitate their style and behaviour. ‘Mirska’, the celebrity proprietor of ‘Mirska’s Virtual Hospital’, complained in an interview that her hospital decoration concept is being copied by others, and that some users are even pretending to be her. Imitation leads some celebrities to adopt a new style. ‘Kalasumppu’ at fansite Kala-stamo writes: As we all know, afro-style hair has long been a favorite among celebs, but at the moment [. . .] it looks like the afro is giving way to its successor, the rasta hair. Having a large number of celebrities as friends is a positive status sign. Being familiar with Habbo culture is also desirable. If one shows a lack of knowledge regarding conversational conventions, the celebrity canon (over which there is much disagreement), popular venues, or the value and history of every virtual commodity in the service, one risks being labelled a ‘newbie’: a new or otherwise insignificant user whose opinions carry little weight and can be marginalised. Sometimes ‘experienced’ users claim to be able to recognise new users based on their clothing style alone. As an example of how the correct interpretation of the meanings of various virtual consumption choices is based on negotiation between users (as opposed to being determined in advance by developers through, for example, graphic design), consider ‘Habbo stables’: rooms where participants assume the role of either a horse or a horse keeper. Since actual horse-avatars are not available, the choice is indicated by skin colour, which in Habbo, as in many other online hangouts, is a consumption preference. Dark-skinned avatars are considered horses, while light-skinned avatars are keepers, a code that, for a moment, seemed quite widely understood. Not understanding would of course have been grounds for being labelled a newbie. Another example of user-negotiated meaning is the emergence of an ad-hoc currency. When Habbo Hotel was first opened, it contained no currency that could be used in user-to-user transactions; all trade was barter. But after a

1071


1072


FIGURE 3

A room full of Plastyk chairs, used as informal currency.

while, a certain virtual plastic chair known as ‘Plastyk’ emerged as a de-facto currency (Figure 3). Prices would be denominated in plastic chairs, and the same plastic chairs could be used as a medium of exchange. The same process lead to different results in parallel versions of the system. In the UK Habbo Hotel, a three-level currency system emerged: one ‘Throne’ equals approximately 34 ‘Club Sofas’, and one Club Sofa equals approximately 75 ‘Rubber Ducks’. The exchange rates of the currency units naturally change as supply and demand fluctuate. Several fansites observe and publish the going rates; the values mentioned here are from HabboX.com.

Conclusions and discussion Through our selective exposition of everyday practices and culture in Habbo Hotel, we sought to clarify the motivations of those who participate in the system and spend money on it as well as to explain some of the more complex social dynamics that shape virtual consumer behaviour there. In this final section, we summarise the findings from our three chosen perspectives, conduct another comparison between virtual commodities and digital information, and conclude with a discussion of virtual consumerism as a ‘real’ form of consumption. Adopting an individualistic perspective on consumer culture, the everyday environment of the virtual world of Habbo Hotel can easily be described as


VIRTUAL CONSUMERISM

highly aestheticised, permeated with abundant imagery and artistic expression similarly to the cityscapes described by Featherstone (1991, pp. 95– 112). Virtual flaˆneurs derive pleasure from wandering around artistically placed stacks of furniture while wearing carefully selected couture. Self-expression, aesthetic considerations and even artistic aspirations are revealed in users’ virtual consumption choices. Similar motivations have been identified in, e.g. The Sims (Paulk 2006). From a more social perspective, it is clear that many virtual commodities are also used in the capacity of markers. They are used to signal distinctions between high status and low status, between membership and non-membership, and between one group and another. In the marker role, the aesthetic and functional attributes of a commodity may cease to matter entirely. Extrinsic attributes, particularly rarity, exclusivity and association with particular individuals or groups become the driving forces behind consumers’ choices. For instance, our material suggests that Habbo collectors are at least equally motivated by status-seeking and utilitarian pursuit for the possession of expensive rarities as they are by a romantic, noble passion for the goods as such, as anticipated by Belk (1995, 2004). The status of a virtual luxury good is not necessarily fixed, however. Like contemporary consumer society, Sulake introduces and re-introduces items and styles so that what was previously the privilege of a select few may suddenly become available to masses. This makes it difficult to rely on ownership of exclusive objects as the only means of establishing social differences, leading individuals to adopt taste and ‘correct’ consumption choices as another way of drawing distinctions. Not only must one adopt an appropriate consumption style out of all the possible combinations, one should also display the ability to classify others’ styles to be considered authentic (cf. Bourdieu 1984). Compared to the static, pre-programmed symbolic values of virtual items in The Sims (Paulk 2006), in Habbo the meaning of goods is negotiated between real people. Some users imitate the fashions of celebrities, which in turn invent new fashions to maintain the distance (cf. Simmel 1904/1957). Others opt for subversive strategies (Featherstone 1991, pp. 92 –93), aiming to discount or even replace existing status games instead of joining them. Keeping abreast of all the changing styles and commodities requires good sources of information. This is provided by fansites, the Habbo equivalent of cultural intermediaries. In all, the dynamics outlined above resemble Featherstone’s ‘economy of cultural goods’(1991, pp. 83 –94). Intricate as the Habbo cultural economy is, the dynamics of taste, fashion and imitation are no strangers to the worlds of ordinary websites, blogs and social media platforms either. Featherstone (1991, p. 17) associated these features with societies that enjoy unrestricted access to goods, information goods in this case. Thus the remarkable thing about virtual commodity platforms is perhaps rather their ability to create the more stable, old-fashioned status

1073


1074


systems: stable markers to which economic value can readily be assigned. That said, even the exchange of digital files can sometimes engender alternative economies with instances of scarcity. For example, Don Slater describes a social world where digital sex pictures are treated similarly to tradable commodities (Slater 1998). The images are digital and abundant, but finding ‘good’ or ‘rare’ ones from the global haystack is difficult. When a good specimen is uncovered, it enjoys a period of natural scarcity until it has been propagated to everyone, creating momentary difference between haves and have-nots. Yet even without considering scarcity, a significant difference between digital files and virtual commodities can be found in their life-course. Lash and Lury’s (2007) comparison of the products of the old and new culture industries illustrates this point well. According to Lash and Lury, Adorno’s commodities were atomistic: identical to each other and externally caused. In contrast, the products of the global culture industry are Leibniz’s monads: ‘all different from each other, because each carries its own trace’ (Lash & Lury 2007, p. 12). In Habbo Hotel, users can discuss the age, provenance and past owners of each instance of a very rare item. This makes each copy distinct and of different value. For a rare sex picture, on the other hand, it does not make sense to discuss the provenance of individual copies: they are all the same. How do we address the charge levelled from the ‘production of consumption’ perspective: that Habbo Hotel and similar services are deceiving consumers into spending money on an illusion? One strategy is to note that everywhere in our economy, goods have a social life beyond their physical qualities. Even the consumption of food and drink is far removed from simple fulfilment of physical needs, in terms of quality, quantity, form and ritual. Positional goods are frequently made artificially scarce to enhance their value as markers (Leiss 1983). Thus it is not so much a question of Habbo luring the consumer from ‘real’ to ‘illusionary’ consumption, but from one arena of socially oriented consumption to another. We have seen that virtual commodities are able to fulfil the same kinds of aesthetic and social roles as material commodities in their respective subcultures (cf. Thornton 1996). When we label them ‘virtual’, we do not mean to imply that they are less ‘real’ – only that they are computer-mediated. If there is an unreal air to how intangible objects can be worth lots of money, it is an observation regarding the nature of our consumer culture in general, of which virtual consumerism is only a naked example. In this sense, all consumption is virtual. Yet it is obvious that from a parent’s point of view, no matter how real they are, social arenas that are computer-mediated may not be very attractive targets for investment, especially if they steal resources away from the ‘big game’ of school and career. Marginalising them as ‘not real’ may thus be an appealing strategy. However, as more and more aspects of life from work to social relations become increasingly computer-mediated, it becomes progressively more difficult to distinguish between these arenas (on work, see e.g. Pinckard 2006; Yee 2006).


VIRTUAL CONSUMERISM

If a quarter of all Finnish teenagers use Habbo Hotel, there may well be school classes where almost every pupil is a user. In this case, what is the relationship between the pupils’ Habbo identities and their schoolyard identities? One possibility is that virtual and material consumption profiles are positively correlated. Previous research suggests that young people’s ways of using new technology correspond with their consumption patterns in general: those who are interested in and competent with new technology are likely to be more trend and brand conscious consumers, and vice versa (Wilska 2003; Wilska & Pedrozo 2007). Another possibility is that status obtained in Habbo Hotel could act as a substitute to material status consumption within a reference group. In other words, lack of trendy sneakers at school could be compensated with virtual dragons and record players. If this was the case, it could have interesting ecological ramifications. Assuming fixed income, any money spent on virtual goods must necessarily mean less money spent on material goods. A virtual item carrying the same price and symbolic payload as a material commodity could potentially have a significantly smaller ecological footprint. The production, distribution and disposal of virtual commodities requires only negligible resources, although their design and use consumes computing related resources such as electricity. Serious studies on the environmental impact of virtual commodities are yet to be conducted, but in July 2005, Sulake received an award from Dodo, a Finnish environmental organisation, for the ‘biggest inadvertent environmental feat of the decade’ (Dodo 2005). Dodo believed that by providing a venue for virtual consumption, Sulake had created a greener alternative for consumption crazy teenagers to spend their allowances on. Our notion of virtual consumerism is intended to draw attention to this societal side of the real-money trading phenomenon. Compared to the innocuous consumption games of the past, today’s virtual consumption is increasingly a part of the surrounding society, a new extension of the playfield of contemporary consumerism. For now, it seems the rules remain largely the same, from global division of labour to systems of symbolic exclusion. But in a mediated society, virtual consumerism also carries transformative potential, at least towards the material aspect of consumption.

Notes 1 2

Except for the gradual wearing down of computer hardware and network infrastructure, and the electricity they consume. Note that Lash and Lury (2007, p. 182) use the similar term virtual object to refer to ‘potentials that generate a succession [. . .] of actual forms’, e.g. a brand that is actualised as a series of sporting events. In this parlance, virtual commodities are instances of ‘actual form’. However, due to their

1075

1076


3


4

5

digital nature, virtual commodities can remain extremely close to the objects that spawned them, e.g. a digital brand image. Copyright law of course tries to persuade users that media representations are commodities that can be sold by the piece. The ease at which digital information is duplicated works against this interpretation being perceived as correct by the consumers. So-called Digital Rights/Restrictions Management technologies attempt to turn media products into actual virtual commodities. According to Sulake (2005), Habbo Hotel Finland had 280,000 monthly unique visitors in 2005, although some probably just visited the front page. According to Sulake (2008a, p. 189), 57 per cent of Finnish Habbo users are 13 – 18 years old. According to Statistics Finland (2008), there are approximately 640,000 people aged 10 –19 years in Finland. Thus at least 280,000 ! 57 per cent/640,000 ¼ 24.9 per cent of 13 – 18-year-olds were visiting Habbo in 2005. It is not unreasonable to assume that similar penetration continues today, given that Sulake (2008b) claims strong growth in total user base globally. These ‘Habbo celebrities’ should not be confused with occasional visits by actual celebrities such as Ozzy Osbourne, which are organised by Sulake as part of marketing and promotional activities.

References Appadurai, A. (ed.) (1986) The Social Life of Things: Commodities in Cultural Perspective, Cambridge University Press, Cambridge. Baudrillard, J. (1971/1988) ‘Consumer society’, in Jean Baudrillard: Selected Writings, ed. M. Poster, Polity, Cambridge. Baudrillard, J. (1994) ‘The system of collecting’, in Cultures of Collecting, eds J. Elsner & R. Cardinal, Reaktion, London, pp. 7–24. Baudrillard, J. (1996) The System of Objects, trans. J. Benedict, Verso, London. Bourdieu, P. (1984) Distinction. The Social Critique of the Judgement of Taste, Routledge, London. Belk, R. W. (1995) Collecting in a Consumer Society, Routledge, London. Belk, R. W. (2004) ‘The human consequences of consumer culture’, in Elusive Consumption, eds K. M. Ekstro¨m & H. Brembeck, Berg, Oxford, pp. 67–86. Campbell, C. (1987) The Romantic Ethic and the Spirit of Modern Consumerism, Blackwell, Oxford. Campbell, C. (2004) ‘I shop therefore I know that I am: the metaphysical basis of modern consumerism’, in Elusive Consumption, eds K. M. Ekstro¨m & H. Brembeck, Berg, Oxford, pp. 27–44. Castronova, E. (2006) ‘Virtual worlds: a first-hand account of market and society on the Cyberian frontier’, in The Game Design Reader: A Rules of Play Anthology, eds K. Salen & E. Zimmerman, MIT Press, Cambridge, MA, pp. 814–863.


VIRTUAL CONSUMERISM

Dibbell, J. (1998) My Tiny Life: Crime and Passion in a Virtual World, Henry Holt, New York. Dodo (2005) ‘Dodon paä¨sia¨sissaaripalkinto jaetaan virtuaaliyhteiso¨n kehitta¨ja¨lle’, press release, 19 July 2005, [Online] Available at http://www.dodo.org/ lehdistotiedotteet/paasiasissaaripalkinto_jaetaan.html (17 July 2008). Douglas, M. & Isherwood, B. (1978) The World of Goods, Basic Books, New York. Fairfield, J. (2005) ‘Virtual property’, Boston University Law Review, vol. 85, no. 4, pp. 1047–1102. Featherstone, M. (1991) Consumer Culture & Postmodernism, Sage, London. Fung, A. (2006) ‘Bridging cyberlife and real life. A study of online communities in Hong Kong’, in Critical Cyberculture Studies, eds D. Silver & A. Massanari, New York University Press, New York, pp. 129–139. Grimes, S. M. (2006) ‘Online multiplayer games: a virtual space for intellectual property debates?’, New Media & Society, vol. 8, no. 6, pp. 969–990. Guo, Y. & Barnes, S. (2007) ‘Why people buy virtual items in virtual worlds with real money’, The DATA BASE for Advances in Information Systems, vol. 38, no. 4, pp. 69–76. Herman, A., Coombe, R. J. & Kaye, L. (2006) ‘Your Second Life? Goodwill and the performativity of intellectual property in online digital gaming’, Cultural Studies, vol. 20, nos 2 –3, pp. 184–210. Horkheimer, M. & Adorno, T. (1976) Dialectic of Enlightenment, Continuum, New York. Huhh, J. S. (2008) ‘Simple economics of real-money trading in online games’, working paper, [Online] Available at http://ssrn.com/abstract¼1089307 (22 July 2008). Hyatt, N. (2007) ‘Microtransactions actually work, proof from MapleStory in the US’, [Online] Available at http://nabeel.typepad.com/brinking/2007/03/ microtransactio.html (22 October 2008). Jin, G. (2006) ‘Chinese Gold Farmers in the Game World’, Consumers, Commodities & Consumption: A Newsletter of the Consumer Studies Research Network, vol. 7, no. 2, [Online] Available at https://netfiles.uiuc.edu/dtcook/www/CCCnewsletter/ 7-2/jin.htm (24 July 2007). Johnson, M. & Toiskallio, K. (2005) ‘Fansites as sources for user research: case Habbo Hotel’, Proceedings of the 28th Conference on Information Systems Research in Scandinavia (IRIS’28), [Online] Available at http://www.soberit.hut.fi/ johnson/Johnson_IRIS_2005.pdf (24 July 2007). Johnson, M. & Toiskallio, K. (2007) ‘Who are the Habbo Hotel users – and what are they doing there?’, Proceedings of the Nordic Consumer Policy Research Conference 2007 (Consumer 2007), [Online] Available at http://www.consumer2007.info/ wp-content/uploads/ProceedingsConsumer2007v2.pdf (22 October 2008). Lash, S. & Lury, C. (2007) Global Culture Industry, Polity, Cambridge. Lastowka, F. G. & Hunter, D. (2004) ‘The laws of the virtual worlds’, California Law Review, vol. 1, no. 92, pp. 1–73.

1077


1078


Lehdonvirta, V. (2005) ‘Real-money trade of virtual assets: ten different user perceptions’, Proceedings of Digital Art and Culture 2005, IT University of Copenhagen, Denmark, [Online] Available at http://virtual-economy.org/files/ Lehdonvirta-2005-RMT-Perceptions.pdf (22 July 2008). Leiss, W. (1983) ‘The icons of the marketplace’, Theory, Culture & Society, vol. 1, no. 3, pp. 10–21. Lury, C. (1996) Consumer Culture, Polity, Cambridge. Malaby, T. M. (2006) ‘Parlaying value: capital in and beyond virtual worlds’, Games & Culture, vol. 1, no. 2, pp. 141–162. McCracken, G. (1988) Culture and Consumption: New Approaches to the Symbolic Character of Consumer Goods and Activities, Indiana University Press, Indianapolis. Miller, D. (1987) Material Culture and Mass Consumption, Blackwell, Oxford. Paulk, C. (2006) ‘Signifying play: the Sims and the sociology of interior design’, Game Studies, vol. 6, no. 1, [Online] Available at http://www.gamestudies. org/0601/articles/paulk (22 July 2008). Pinckard, J. (2006) ‘Is world of warcraft the new golf?’, article in 1UP.com, 8 February, [Online] Available at http://www.1up.com/do/newsStory?cId¼3147826 (22 July 2008). Simmel, G. (1904/1957) ‘Fashion’, American Journal of Sociology, vol. 62, no. 6, pp. 541–548. Slater, D. (1998) ‘Trading sexpics on IRC: embodiment and authenticity on the internet’, Body and Society, vol. 4, no. 4, pp. 91–117. Slater, D. & Miller, D. (2000) The Internet: An Ethnographic Approach, Berg, Oxford. Statistics Finland (2008) ‘Population by age 31.12.2007’, Population and Cause of Death Statistics, [Online] Available at http://stat.fi/til/vaerak/2007/vaerak_ 2007_2008-03-28_kuv_002_en.html (22 October 2008). Sulake (2005) ‘Habbo Hotelin luojat palkittiin Opetusministerio¨n Suomi-palkinnolla’, press release, 8 December 2005, [Online] Available at http://www. sulake.com/press/releases/2005-12-08-Habbo_Hotelin_luojat_palkittiin_ Opetusministerioen_Suomi-palkinnolla (22 October 2008). Sulake (2008a) Global Habbo Youth Survey, Sulake Corporation, Helsinki. Sulake (2008b) ‘Habbo – where else?’ Introduction to Habbo on Sulake Corporation’s website, [Online] Available at http://www.sulake.com/habbo/ (22 July 2008). Taylor, T. L. (2006) Play Between Worlds: Exploring Online Game Culture, MIT Press, Cambridge, MA. Thornton, S. (1996) Club Cultures: Music, Media and Subcultural Capital, Wesleyan University Press, Middletown, CT. Veblen, T. (1899/1955) The Theory of the Leisure Class, Mentor Books, New York. Wilk, R. (2004) ‘Morals and metaphors: the meaning of consumption’, in Elusive Consumption, eds K. M. Ekstro¨m & H. Brembeck, Berg, Oxford, pp. 11–26. Wilska, T.-A. (2003) ‘Mobile phone use as part of young people’s consumption styles’, Journal of Consumer Policy, vol. 26, no. 3, pp. 441–463.

VIRTUAL CONSUMERISM


Wilska, T.-A. & Pedrozo, S. (2007) ‘New technology and young people’s consumer identities. A comparative study between Finland and Brazil’, Young. Nordic Journal of Youth Research, vol. 15, no. 4, pp. 343–368. Yamaguchi, H. (2004) ‘An analysis of virtual currencies in online games’, unpublished manuscript, [Online] Available at http://ssrn.com/abstract¼544422 (22 July 2008). Yee, N. (2006) ‘The labor of fun’, Games and Culture, vol. 1, no. 1, pp. 68–71. Vili Lehdonvirta works as a researcher at Helsinki Institute for Information Technology, focusing on virtual consumption and real-money trade of virtual property. He graduated from Helsinki University of Technology in 2005 with an MSc (Tech.) in Knowledge Intensive Business and is currently finishing a PhD degree in Economic Sociology at Turku School of Economics, Finland. Between 2006 and 2008 he was Visiting Scholar at Waseda University, Tokyo. He is the co-founder of Virtual Economy Research Network, http://virtualeconomy.org. Address: Helsinki Institute for Information Technology, PL 9800, Espoo, 02015 TKK, Finland. [email: [email protected]] Terhi-Anna Wilska is Professor of Economic Sociology at Turku School of Economics. She obtained her PhD in Sociology in 1999 from Lancaster University, UK. Her special interest areas are consumption behaviour and lifestyles of children, young people and families, as well as social implications of new media. Address: Turku School of Economics, Turku, Finland. [email: [email protected]] Mikael Johnson works as a researcher at Helsinki Institute for Information Technology, focusing on developer – user dialogue in digital communities. He graduated from Helsinki University of Technology in 2002 with a MSc (Tech.) in usability engineering and is now pursuing a PhD degree at the Computer Science and Engineering Department. He is currently Special Researcher at National Consumer Research Centre, Finland. Address: Helsinki Institute for Information Technology, PL 9800, Espoo, 02015 TKK, Finland. [email: [email protected]]

1079

ARTICLE

V


Mikael Johnson and Tanja Sihvonen (2009). "On the Dark Side: Gothic Play and Performance in a Virtual World". Journal of Virtual Worlds Research 1 (3). Reprinted with permission.

263

Vol. 1. No. 3 ISSN: 1941-8477 “Cultures of Virtual Worlds” February 2009

On the Dark Side: Gothic Play and Performance in a Virtual World By Mikael Johnson, National Consumer Research Centre, Finland; Tanja Sihvonen, University of Turku, Finland

Abstract

This article investigates gothic play and performance in Habbo, a virtual world for children and teenagers. In addition to analyzing the aesthetics of the environment and its player characters, the authors seek to discover how gothic players figure in its appropriation and redevelopment. Countercultural player activities, such as playing goth in Habbo, are often treated as disruption, but we argue that gothic players are in fact rather resourceful and productive members of the community. Observations of player-generated content, events, rituals, appearance, and group discussions indicate that goth is more than a style – it is also play and performance. Gothic players are not only consumers, but also content providers who inspire the developers at Sulake, the company behind Habbo. The influence of Habbo goths is evident, as emergent gothic player-created content has repeatedly been incorporated in newer releases of the game platform.

Keywords: goth; gothic; virtual world; subculture; counterculture; play; performance.

This work is copyrighted under the Creative Commons Attribution-No Derivative Works 3.0 United States License by the Journal of Virtual Worlds Research.

Journal of Virtual Worlds Research – On the Dark Side 4

On the Dark Side: Gothic Play and Performance in a Virtual World By Mikael Johnson, National Consumer Research Centre, Finland; Tanja Sihvonen, University of Turku, Finland

The foundations of virtual worlds, especially massively multiplayer online games (MMOGs), may be based on their developers’ intentions, but it is clear that the social dynamics present in them are mainly the result of player interactions. To understand what constitutes a virtual world in this cultural or social sense, it is mandatory to investigate what kinds of practices and ideas characterize the social interaction between players. What is also important to consider is how players reproduce their player identity on the internet and how they negotiate and talk about their gameplay habits outside of the game in related discussion boards and wiki pages. In this study, we concentrate on analyzing the countercultural player activity we conceptualize as “gothic” and the ways the particularly gothic behavior is manifested in online immersive 3D environments. As a specific point of reference we look at the player behavior in Habbo (Sulake, n.d.), which is an open and welcoming game environment and platform for social interaction for all kinds of players – as long as they play by the rather strict rules. Since Habbo is a virtual world for children and teenagers – 90 percent of the players are between 13 and 18 years of age – it is subject to notable monitoring and moderation from the part of the players’ parents as well as of the developer company, Sulake. All activities referring to violence or sex, even implicitly, are prohibited. Despite the parental concerns and far-reaching monitoring mechanisms, “goth” is played out in various ways also in Habbo. Gothic themes, such as the utilization of morbid aesthetics in the creation of game character habitats, seem to be fairly widespread in this online environment. It can be suggested that presenting one’s avatar or room as gothic in Habbo is a countercultural activity as it is not always in accord with developer intentions; with that said, it has to be pointed out that gothic play can be done in diverse or inconsistent ways. Gothic themes in the context of Habbo range from experimental and ironic performance to the player contemplation of deep identity issues. What are the ways gothic is manifested, then, and how do gothic players figure in the development and appropriation of virtual worlds such as Habbo? Our starting point is to illuminate the diversity of gameplay practices and their importance by concentrating on the gothic subcultural style of play and performance. Before delving into the analysis of gothic behavior and performance in virtual worlds, however, we will provide an overview of the research context of virtual worlds to which this text seeks to contribute. Contextualizing Player Behavior Habbo is a virtual environment where children and teenagers meet, socialize, and play many types of games. It was first launched in August 2000 in Finland as Hotelli Kultakala (“Hotel Goldfish”) and it was based on the developers’ two earlier online services. At the time of writing, there are Habbo hotels in thirty-three countries, and 11.5 million players visit Habbo each month (Sulake, n.d.). Instead of an entrance or a monthly fee, the profit model is based on 4


micro-payments in the hotel. Virtual furniture, mini-games, and membership in the Habbo club are bought with Habbo credits. These credits can be purchased (depending on the country) with pre-paid cards, bank transactions, credit cards, or special text messages that add a specified amount of money to the customer’s mobile phone bill. The social interaction in Habbo is multifarious. In the design of Habbo, clear winning conditions and gameplay rules have been avoided, and instead, players are encouraged to create their own objectives beyond chatting, room decoration, and meeting friends. The provided environment for these activities is a hotel consisting of public and private rooms, where the virtual hotel visitors, called Habbos, chat, buy virtual furniture, decorate rooms, play minigames, and arrange social events. Most of the teenage players log on after school, and according to Sulake, the developer company, on average they spend around forty-five minutes per day in the hotel or on its related discussion forums. Even though Habbo is an open virtual world with built-in social networking tools (Messinger, Stroulia, & Lyons, 2008), it has very active player forums outside of the gameworld. Habbos, the game characters, can be thought of as avatars mainly because they can be used also in role-play. What is notable in the player behavior in and around Habbo is the players’ own creation: the players design their characters and their individual rooms, but they also use the available means to organize social and game events often on a specific theme. Sulake has even incorporated specific “Habbo homepages” for each avatar to support player creativity and selfexpression. The player-generated content that we consider in this article includes avatars, virtual objects, and shared virtual spaces. Concepts like participatory culture, social media, online community, user innovation, player-generated content, and crowdsourcing have recently been applied to describe the changing relations between producers and consumers of online products and services like virtual worlds (Fischer, 2002; Jenkins, 2008; Sotamaa, 2007). Generally, in contrast to non-digital massproduction where consumption decreases the value of a product, “consumption” of virtual worlds increases their value. Gameplay is an activity that often results in something tangible, which can then in turn be transformed into something of value. A max-level avatar or an extremely rare weapon are examples of in-game items that may have monetary value both in the gameworld as well as outside of it on game-related auction sites. Players are naturally productive members of the gaming community, and they have active roles in all stages of the innovation process. By playing the game the players incorporate game development ideas and practices into the game culture they create and maintain. In the long run, nobody wants to be in an empty virtual world; just by being there, players play a key role in the maintenance of those virtual worlds.1 Players help each other, forming teams and guilds to achieve shared objectives. In addition to these visible roles, some players also carry out other, more subtle experiments and try out different positions in and out of character. Some create auxiliary game-related websites, some act as beta-testers, others develop technical or content

1 On the other hand, it is recognized that solitary play is common among some player groups (Ducheneaut, Nicholas Yee, Nickell, & Moore, 2006). In all game-like virtual worlds, practices like leveling up and “grinding” are often done alone; similarly, preparing for big, public fights or events are usually done outside of the social context of play. Even in Habbo, it is customary that players tune up their game characters and organize their virtual assets in solitude for the public show-off.

5


modifications, and almost all are providers of feedback to the developer company in one way or another. Virtual worlds emerge as the result of complex processes of negotiation between developers and players from the start of the design process. Developers work with a vision of a good game and fun gameplay, which implicitly shapes future players. Developers consider what the players might do and how they might play. The vision is then realized during game development into particular features. As actual players put the features into use, they appropriate them for their own purposes, and also make up their own, novel ways of using the game (Silverstone & Hirsch, 1992/1994). These creative gameplay processes feed back to the developers, who learn, develop their vision, and adapt the features to suit the emergent use of the game environment. As new features are developed and appropriated, this slowly paced dialog continues between developers and players through the (co-realized) material foundations of the virtual world. Graphical player-generated content has changed the web user experience and given players new possibilities for online self-expression. The internet itself has changed from having been mostly textual (e.g., multi-user dungeons and newsgroups) towards a more graphical basis that builds either on computer-generated 3D graphics (MMOGs) or photos and other visual material (e.g., social networking sites) (King, 2008). Games-related communication in particular is not only based on chats, social forums, or messaging services, but it also relies on the making and distribution of skins and clothes for game characters, as well as virtual objects and decorations for in-game locations that are usually presented to other players via screenshots. In addition to the game platform itself we also consider Habbo fansites as relevant research material. Fansites, or player-authored websites and forums, are essential in considering the characteristics of player behavior in the context of virtual worlds. Unlike traditional singleplayer games without any network connection, stand-alone games with internet connectivity make it part and parcel of the game itself for players to show their avatars, skills, and possessions to other players. Besides designer-provided game objectives, the online sharing and trading of these virtual assets has become an activity itself in the context of Habbo. If not possible within the gaming platform, fansites and other third party platforms often provide complementary services to the realms of virtual worlds (Johnson & Toiskallio, 2005). Many game-themed supporting websites and fansites created by active players have assumed an important role in the maintenance of game cultures. These fansites help players and support gameplay by sharing hints, cheat codes, secret paths, guidelines, stories, and perspectives about what players regard essential. By doing so, the fansites render the player culture around a virtual world visible and material, and they also contribute to the general understanding of the game in an important way. The fansites may also reproduce and reinforce player “career paths” and the legitimization of player groups in specific ways (Sihvonen, 2009). With their online activities and player-generated content, MMOG fansites often expand and comment on the media content produced by the game developers. The processes of giving feedback and negotiating with the available game contents may present themselves as harmonious and constructive, a win-win situation for both the developers and the players, as they both aim for the common goal of enriching the gameplay experience. However, there are situations where the conflicting interests of these parties unexpectedly surface. Practices like hacking, scamming, and grief play – a playing style where a player intentionally aggravates and

6


harasses other players (Foo & Koivisto, 2004) –, as well as specific MMOG phenomena such as mass protests and virtual item trade may seriously undermine the developer-player relations, and thus, the virtual dynamic constantly needs to be addressed by the developer company. Conflict management and governance are much needed also in virtual worlds. One of the important underlying questions in MMOGs is how governance can be organized so that it best supports the expression of different opinions while not suppressing the identity building processes game players engage in (Taylor, 2006). Gothic play can be interpreted in the context of identity politics, or as adoption of a certain counter- or subcultural position. This usually manifests itself in the stylistic choices the players employ as part of their gameplay. However, playing goth can not simply be considered as grief play, or a countercultural activity in the sense that it would self-evidently disrupt the logic of play originally intended by the game developers. Instead, we see goth more as a subcultural activity, as something that is not necessarily anticipated by the developers or the mainstream player community. Gothic play is rather like toying with the hidden subtext in the layers that make up the cultural product as a whole. At the same time, gothic themes can also be rather effortlessly incorporated as part of general game (re)development taking place through the developer-player dialogue described above. An example of such incorporation is the prevalence of gothic characters, such as the vampire and the werewolf, in the expansion packs of The Sims 2 as the result of players wanting to focus on such themes. Gothic Themes in Teenagers’ Virtual Playground In the context of studying gothic behavior and aesthetics it is important to define the convoluted terminology. Words like “goth” or “gothic” have different connotations depending on the context of their use and origin. Furthermore, gothic themes have not been present in academic research to the degree that they have in general popular culture, and therefore, the significance of these words is largely associated with themes present in more popular discourse. In the context of popular culture, the Oxford English Dictionary defines gothic as ”a genre of fiction characterized by suspenseful, sensational plots involving supernatural or macabre elements and often (esp. in early use) having a medieval theme or setting” (2008). However, at least regarding virtual worlds, this definition lacks the sense of irony and black humor which are often considered important in the appropriation of gothic themes. We argue that a more fruitful approach to gothic would be interpreting it rather in the vein of “harnessing the dark forces in an ironic spirit,” and hereby we follow writers like Gavin Baddeley (2006) and Paul Hodkinson (2002, 2006). Fantasy and horror are easily applicable, everyday themes in many virtual worlds and game environments; they are present also where one would not expect it. World of Warcraft and Second Life (Gothic Second Life, n.d.) may be obvious candidates for exploring provoking and macabre topics associated with gothic aesthetics, but even Facebook users are regularly bitten by virtual vampires (Vampires, n.d.). Monster-like characters as well as horrifying extra-terrestrial life-forms play an important role in current virtual worlds, but they are especially manifest in environments that are based on a confrontation between good and evil. The introduction of such fearful forces has been a welcomed point of departure and feature in the creation of suspenseful and sensational plots for digital games. Culturally meaningful and multidimensional monsters have the power to fascinate, where as hastily sketched and easily defeated opponents do not.

7


However, not all virtual worlds are violent or about monsters. Habbo, for instance, is a cartoon-like virtual world popular among millions of teens, set in a contemporary indoor setting. Habbo players are encouraged to interact in a non-violent fashion and it is not possible to inflict harm on avatars or break things in Habbo. The only non-player characters are pets – such as cats, dogs, and crocodiles – that live forever and are designed to have a friendly appearance (Habbo UK, n.d.). Even though Habbo is the perfect opposite to violence and horror, goths are encountered in Habbo, which begs two questions. First, how is gothic represented in style – that is, how do players make a cheerful cartoon-like avatar or a room that looks gothic? Second, in addition to this representational dimension, the question of how goth is played out and performed needs to be answered; what does it mean to engage in gothic activities and social performances in this kind of a benevolent virtual world? As we set out to understand how gothic is manifested in Habbo, a few notes on our methodology are necessary. The material discussed here was gathered online through observations of player-generated content, online behavior, appearance, and discussions. This research is based on fansites and discussion forums, online player and group profiles, online images of avatars, and shared virtual spaces in Habbo (Johnson & Toiskallio, 2007). In addition, our understanding of Habbo draws on developer (N = 10) and player interviews (N = 12), as well as ongoing developer collaboration in a research project (Johnson, 2007). The Habbo player communities are based on localized Habbo sites that exist independently in each of the thirty-two Habbo hotel countries. Each country has a thriving community with bigger and smaller fansites. Johnson and Toiskallio (2005) found footprints of 173 player-created Finnish Habbo fansites in 2004. The 6,850 threads regarding Habbo fansites in the UK is also a clear sign of an active player community (Habbox Forum, n.d.). In 2007, the Habbo developer company Sulake extended the virtual game world to include social networking features such as player and group profile web pages, and gradually the fan community landscape started to change. At present, player discussions are prevalent on the sites provided by Sulake. Furthermore, the new Habbo website structure with social networking features, and particularly tags enable searching for players and groups with particular interests, such as “goth.” Habbo players use tags to describe their avatars and groups, but these tags are also interactive. When players click on a tag, the system shows an index of all avatars and groups that use that tag. Compared to previous research on goths online, such as those conducted by Hodkinson, (2002, 2006), this study focuses more on visual culture than on the structure (central vs. dispersed) of textual discussions. Especially interesting is how the visual culture associated with gothic themes is influenced by the operating logic of the technical platform in question. Not all online games or media are similar, which makes it important to shed light on how the gothic subcultures are manifested in different ways in their specific contexts. The selection and presentation of gothic examples in this article fall into two categories: the players themselves explicitly identify their creations as gothic – in names and descriptions of player characters and hotel rooms – or we discuss specific themes (e.g. death, suffering, monsters) commonly considered as gothic in the context of Habbo. It is difficult to give an exact estimate of the popularity of a subculture such as “gothic players.” However, thanks to recent developments in the Habbo platform, it is possible to give some quantitative figures. Table 1 contrasts Habbo tags that relate to the gothic subculture (goth, gothic, goths, dark, vampire, death) with tags that relate to mainstream music styles (rock, rap)

8


and phenomena (love, football, emo, cats) as well as a three other subcultural music styles (hiphop, punk, metal). The tags are counted in the US and UK Habbo hotels. Table 1. Selected Totals of Habbo Tags in US and UK as of September 2008.

tag

habbo.com

love rock football rap emo metal punk cats hiphop goth death vampire dark gothic goths

25149 20483 13302 10016 7027 3768 3336 3243 2710 1559 1124 1080 622 543 50

habbo.uk 1849 1509 2201 236 735 394 249 737 60 226 166 113 74 63 33

This table indicates that the gothic subcultures, defined through the gothic-themed tags in Habbo US and UK, are ten times smaller than the most popular tags. However, the gothic tags outnumber other tags shared by only a handful of avatars. This implies that goths in Habbo are few enough not to be mainstream, but large enough to represent more than a scarce, individual interest. Because the numbers are fairly similar in Habbo US and UK, gothic play does not seem to be a sign of anything particularly national; gothic in this sense is rather a reference to transnational (or translocal) subculture that is possibly global in scope (cf. Global Gothic, n.d.). To sum up, even though gothic themes may be somewhat underground, they are widely dispersed in the undercurrents of popular culture throughout the Western world. Understanding the Dark Gothic Core of Popular Culture Traces of gothic themes prevail in popular culture. Gothic has manifested itself in the history of digital games, and especially the role-playing game (RPG) subculture, in numerous ways. The RPG culture has been occupied with gothic thematics, figures, and sceneries to the point of being accused of obsessing over the “dark forces.” David Waldron (2005) traces the gothic roots of RPGs to the early 1980s and the Christian moral panic in the US, the objective of which was to ban RPGs through various media campaigns and legal actions. Schools, parents, and Christian associations joined forces in the campaign against games such as Dungeons & Dragons.

9


One of the long-lasting effects of this RPG moral panic was the emergence of a stronger need for identity-building among game developers and player groups. Role-players in the 1980s were young enthusiasts, and no ideologies or political agendas per se were associated with the leisure activity. As role-players were attacked in public debate, they had to rise to the challenge by organizing themselves into groups and more clearly defining the nature of their hobby and gameplay activity. These practices gradually aided also the consolidation of the subcultural identities linked with RPGs. Furthermore, according to Waldron, the subcultures that were formed around role-playing were associated with the strengthening of gothic influences in the more general popular culture at the time. Role-players started to connect more tightly with fans of other media products, such as horror, science fiction, and fantasy literature, movies, TV shows, as well as computers and computer networking. Fan cultures around RPGs expanded by incorporating player created fanzines (fan magazines), cons (meetings), and discussion forums. A certain jargon became established, and members could be identified through a certain way of dressing – general characteristics to many other subcultural manifestations, as well. Nevertheless, the connections among goth, subculturality, and role-playing are not simple and straightforward. Goth is about a certain flexible sensibility, the origins of which can be traced to artistic movements in the English-speaking world in the eighteenth century. This socalled neo-gothic style took inspiration from Europe’s pagan past and created imageries of monsters – the vampire, Frankenstein’s monster, the living dead (zombie), and the werewolf – of which many lived on for centuries (Davenport-Hines, 1998). What is typical to the gothic sensibility is its simultaneous sympathy and antipathy for these monster figures, resulting in complex settings and scenarios. In gothic culture, non-human characters tend to have human characteristics that make us want to identify with them, despite their evil and inhuman origins. These human traits can simultaneously be both hidden and exaggerated. According to Gavin Baddeley (2006), gothic is a cultural aesthetic that lives on oppositional, anti-mainstream, and subcultural mentalities. “Performing goth,” however, remains open for interpretation and individual appropriation. Gothic more likely signifies a viewpoint, subversion, and counter-action, or even a lifestyle, than it does a mere aesthetic choice. In the contemporary cultural context, goth signifies a subcultural phenomena that can be first located in the early 1980s Britain. Gothic style incorporated elements from the Romantic and Victorian traditions as well as fantasy, the glorification of the mystical and the supernatural, and marginal ideas and ideologies outside of the mainstream. Gothic music emerged in the aftermaths of punk inspired especially by horror literature and movies (Davenport-Hines, 1998). Many cultural elements were mixed and matched with an attitude that had a hint of irony or camp aesthetics. Since the early 1980s, a great deal of gothic horror and monster thematic has been applied to games played on different platforms. For instance, the narratives of the two Dracula games that were made for early hand-held gaming devices, were loosely based on Bram Stoker’s classic novel of the vampire count (Imagic from 1983 and Epoch from 1982). The idea of the games is to combat werewolves and vampire bats, enter Dracula’s castle and steal his gold. Expanding this kind of a simple, widely-known, and easily malleable back story to different game platforms has proven an extraordinarily profitable starting point for game design and development.

10


Today, there seems to be an abundance of dark fantasy and horror games: Call of Cthulhu, Gothic3, The Vampire, and Werewolf games from the World of Darkness series are examples of rather mainstream utilization of gothic themes. The aesthetic influences in these games often hark back to movies and other audiovisual culture products. A great deal of the work done by directors such as Tim Burton circle around gothic themes, as films like Edward Scissorhands, The Nightmare Before Christmas, Corpse Bride, and Sleepy Hollow illustrate. These types of cultural products act as important models and testing grounds for the stylistic choices associated with gothic sensibility, therefore providing a basis for the maintenance of a subculture also visible in the context of a non-violent online game. How Habbos Are Tweaked Gothic It is typical for subcultures that their ”subculturality,” or means of distinction from mainstream culture, is a process of continuous negotiation. To place oneself in the margin is an active act that constantly needs to be reproduced through various actions and visible symbols. Gradually these markers may form a style that can be associated with a specific subculture and behavior. Based on these stylistic distinctions, different degrees of membership emerge: ”real” members of the subculture may want to be distinguished from “wannabe” members or the ones who only “flirt” with the margin of the group (Hebdige, 1979/1988).

Figure 1. Gothic Style Avatar (Calliope, n.d.).

Figure 1 show a gothic style avatar from a Finnish fansite discussion (Calliope, n.d.). In the accompanying description of the gothic style, the difference between ”real goths” and ”wannabe goths” is actively reproduced and negotiated. According to these self-proclaimed Habbo specialists, for instance, wannabe goths temporarily try out a white avatar skin shade for fun, where as real goths keep their avatar skin white consistently and continuously. Calliope writes:

“Gothic. In addition to punk, this style is really popular in Habbo, and in real life. White skin and dark clothes guarantee the beauty of the gothic style. But there are also wannabe-goths, who just use gothic clothes in Habbo for fun, if something is really bothering. You recognize a wannabe when you see that they don’t wear the previously mentioned clothes on the day after. Gothic colors are for example black, red, and violet”. An example of a REAL goth: Wrox (translated by the authors from Finnish to English from Calliope, n.d.).

11


Especially in those virtual worlds, where human-like avatars are common, it is typical that avatar subclasses emerge, as do players who distinctively want to play a certain kind of avatar (based, for instance, on their real-life self-identification). References to gothic style in this sense can be found already in the forums of Nerokala – a Finnish community that formed around Habbo in 2002, but later on branded itself as an online youth forum – in discussions about different avatar styles and potential clashes between the player’s normal everyday style and the appearance of her avatar. Even though gothic player cultures are not extraordinarily vast in Habbo, gothic elements found on the player forums support the subcultural character of playing goth in particular ways. It can be suggested that gothic in the context of these games is one way of standing out in the crowd and opposing the mainstream player communities, as well as the developers of the game with their rules and monitoring – that is, power – they impose on players. While many avatars designed for virtual worlds such as Habbo incorporate hegemonic beauty ideals – they are often young and “white” characters with colorful and trendy outfits – there are implicit as well as explicit gothic, horror, and dark fantasy elements present in Habbo character creation. In the design of gothic avatars, existential agony, Weltschmerz, pain, and suffering related to death and love emerge as central topics, often in overtly exaggerated ways and through the use of melodramatic expressions (possibly typical to the teen years). Even though all themes relating to violence are prohibited in Habbo, there are ways to offset this rule. These subversive practices are visible in the look and feel of many player profiles in the community.

Figure 2. Two Different Gothic Avatars (XXXDominatrix, n.d.; GothicaTheLost, n.d.).

The avatar on the right in Figure 2 displays perhaps the only way of explicit avatar violence possible in Habbo. By positioning a pin on the same spot as the avatar (on the avatar homepage, Habbo Home), it gives the impression that the avatar – or more specifically, this avatar’s heart – is pierced or stuck by a giant pin. This way, even these innocent and non-violent Habbo features can be tweaked to portray suffering. However, not all Habbo avatars that tag themselves as “goth” follow the dark gothic style. Figure 2 shows also another kind of gothic avatar on the left that is also tagged with “emo” and “punk,” which is an example of using multiple style or group identity markers

12


simultaneously. The tags can be interpreted as a sign of overlap between these styles, an individual selection of styles, or a low degree of commitment to any styles in particular. It has been suggested that portraying a style such as goth can be more like a flimsy experiment or a phase rather than an expression of identity, as is detectable in the attached chat posting: “Unlike normal text chats, in Habbo one can be a sexy blond or a blackish goth, even if reality would be far from that. And it is always so sweet, if 94’ers find a habboadventurefriend or something. Oh that youth bliss”. (translated by the authors from Finnish to English from Ferithem, 2005). Another aspect that undermines the significance of gothic performance in Habbo is the fact that the player can only choose the name, appearance, and gender of his or her avatar, and the aesthetics of the virtual world resemble those of highly abstracted and stylised Lego or Duplo rather than the more generic fantasy-themed MMOGs. This naturally restricts the means of visual expression. In Habbo, being a goth is more likely to be signalled by writing something intriguing and provoking in the one-line avatar description than mere looks. In the more general gothic subculture, a certain erotic appeal, particularly the kind concentrating on an active woman, or the “femme fatale,” is clearly visible. Men and masculine characters such as vampires are also often eroticized. In the world of Habbo this kind of simple aesthetics-based erotic charging of characters is not possible because the looks of Habbo avatars are more based on the appearances of children or young teens instead of grown-ups. This directs the mind more towards the innocence of youth, from which follows the fact that potential sexual and violent themes are much (and understandably) less prevalent. Gothic Ambience in Game Space Habbo players can create a room of which the style can be categorized as gothic, and this can be done rather effortlessly by using furniture from the lines such as the “Gothic” or “Halloween.” Along with “Valentine’s Day,” “Easter,” and “Christmas,” Habbo also launched furniture for Halloween in 2002–2003 (Habborator, n.d.). In response to gothic player activities, a Gothic line of furniture was launched in 2007. Figure 3 shows three examples of Halloween furniture (left), and three pieces from the more recent Gothic series (right).

Figure 3. Examples from two lines of gothic furniture in Habbo. (Habborator, n.d.)

The creation of gothic-themed rooms has been rather popular in Habbo, as there are more rooms with words like “gothic” and “goth” in the title than the fifty rooms that can be displayed in the room search function. Most of the rooms found using general search terms in the room search function, including most rooms named gothic-something, appear as empty or abandoned,

13


which indicates that they are more like archaeological relics than actual avatar habitats. Players have been creating gothic items and furnishings in Habbo, and the traces of this activity can be found in the room names and descriptions. Some of these rooms have preserved decorations, even though no avatars are visible. Figure 4 portrays a room made by Cosmo85 called ”Gothic Mansion”:

Figure 4. A Gothic Room in Habbo Finland (cosmo85, n.d.).

This room’s decoration is gothic in many visible ways. The candles, the skulls, the bats, and the red-black-gray background colors with white as contrast give the room a mystical atmosphere. The bed and the red chairs reserved for ”the audience” along the walls, as well as the use of the symbols of love – roses, the heart, and Amor – suggests transcendent rituals. Tension is in the air as the visitor clicks the post-it note above the door with the heart shape, and a poem pops up. This poem at first sight appears to be about rather straight-forwardly gothic thematics of blood, poison, love, and death; the lyrics of Alice Cooper’s mega-hit Poison (1989) emerge from the words. But the question surfaces: how can this hard rocker be considered gothic in Habbo and not too “old,” mainstream, or too big a star to be approved by real Habbo goths? The most important influences to contemporary gothic rock music come from the late 1970s post-punk music, where a particular musical style began to play a big role. The pale

14


representatives of gothic rock and ”death rock” dressed darkly and created their shows on stage with impressive horror film elements and references. Smoke machines, spider webs, rubber bats and other animal replicas, as well as various edged weapons began to form the cornerstones of the style (Baddeley 2006). For instance, the nightclub show by rock group Bauhaus on the piece ”Bela Lugosi’s Dead” in the vampire movie The Hunger (1983) is a poignant example of this early gothic style. An important characteristic of the early gothic rock was however the ironic play with horror themes. Bloody stage performances by Black Sabbath, where some ”night creatures” had their heads bitten off, cannot be considered as anything else except a joyfully horrific provocation. Alice Cooper and Poison belong to this second gothic wave from the early 1980s. In the music video, Alice appears from smoke in the middle of young, half-naked girls with a wicked expression in his eyelined eyes. The Habbo player’s connection between gothic romanticism and Alice Cooper repeats the same ironic vibe that has been very typical to gothic culture, especially in its early stages. Nevertheless, the question of what constitutes “real” gothic music is still a serious issue (Goodlad & Bibby, 2007), as there have been fierce debates over what counts as credible music styles also in Habbo. For instance, in Finnish advice for goths (Goth 101, n.d.) , it is mentioned that one should despise metal rock because it is not authentically goth, so it is debatable whether the ”Gothic Mansion” room represents any approved gothic Habbo style or not. Actually, the fact that we recognize the lyrics of Alice Cooper’s song is an indication of its slippage into mainstream culture. It also hints that this room might be created with a twinkle in the eye. It is very possible that the room decorator wanted to try out a mystical theme to attract visitors just for chit-chat. This interpretation is supported by the fact that the room creator was one of the Finnish Habbo’s most famous celebrities for several years. Gothic Behavior and Play in Habbo As we are interested in finding out how goth is being acted out and performed, we can not limit our observation of the gothic style to the surface; we also need to look beyond. Even though gothic aesthetics are distinguishable in virtual worlds, they cannot be reduced to visual style alone. Habbo allows for player-created gothic environments and avatars, but it is still within the limits of the game environment and the imposed rules. What could then be termed as gothic player behavior and performance in virtual worlds such as Habbo? Are there characteristics that prevail in other virtual worlds including built-in “gothic” features, such as causing problems for other avatars (griefing), torture and even the possibility of death? Dealing with gothic themes and twisting something to represent the gothic sensibility is a way of handling difficult topics, such as violence, disease, and death. By re-enacting gothic rituals, players seem to explore and contemplate their own emotional responses to these issues. There are rooms and groups dealing with both heaven and hell and even death-themed spaces. Figure 5, from player THEBLOWFISH, shows a Habbo version of a funeral parlour, where one player lies on a bed pretending to be dead and another player enacts the role of a bereaved visitor.

15


Figure 5. Funeral Parlour in Habbo UK (THEBLOWFISH, 2008).

When comparing the use of gothic aesthetics in Habbo to other online forums, the rather unique characteristic of virtual worlds like Habbo is made visible: since it is a game environment, players can create gothic items, objects, and sceneries for playful interaction. Even though the hotel is situated indoors, the popular Halloween furniture series includes a grave item, from which – when clicked upon – a skeleton emerges. As an example, Figure 6 shows a creative use of the grave furniture item: by putting several items besides each other, an interesting masseffect can be created.

Figure 6. Massive Use of the Grave Furniture Item Showing its Skeleton Animation (UNDROZE4VALO80, 2008).

In general, gothic player behavior in Habbo seems to be of the more benevolent kind. Like other online forums, Habbo provides opportunities for goths and wannabe-goths to discuss topics such as gothic books, films, music, and style. Even though what is considered gothic is debatable, this debate also takes place in Habbo. Habbo players can create discussion groups to discuss topics close to their hearts. In Habbo UK, there are close to thirty discussion groups that are tagged gothic. Figure 7 shows the tags that are related to the tag “gothic” in Habbo, which gives an idea of what counts as gothic in Habbo. Active Habbos have also created fan groups for gothic musicians and bands like Marilyn Manson, Gothminister, Turmion Kätilöt, Dragonforce, and others.

16


Figure 7. Gothic Related Tags in Habbo UK (Habbo UK, 2008).

Gothic Habbos also regularly make gothic versions of the established player activities in Habbo. Typical Habbo activities include sports, contests, TV-show imitations, and so on, as well as boarding schools and mafias. A boarding school imitates a school environment, where the players play teachers and pupils, where as a mafia is a hierarchically strict group with playerinvented ranks and missions. Figure 8 shows an example of a vampire boarding school that, in contrast to most, does not require special school uniforms.

Figure 8. Vampire Boarding School in Habbo UK (megask8tr, 2008).

It can be argued that the core of performing goth consists of role-play. A gothic room interior and avatars representing the gothic style is all that is needed for gothic role play in Habbo. Perhaps the biggest and oldest role-playing subcommunity in Habbo Finland is Enelya (Enelya, 2008). It has been active since 2002, has 1,055 members, and essentially consists of hundreds of player-created rooms. The theme of Enelya is medieval fantasy, and all participating Habbos play a specific race, the specificities of which are made up by the game master and role-players. Figure 9 shows digitally edited screenshots of some of the available races: humans, demons, undeads, mutants, giants, dwarves, and fairies.

17


Figure 9. Digitally Edited Screenshots from Habbo Showing Player-created Races in a Player-created Roleplaying Game (Enelya, 2008).

Role-playing a goth also seems to be one way of avoiding annoying pickup lines in chat environments in Habbo. One writer explained how the gothic style communicates “don’t come near me” to other visitors. This strategy might work well; however, for many, Habbo is first and foremost a chat, which renders this “don’t come near me” attitude as strange and unusual. This topic is discussed on a British fansite2 by Habbo-reporter .:Luna-Lovegood:.. As she entered a gothic room, sat down beside another avatar, and asked ”Hi, how are you?,” the response was ”Go away.” This was repeated as she moved to talk to another group: — Hi how are you? No reply. — hello? Still no reply. — How are you? Then one of them said: — Cant you guess, we’re off people (.:Luna-Lovegood:., 2006).

After the experience, .:Luna-Lovegood:. was left wondering why these players were in the chat in the first place. It can be suggested that this is an example of the ways the very basic gothic principles of “anti-behavior” are performed and reproduced: the point of resorting to only talking within a small group of players was to be noticed, to be alone together with like-minded people, and to provoke mainstream players. In this excerpt, Luna seems to represent the mainstream to the gothic players, wondering about “these goths”, who were successful in their subtle provocations. Conclusions In this article we have analyzed various forms of online gothic style in avatars, virtual objects, and places. Players of MMOGs usually use the available means to achieve the typical goth appearance of pale skin and dark clothes for their avatars. The reproduction and transformation of the gothic style is also rather common in Habbo, although this MMOG is initially designed to support benevolent and playful online behavior. 2

Habbo Paper used to live at http://www.habbopaper.co.uk, but has transformed into a Habbo group on the Habbo website. The article is not online anymore, but traces of .:Luna-Lovegood:.’s “Habbo, a chat site?” of it can be found in the Internet Archive.

18


Furthermore, what we discovered is that performing goth online is not just based on a visual style, as it is also a way of playing in Habbo. Being a goth in Habbo can mean different ways of being noticed and being alone together with like-minded players, as well as provoking the more mainstream players. Actively discouraging contact is one such norm-breaker, and another example is the gathering of “off people” players in a room, where everyone sits silently in the chat. As we established, goths in Habbo play an interesting role in the development of the virtual world, as well as in the maintenance of it as a platform for social interaction. From the player’s point of view, Habbo is a place where gothic players can find the company of other gothic players. The player-created gothic scenery – avatars and rooms in gothic style – provide a fun environment for the interaction. Gothic events, rituals, and groups structure the interaction and enable participation in something bigger than just a conversation. Talking about gothic topics can be fun or ironic for some gothic players, but it is also a way of dealing with personally touching and emotionally charged topics. What Habbo goths have from the developers’ point of view is a specific role in the innovation process. Goth players provide the kind of content for the virtual world that interests many teenagers and which can be further utilized by the developer company. Some gothic players are consumers – they pay real money for virtual furniture – but there is more to it. For example, gothic players started out by using the “Halloween” furniture line of candles, skulls, and bats for their own purposes. The developers noted the popularity of the “Halloween” line, and in 2007, they incorporated parts of the gothic subculture into the core of Habbo. A gothic line of furniture emerged as a set of its own, which shows the innovative impact of gothic players. Without the content and feedback provided by gothic Habbos, the gothic line of furniture would probably not have been launched by the developers. These two perspectives on goth subculture in Habbo make an interesting duality. On the one hand, Habbo goths have created a provocative anti-mainstream playing style; on the other hand, the developers have managed to turn it into a business benefit. This incorporation of subcultural activities has not meant the end of playing goth, as early subculture theory by Hebdige (1979/1988) suggests. On the contrary, the gothic subculture is still vibrant. The gothic subculture in Habbo has survived commercialization, which is in line with recent research on the new relationships between gothic subculture and the mainstream (Goodlad & Bibby, 2007). However, in contrast to previous research where gothic discussion groups and blogs are described, the boundary processes are different in virtual worlds. Hodkinson (2002, 2006) describes these as online places where few outsiders intentionally or accidentally appear. In contrast, Habbo is a virtual dwelling place with meeting points for everyone, which puts the intermingling of goths and non-goths more in the foreground. The chance that gothic players meet non-gothic players is much greater than it is in general discussion groups or blogs. Gothic play is an interesting challenge for player models of virtual worlds. Research on player motivations typically mentions motivational factors such as achievement, socializing, and exploring/immersion behind player participation online (Yee, 2006). Most goth players would probably score low on these mainstream motivational factors because of their anti-social, but still at many times pacifist, behaviour. Because of goths provoking, but not actively disturbing style, they cannot be described as typical grief players either. Further work is needed to develop player motivation models that incorporate also the anti-mainstream, such as gothic, behaviour.

19


It is also worth noting that gothic players in Habbo are neither disruptors nor a kind of grief players. Grief players intentionally disturb other players and disrupt their play patterns. Sure, some goths in Habbo aim to provoke, but it is not their goal to actively disturb others, as the “off-people” example above illustrated. Neither is gothic just an aesthetic style in Habbo. Gothic is performance and play, as all the above-discussed gothic examples of the Habbo activities illustrate. For some players, exploring gothic themes and topics may be a way of experimenting with the “dark side” in a safe environment. For others, it is a way of tackling difficult and emotionally-loaded issues. And for some, the gothic sensibility can provide important mechanisms for self-expression and exploration of the facets of identity. All of these forms of player behavior contribute to the social dynamics and the re(development) processes of the online virtual world itself. Understanding them is key in deciphering how and why virtual worlds such as Habbo function so well in the way they do.

Acknowledgements We wish to thank Sulake Corporation Oy and the Mobile Content Communities research project for enabling this study, as well as Sampsa Hyysalo, Asko Lehmuskallio, Mikko Rask, and Petteri Repo for constructive comments.

20


Bibliography Baddeley, G. (2006). Goth chic: A connoisseur's guide to dark culture (2nd ed.). London: Plexus. Calliope. (n.d.). Pukeutumistyylit. Kriisipalvelu.net. Retrieved January 12, 2009 from http://web.archive.org/web/20060625121937/kriisipalvelu.net/index.php?s=artikkelit/puke utumistyylit&p=php cosmo85. (n.d.). Gothic Mansion, Habbo Room. Retrieved March 22, 2006 from http://www.habbo.fi/ Davenport-Hines, R. (1998). Gothic: Four hundred years of excess, horror, evil and ruin. Fourth Estate. Ducheneaut, N., Yee, N., Nickell, E., & Moore, R. J. (2006). "Alone together?": Exploring the social dynamics of massively multiplayer online games. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. Montréal, Canada: ACM, pp. 407416. Enelya. (2008). Enelya. Retrieved January 12, 2009 from http://koti.mbnet.fi/evk/ Ferithem. (2005). Habbo Hotel Kultakala [Msg 54]. Message posted to FFfin Foorumit archived at http://foorumit.fffin.com/archive/index.php/t-10549.html Fischer, G. (2002). Beyond Couch Potatoes. First Monday, 7. Retrieved from http://firstmonday.org Foo, C. Y., & Koivisto, E. M. I. (2004). Defining grief play in MMORPGs: player and developer perceptions. In Proceedings of the 2004 ACM SIGCHI nternational Conference on Advances in Computer Entertainment Technology. Singapore: ACM, pp. 245-250. Global Gothic. (n.d.). Events. CFP: Globalisation and the Gothic. Retrieved January 12, 2009 from http://www.globalgothic.stir.ac.uk/show_event.php?id=8&content=27 Goodlad, L. M. E., & Bibby, M. (2007). Goth: Undead subculture. Durham: Duke University Press. Goth 101. (n.d.). Goth 101. Retrieved August 31, 2008 from http://www.tribar.net/g101/ Gothic Second Life. (n.d.). Flickr Group. Retrieved January 12, 2009 from http://www.flickr.com/groups/363644@N20/ GothicaTheLost. (n.d.). Habbo home. Retrieved January 12, 2009, from http://www.habbo.com/home/GothicaTheLost

21


Habbo Paper. (2006). News, editorials, interviews, competitions and much more! Retrieved January 12, 2009 from http://web.archive.org/web/20060614193708/www.habbopaper.co.uk/articles2.php Habbo UK. (n.d.). Habbo pets. Retrieved January 12, 2009 from http://www.habbo.co.uk/help/2 Habborator. (n.d.). Habbo hotel international furniture charts. Retrieved January 12, 2009 from http://www.habborator.org/furniture/index.html Habbox Forum. (n.d.). Show off your website here - including Habbo fansites (forum thread). Retrieved January 12, 2009 from http://www.habboxforum.com/forumdisplay.php?s=d757 4da77151c4e53f0cc8cae37acf11&f=45 Hebdige, D. (1979/1988). Subculture: The meaning of style. London: Routledge. Hodkinson, P. (2002). Goth: Identity, style and subculture, dress, body, culture. Oxford: Berg Publishers. Hodkinson, P. (2006). Subcultural blogging: Online journals and group involvement among UK goths. In A. Bruns & J. Jacobs (Eds.), Uses of blogs. New York: Peter Lang, pp.187-197. Jenkins, H. (2008). Convergence culture: Where old and new media collide. New York: NYU Press. Johnson, M. (2007). Unscrambling the “average user” of Habbo hotel. Human Technology, 3, pp. 127-153. Johnson, M., & Toiskallio, K. (2005). Fansites as sources for user research: Case Habbo hotel. In 28th Information Systems Research Seminar in Scandinavia. Kristiansand, Norge. Retrieved January 12, 2009, from http://www.hiit.fi/u/johnson Johnson, M., & Toiskallio, K. (2007). Who are the users of Habbo hotel? In Mobile Content Communities. HIIT Publications 2007-1, p. 89-107. Retrieved from http://www.hiit.fi/publications King, A. (2008). Average web page size triples since 2003. Retrieved January 12, 2009 from http://www.websiteoptimization.com/speed/tweak/average-web-page megask8tr. (2008). Vampire boarding school, Habbo room. Retrieved November 7, 2008 from http://www.habbo.co.uk Messinger, P. R., Stroulia, E., & Lyons, K. (2008). A typology of virtual worlds: Historical overview and future directions. Journal of Virtual Worlds Research, 1. Retrieved from http://jvwresearch.org Nerokala. (n.d.). Web community. Retrieved January 12, 2009, from http://nerokala.com

22


Sihvonen, T. (2009). Players unleashed! Modding The Sims and game culture. Forthcoming PhD dissertation, Faculty of Humanities, University of Turku, Finland. Silverstone, R., & Hirsch, E. (Eds.) (1992/1994). Consuming technologies: Media and information in domestic spaces. London: Routledge. Sotamaa, O. (2007). Let me take you to the movies: Productive players, commodification and Transformative Play. Convergence, 4, pp. 383-401. Sulake. (n.d.). Habbo - Where else? Retrieved January 12, 2009 from http://www.sulake.com/habbo Sulake. (n.d.). Corporation website and press releases. Retrieved January 12, 2009 from http://www.sulake.com Sulake. (n.d.). Habbo Way (US). Retrieved January 12, 2009 from http://www.habbo.com/help/51 Taylor, T. (2006). Beyond management: Considering participatory design and governance in player culture. First Monday, Special Issue 7. Retrieved from http://firstmonday.org THEBLOWFISH. (2008). Habbo funeral parlour, Habbo room. Retrieved November 7, 2008 from http://www.habbo.co.uk UNDROZE4VALO80. (2008). Habbo skulls!!! WiCkEd! Retrieved January 12, 2009 from http://www.youtube.com/watch?v=onqQPAuTDV0 Vampires. (n.d.). Facebook application. Retrieved January 12, 2009 from http://www.facebook.com/apps/application.php?id=2458301688 Waldron, D. (2005). Role-Playing games and the christian right: Community formation in response to a moral panic. The Journal of Religion and Popular Culture, IX, pp. 50-78. Wikipedia. (n.d.). Griefer. Retrieved January 12, 2009 from http://en.wikipedia.org/wiki/Griefer Wikipedia. (n.d.). Nightwish. Retrieved January 12, 2009 from http://en.wikipedia.org/wiki/Nightwish XXXDominatrix. (n.d.). Habbo home. Retrieved September 1, 2008 from http://www.habbo.com/home/XXXDominatrix Yee, N. (2006). Motivations for play in online games. CyberPsychology & Behavior, 9, pp. 772775.

23

ARTICLE

VI


Mikael Johnson (2010). "User Involvement, Social Media, and Service Evolution: The Case of Habbo". 43rd Hawaii International Conference on System Sciences. Kauai, Hawaii, January 2010, 1–10, (Nominated for Best Paper Award). Reprinted with permission.

287

Proceedings of the 43rd Hawaii International Conference on System Sciences - 2010

User Involvement, Social Media, and Service Evolution: The Case of Habbo Mikael Johnson National Consumer Research Centre [email protected]

Understanding user needs and involving users are key success factors for service development. However, it is unclear how to apply old wisdoms from traditional product development settings in a social media context. This article analyses user involvement practices in a longitudinal case study of the virtual world and social networking service called Habbo. The focus is on how particular social media aspects, such as persistent user-created content and developers’ easy access to online user action, shaped user involvement and how it evolved over time. In line with previous research, the findings indicate that user involvement becomes more formal as the user community grows and the servicemarket combination matures. In addition, user involvement regarding social media appears to be an issue of timing, trans-nationality, and governance. The results suggest that the usability practitioner’s ‘toolbox’ needs tuning regarding social media.

the service, and after the research project the author has informally followed the case. In this article, the focus is not on the minute details of a particular user involvement method, but more on a strategic level. During ten years, Sulake developers have explored various user involvement methods ranging from informal observation, focus groups, market surveys, and new social media possibilities. The aim of the case study was to answer the question of what practices social media developers have regarding user involvement. In this article, we focus on two aspects in more detail. First, how the social media context influences the possibilities for user involvement, considering especially persistent content created by users and developers’ easy access to online user action. Second, how user involvement practices change as the service evolves, with particular interest in the increased complexity, signified by the increasing numbers of both users and developers.

1. Introduction

2. User Involvement Approaches

Involving users in service development is a key strategy to build better services and democratize the innovation process [1-3]. While there are mixed results on the cost-effectiveness of user involvement [4], there is an abundance of methods available to bridge the gap between developers and users [5-9]. Many user involvement methods have been created and applied in contexts like commercial development for offices, government health care, but few consider the emerging social media. Is the standard textbook advice ‘study first, design second’ applicable in social media development, or is it the other way around [10]? This article gives an overview of user involvement practices from a longitudinal case study: the virtual world and social networking service called Habbo that was launched in 2000. The Habbo producer, Sulake Corporation, was involved in a research project in 2003-2006, during which the author followed the user involvement approaches used. In retrospective interviews developers shed light on the first few years of

There are many arguments for user involvement: reducing risks of failure in technological innovations [1], democratizing the innovation process [2], building better products, etc. Despite a consensus that understanding user needs is a key factor in commercial success [3], this is not easily translated into action in a particular innovation context. There is an abundance of methods to solve the ‘sticky information’ problem: information about users’ needs and developers’ capabilities is highly contextual, tacit and difficult to transfer [2]. However, which user involvement approach is applicable in which particular design situation is still an open question. In the same way as using a particular research method implies a particular role for a researcher, any user involvement approach imply a particular userdeveloper relationship. A researcher can be anything between a detached and uninvolved observer to a fully engaged participant in a group or organisation under investigation. Developers’ information about users

Abstract

978-0-7695-3869-3/10 $26.00 © 2010 IEEE

1


Table 1: Pros and cons of user involvement methods [12]. Approach to learning about end-users Surveys and interviews

Pros

Cons

Systematic approach to data collection Surveys provide the possibility to poll representative samples

Prior research, particular theoretical perspectives

Sound theoretical base can guide observations and help to make sense of energy-related behaviour

Experience from prior projects and similar examples User-driven project (or pilot project)

Sound experience creates confidence and practical skills/solutions that are difficult to codify End-users know about their needs and circumstances and can contribute to contexttailored and user-friendly designs End-users are motivated and engaged from the start, thus ‘less work’ is left for the programme manager. Informal interactions allow for a rich exchange of information (including non-verbal communication) Familiarity creates trust and mutual confidence

May not always feed into programme design Surveys may be designed to confirm existing preconceptions, may fail to bring up new insights Conducting good research may be expensive and require specialized skills Strong commitment to prior findings or theories may lead to overlooking contextual particularities Overly theoretical background can lead to complex and confusing designs ‘Competence trap’: overconfidence and failure to learn new skills in new contexts

Familiarity and informal interaction with the target group

originates in similar relations: statistics of user practices can easily be detached from a particular situation, aggregated and analysed from a distance, where as participant observation of situated use implies that actual developers and users meet. In the tradition of designing for usability, developers refer to these situations as usability evaluation with or without users, and methods have emerged that are tuned to diverse design situations [11,5,6]. Table 1 lists pros and cons of particular approaches to user involvement, taken from experiences in energy demand-side programmes. User involvement methods for established markets differ from the methods that are developed for new markets. Leonard-Barton [3] found that traditional market research methods (latent needs analysis, lead users, surveys, focus groups, mall studies) fit well with an established market and a high maturity of technology design. For new markets, more used were methods like market experimentation, future scenarios, and trends exploration. In the grey area between these typical situations, Leonard-Burton argued for less traditional market research methods under the rubric of empathic design: studying actual observed customer behaviour, direct interaction between developers and users, and considering existing technological capabilities. Empathic, participatory, and other user-centred design approaches emerged because user requirements engineering techniques at the time lacked an under-

End-users may not be fully aware of their behaviour and all the factors underlying it ‘Upscaling’ from small user-driven pilots to broader group of end-users can be difficult

It can take a lot of time and commitment to build up the level of familiarity needed to execute a successful programme Contacts may be biased: some end-users are more familiar than others.

standing of the culture and practices of users [12-14]. However, while the initiatives recognise the complexity and diversity of user settings, they seem to have had modest influence over system design overall [15]. Stewart & Williams [16] warn against a pitfall they call the design fallacy: the assumption that from the diversity of user needs follows that extensive knowledge about user contexts must be built in the product. In contrast, they report that generic products that can be customised and configured seem to have had more impact on system design than user-centred design [15]. There is no consensus on why potential initiatives such as user-centred design have had little impact. Recent debates suggest that standard methods such as ethnographic field studies or usability evaluations can do more harm than good if not applied carefully [17,18]. This reflective debate can be interpreted as a sign of maturity in the human-computer interaction field as it highlights specific problems that must be addressed when mixing particular research disciplines with product development. Another suggestion is that the boundary conditions of product development projects have not been addressed thoroughly. Svanæs and Gulliksen [19] suggest that the design context needs greater attention, that is, the internal structure of the developer and the client organizations, contractual and tender issues, software engineering tools, and stakeholder agendas and relations.

2


2.1. Social Media New online web services often referred to as digital media, web 2.0, and social media have changed the possibilities for user involvement. On the one hand, people discuss particular products and brands in online discussion forums and communities. Product developers can more or less openly visit these forums to get customer feedback. On the other hand, when people use web 2.0 services, they leave an electronic trail. What people click on, what they search for, what they select, and how they navigate on a web site can all be collected and aggregated. Service developers can use this log statistics to fine-tune the service. Furthermore, in some services it is customary for users to openly share their preferences, bookmarks, opinions, and other content with other users. This participative web and user-created content [20] sometimes plays a major role in the service and can also be used by service developers as inspiration for new features. Social media is clearly not only win-win situations, since many discussions on privacy, advertising, intellectual property rights, and other conflicts between commercial actors and consumers have emerged. Lehmuskallio et al. [21] observed that these discussions relate to communicational changes in access to, scale, distribution, and persistence of computermediated communication. For instance, web 2.0 business models and privacy are often difficult to combine, as private information can be used to target ads, be aggregated and sold to third parties, as well as get into unexpected hands when big commercial players buy web 2.0 start-up companies. As social media change user-producer relationships, new research questions related to user involvement emerge. Do the short release cycles frequently associated with social media – the infamous ’forever beta’ approach – give more or less room for user feedback? How do social media developers apply user involvement methods as the service evolves over time? In a discussion of community building, Kim [22] suggests that many online communities start with the developers as central figures in the community, but as the community grows, the developers’ roles and influence decrease. Is such a change in the developeruser relationship reflected in the use of user involvement methods?

3. Case Habbo Habbo is a virtual environment where children and teenagers meet, socialize, and play many types of games. It was first launched in August 2000 in Finland as Hotelli Kultakala (‘Hotel Goldfish’) and it was based on the developers’ two earlier online services. At

the time of writing, there are Habbo hotels in more than thirty countries, and 13 million players visit Habbo each month [23]. Instead of an entrance or a monthly fee, the profit model is based on micropayments in the hotel. Virtual furniture, mini-games, and membership in the Habbo club are bought with Habbo credits. These credits can be purchased (depending on the country) with pre-paid cards, bank transactions, credit cards, or special text messages that add a specified amount of money to the customer’s mobile phone bill. The social interaction in Habbo is truly diverse. In the design of Habbo, clear winning conditions and gameplay rules have been avoided, and instead, players are encouraged to create their own objectives beyond chatting, room decoration, and meeting friends. The provided environment for these activities is a hotel consisting of public and private rooms, where the virtual hotel visitors, called Habbos, chat, buy virtual furniture, decorate rooms, and arrange social and game events. Most of the teenage players log on after school, and according to Sulake, the developer company, on average they spend around forty-five minutes per day in the hotel or on its related discussion forums.

3.1. Data and Method The Habbo producers, Sulake, was established in 2000 and focuses on virtual worlds and social networking. Sulake’s strategy is to be a leader in community based entertainment with a portfolio of properties addressing a wide range of target audiences. During 2003-2006, Sulake, together with ten other commercial actors of the Finnish gaming industry, participated in a research project coordinated by the Helsinki Institute of Information Technology [24]. This research context made different design research studies and collaboration with Sulake’s product development possible. The focus of the Habbo study was user involvement strategies in the design and management of social media. Since its launch, the user communities have participated actively in the shaping of Habbo [25,26]. Some as active playmakers [27] and others as equally important participants and audience. The author had the opportunity to conduct both quantitative and qualitative research to understand the Finnish Habbo communities. The project started with participant observation in Habbo, pilot interviews, a community manager survey (N=4), as well as an explorative survey (June 2004) on the visitor profiles (N=10000) as Sulake’s first global youth survey was published two years later. The author participated in Habbo, explored the features and their affordances, analysed default values and users’ degrees of freedom as a routine when new features were launched. The

3


visits to Habbo occurred on the average about twice per month throughout the first three years of research, and a visit every second month from 2007 onwards. Fansites became an important source of knowledge about the user communities. In 2004, Finnish Habbo fansites were analyzed in detail [28]. Whereas the survey findings provided background statistics, the fansites and forum discussions allowed an insight into active user groups, popular activities, and hotel history. Since its very beginning Habbo fans have kept several sites devoted to documenting changes in the hotel, for instance in kinds and appearance of furniture, characters, mini-games, and bugs. In 2005, the author conducted ten thematic 2-3 hour interviews with ten Habbo developers, or about two thirds of the game development organization at the time. Six of the interviewed developers (graphical designers, and client and server developers) had been in the organization since the beginnings, five years earlier, while four developers had about one year of Habbo experience. This was followed by 2-3 hour individual, pair, and group interviews with 11-16 years old users (N=6), and 30+ users (N=6) to focus on their participation histories, motivations, meanings they give to Habbo, and groups they participate in. Habbo has also been a topic of collaboration between the author and other Habbo researchers. A Sulake employee’s master’s thesis on communication and action in Habbo provided important secondary sources to Habbo data, as did several youth work studies1. Usability research students at Helsinki University of Technology also provided secondary data, as part of to their efforts on both a usability test on new 10-12 year old Habbo users (N=8) in 2004, and a software engineering project in 2004-2005 to develop a fansite starter kit for active Habbo users. After these activities, the author had the opportunity to take part in an intervention study with Sulake. For release 9 of Habbo in 2006, a set of user feedback methods were explored with different stakeholders inside Sulake. This included confidential data sets: database statistics and surveys from two countries on the use of a new feature. In addition to these research activities, Sulake representants participated in project partner seminars arranged every 6 months, two workshops on virtual economy, and many project meetings arranged by the research project. These meetings made informal discussions and a continuous dialogue with Sulake possible. Taken together these bodies of data provide an excellent view to the varying forms of dialogue between the users and developers of this virtual world. 1

These are only available in Finnish, but noted here for further inquiries: Pietiläinen 2004, Sihvola 2005, Koskinen 2006, Merikivi 2007.

The data analysis has proceeded in multiple waves over the years, including quantitative and qualitative clustering of Habbo user profiles [29], coding and examining interviews with regard to participation and development histories in Habbo, analysis of fansite contents and developer views to them [25,28], and here, constructing the changes in Habbo user-developer relations over time.

3.2. Habbo Service Evolution Like many other digital startups, Habbo is a moving target for many reasons. In 2003 there were four hotels and 1 million users, but at the time of writing, these numbers are more than tenfold. The producer company has grown from a small group to a 300 people corporation, with local offices in many hotel countries. To aid the analysis process, a division of Habbo service evolution into four stages will be outlined below. Habbo was based on two previous launches, Mobiles Disco (Oct 1999) and a snowball game called Lumisota (Feb 2000). Mobiles Disco provided the basic hotel infrastructure in the pixelated retrostyle, whereas Lumisota featured payments through mobile phone messages, which at the time was a practical solution for the Finnish market. Starting with the first international step to UK, other payment systems were implemented to suit markets with less wide use of mobile phones among teens. Phone call systems, youth cash cards made of paper, check systems came first, but soon followed payment by credit cards, Paypal, and bank transfers. With the UK hotel also followed the teen invasion of Habbo [30]. Habbo was not designed for teens at first and the large number of teen visitors also meant a large number of concerned parents. To keep all this together, the community management received a lot of attention. A set of guidelines for good behaviour was established, called Habbo Way, and tools for governing it were developed. During 2002-2003 Sulake developed its business, the hotel and the community simultaneously. Strategic partnerships were made and brands like Mountain Dew and Britney Spears entered Habbo. Later these brands have disappeared as Sulake tried another approach with Coca-Cola and Warner Cinema. The brands got customised versions of the technical game engine behind Habbo, which helped fund the development of the hotel. The hotel’s technical architecture was stabilized and security improvements were made to package the hotel into a product that could be more easily rolled out in new countries. During 2004-2005 around ten new hotels and local offices were established in different parts of the world.

4


Habbo actively wanted to become the largest teen brand in the world. The English-speaking UK hotel was split into separate hotels for the US, Canada, and Australia. Also game development recruited more developers in 2004, and new features were packaged into simultaneous releases in all hotel countries. Sometimes during 2006 and 2007, the strategies changed again as Sulake entered the social networking market. Not only did Habbo get new social networking features, but Sulake acquired IRC-Galleria, the most popular social networking service in Finland. As IRCGalleria users were a little bit older, average age around 20, the strategy to concentrate on teens was opened up to a variety of age groups. Based on this information, the author has grouped the service evolution into four stages. Table 2. Habbo Service Evolution Stages Concept Development 1999-2000

Beta Testing 2001-2003

Expansion 2004-2005

Complexity Management 2006-

Concept Design refers to the first prototypes in 1999 and 2000: Mobiles Disco, Lumisota, and Hotelli Kultakala, while Beta Testing refers to the time period between 2001-2003, when much of the basic functionality was completed. Expansion refers to 20042005 when the product was packaged so that it made a roll out possible in more than 10 new countries during one year. Complexity Management refers to 2006 and onwards when the product was extended to a social networking service.

4. Habbo User Involvement Methods The user involvement methods reported here attempts to be an overview of the methods used by Sulake from 1999 to 2009. The criteria for what counts as a method is a more or less organized practice where developers learn about the users, which then shapes the service. The user involvement methods are organized according to the service evolution stages.

4.1. Concept Development Avatar activities. Avatar activities refer to what users do online in the service. Developers have easy access to users in the sense that the developers can log on to Habbo and check what is going on. In the same way as users find out what happens where in the hotel, developers can also use the list of most popular rooms or tags to find the trends. In the room, a developer can choose to participate in the ongoing activity and take note of which users are present. The names of the user

avatars and their descriptions reveal what kind of Habbo crowd is present. By studying the arrangement of Habbo furniture and guidelines written on virtual post-it notes, the developer learns about how particular pieces of furniture are used in specific user activities. Developers as users. In the early stages of concept development and beta testing, the developers were developing the service for themselves, their friends and their new media colleagues. The first prototype was a hobby project for a friend’s band. The original group of developers started out as insiders in the user community, but as the service become more popular, other insider groups started to form. After a while, the developers became more distanced to the user community as it became more difficult for them to spend time in Habbo. Each time a Habbo developer’s avatar was online, it did not take long before Habbo newcomers gathered around and wanted to become Habbo friends, or even asked for freebies. Habbo developers soon invented their own tactics to participate in cognito: they created new avatars, whose real identity they only told trusted friends. Informal evaluations. Various informal evaluation practices guided the design early on. In interviews, the developers talked about slogans such as ”easy access, easy play” and ”where else”. A developer described the latter one as ”We won’t make obvious choices, but rather something personal that gives the Habbo world something odd and own, which creates an own persona.” (Interview 19 Apr 2005). The “easy access, easy play” slogan was not only a slogan, but among the developers it also had a visual form called the Habbo Ladder that showed the first steps a new user takes. The Habbo Ladder was made to contrast Habbo with the comparably higher threshold to start playing massively multiplayer online games at the time. The idea was that during the first few minutes that anyone is willing to spend on a new web service, one should be able to login and create an avatar easily, learn the basic navigation, and have a chat with someone else. Each step on the ladder makes more reasons to return to the service, as a developer explained: ”Especially if one gets the first friend on the list of friends, then that it is a reason to return, that you have really got to know someone with whom you might have had an interesting discussion or of whom an interesting image has been conveyed. Then further on, when you have your own room, well that is of course a real investment, even the notion that you have something own going on there, then that is already a good reason to return, especially if you have decorated the room, really purchased something.” (Interview 19 Apr 2005). E-mail feedback. During the first year or so, developers received a lot of e-mail feedback by users,

5


who suggested changes and new features. This e-mail feedback became a handy storage for design inspiration for the developers, who used to return to it later and browse for good ideas. As the user community grew in size, this direct e-mail feedback became impractical as the number of e-mails became too large. Users not only sent feature suggestions per e-mail, but all kinds of support requests, which then as the organisation grew, became the responsibility of a small but growing number of volunteer community moderators and customer service officers. Volunteers. Already before Habbo, in Mobiles Disco, there was a need to keep the shared meeting places in order, in case a user did not follow the shared conventions of appropriate online interaction — the netiquette. Trusted friends of the developers or other insider users received special powers, for instance to remove (kick) a disturbing user avatar from a room, and the responsibility to keep the online place neat. These volunteers followed along from the first prototypes to Hotelli Kultakala and Habbo. Later it became possible to apply to become a volunteer, so called Hobba. However handy early on, and applicable in many hotel countries, this arrangement involved more than 1500 volunteers during the expansion phase in 2005. Sulake changed the volunteering practice in 2005, as the special powers were given to paid moderators only, and the notion of volunteer moderators transformed into volunteer guides, who could guide online newcomers.

4.2. Beta Testing Volunteer forum. To share experiences and moderation policies, the volunteers created an online forum for themselves. Along with the internationalisation and more organised volunteer management, Sulake started hosting a local volunteer forum per hotel country. The volunteers soon got an important role as mediators of user opinions: the developers knew that as the volunteers spent the most time in the hotel, they were always the first to know about the current user concerns, wishes and emergent activities. Weekly newsletter and polls. In an attempt to organise user feedback and create a rhythm of new content, the hotel community management started creating weekly polls and newsletters. The newsletter contained information about new features and events organised by the hotel manager (a Sulake employee). With the newsletter was usually a gallup poll or a competition that the users could participate in. These weekly polls were important to gather immediate user feedback regarding new features. Fansites. Right from the start, Habbo users started writing about and debating Habbo on their own

homepages, and some created totally Habbo-themed websites. Being a Habbo journalist was one way of becoming famous, besides participating in various and creating own online Habbo activities. Some fansites featured discussion forums, which also were an important source for developer inspiration [25,28]. Official Web Fanzine. For two years, Sulake published their own official Hotel Magazine in Finland (www.kultakalankuvalehti.com), but due to limited resources they closed it down. It started out with writings from Habbo developers, and interviews with Habbo users, but soon enough some Habbo users featured as guest authors. In 2004 Sulake launched local competitions in each country to give a few elected large fansites the status of being official Habbo fansites. While these fansites complemented Sulake’s community management, they also helped Sulake in regulating sites that were out to scam the users. Summer Meetings. With the responsibility for teenagers online, Sulake had to develop a policy that discouraged sharing of contact information online and meetings in person. To support this policy, Sulake arranged summer meetings where users could meet each other in person, as part of an organised event. In the first meetings in Helsinki active users and volunteers got to meet the community manager and a few developers. Later on these meetings have been arranged by active users, without much involvement of developers. Sales Statistics. In contrast to other Habbo stakeholders, developers have an additional perspective to the online activities in Habbo. They have a back-end service that keeps track of what pieces of furniture have been sold and how many in which hotel. This means that developers can compare Habbo features on the basis of their economical behaviour, not only based on functional or aesthetic properties. Even though the use of Habbo is anonymous, the registration process asks users to fill in age and gender, which can then be connected to sales statistics on an aggregated level. Customer service. The need for dedicated customer service emerged as the user communities grew and the developers could not find time to handle all support requests. Each hotel country got their own country office, where a local manager managed the volunteer moderators and customer service officers sorted out user requests.

4.3. Expansion Market survey. With the international expansion emerged a need to know whether the user communities were similar or different in different hotel countries. An outsourced market survey in 2004 generated customer segments and their regional distribution. The

6


named customer segments and their demographics were important in the communication about Habbo. Even though demographics seemed less important than user practices from a game development perspective, customer demographics played an important role in business negotiations, regarding advertisement for instance. Focus groups. To prepare for the rapid international expansion that happened during 2004-2005, focus groups were conducted. The applicability of Habbo pixel style graphics and use of colours was evaluated for the Asian market. In addition to traditional market research focus groups, two variants of the focus group method from different research traditions were used, usability evaluation and playability testing. Usability evaluation with users is more like a controlled lab experiment in psychology, where users try out the software with the help of scenarios and evaluators measure issues like usefulness, ease-of-use, learnability, and memorability [31]. Usability Evaluation. The first formal usability evaluation in the fall of 2004 was targeted at checking the usability of service registration and those services in Habbo that were subject to a fee, from the viewpoint of 10-14 year olds. The evaluation was outsourced and conducted by usability students at the local university of technology (TKK). After the first usability evaluation, Sulake established an in-house usability process through a pilot project. The pilot project was related to a product extension for the mobile phone market, and was influenced by usability standards in the field, such as the J2ME guidelines made by Idean Research [32]. Playability testing. While the usability evaluation method was appropriate for features such as login, character creation, furniture purchases, usability methods in general are not tuned to assess playability. For these reasons, Sulake’s R&D lab used playability testing methods developed at Tampere University’s Hypermedia Lab [33] to assess various playability aspects, such as gameplay, game mechanics, appearance, sound, and social playability. CRM system. As some hotel communities grew larger, pressure emerged for customer service to automate their responses. For instance, in a country with several hundreds of thousands of users, a new feature might spawn several thousands of inquiries per day. In 2005, a new customer relationship management system was introduced. It featured a set of standard questions and responses, which reportedly reduced inquiries with 90 percent. Release pilots. As the organic beta testing phase changed into a more controlled release management process, Sulake started piloting the release for one month in one hotel country, before diffusing the release to other hotel countries. This means that users in the

pilot country acted as beta-testers and could comment on the features in the release before they were finalized. The process of selecting the pilot country has changed several times. In 2005, countries with a relatively small number of users acted as pilot countries. However, translation and language become an issue, which is why piloting returned to Finland sometimes around 2007. However, at the time of writing, UK is the pilot country.

4.4. Complexity Management Online user panel. In an effort to gather systematic feedback before the implementation of new features, Sulake recruited 200 volunteers in one country to form an online panel. The online panel was given a weekly task consisting of a set of questions regarding design sketches and an opportunity to share opinions regarding the sketches in a forum. The online panel was popular among developers, who queued to be the one who got to get design-time feedback from the users. Global youth survey. As both the users and competitors change, in 2006 it was again time for a new market survey (GHYS’06). This time the customer segmentation resulted in certain Habbo lifestyles: achievers, creatives, loners, rebels, and traditionals. The report featured extensive countryspecific information on teenagers’ favourite brands and media usage patterns. User, group homepages, tags. In early 2007, Sulake launched social networking features in Habbo as every Habbo avatar got an automatic and customisable Habbo homepage. It was also possible to form Habbo groups, which meant a logo, a group homepage and a discussion forum. In addition, users could ”tag” their avatars, which meant that users could attach a set of clickable one-word descriptors to their avatar. When a user clicks on a tag, the service generates a dynamic index of all the users and groups with that particular tag. User experience testing. Following current research terminology, in 2008 Sulake talked about user experience testing in addition to usability or playability testing. While the usability testing has evolved from a more stand-alone practice, to a tight integration with agile software development, Sulake conducts user experience evaluations with both new and old users, internally called ”live tests”, in one country for every major release, which makes about once a year (Personal communication with User and Market Insight Director, 26 Nov 2008). Personas. As part of further developing usercentred design processes, during Spring of 2009, Sulake applied the Persona method. Six user

7


Table 3. Overview of Habbo user involvement methods. Concept Development

Beta Testing

Expansion

Complexity Management

Avatar activities Developers as users Informal evaluations E-mail feedback Volunteers

Volunteer Forum Weekly newsletter and polls Fansites Official Web Fanzine Summer meetings Sales statistics Customer service

Market survey Focus groups Usability evaluation Playability testing CRM system Release pilots

Online user panel Global youth survey User and Group homepages, tags User Experience Testing Personas

archetypes have been constructed from data to represent the users. The idea is that developers have an updated reference to the goals and needs of Habbo users at hand, which can inform design solutions and evaluations. (Personal communication with User and Market Insight Specialist, 2 Jun 2009).

5. Discussion 5.1. User Involvement and Service Evolution Based on the Habbo case it seems evident that very few service characteristics relevant to user involvement methods remain invariant over time. Some characteristics remained stable: the service concept and the profit model (after small experiments during the first few years), the technical access to online user activities, as well as the service domain and (life-) criticality of the interface. However, a lot has changed over ten years: at least the number of users, the number of developers and resources available, the amount of money involved, competition on the market, the stage of design, the skills and experience of the developers. As the maturity of the market changed, so did the relationship between developers and users, much like in previous research by Leonard-Barton [3] and Kim [22]. In the Habbo case, like with most social media applications, the developers started developing the service for themselves and their friends. While starting out as insiders in the user community, with commercial success, the developers inevitably end up as excluded from the user community, if for no other reason, just by the sheer number of users and their practices. The first usability study and succeeding studies marked two changes in user-developer relations. First, as the number of users increased, the margins for errors in the service were decreased, thus more formal methods were valued. Second, developers’ gut feeling, that is, undocumented previous development experience and participation in the user community, carried less weight, as not everyone in the growing company could be involved in the user communities. Business and development decisions needed more documented

knowledge, which was provided by the use of more formal methods. So far the results from this case study is much in line with previous research. However, the user involvement practices in this case study stands in contrast with the strategy of developing a method and then using it in the same repeatable way over and over. In this case it has been evident that the history of method usage shapes the succeeding use of methods. For instance, as the service as a whole had been usability evaluated, the usability specialists at Sulake found no sense in repeating the same evaluation again. Usability evaluations turned towards smaller details of the service after the overall evaluation. The same finding was found in the market survey practices. After the first overall customer segmentation, the following surveys could fine-tune specific issues that remained open in the previous surveys. This practice can be interpreted as a sensible ongoing tailoring of methods that becomes necessary when dealing with as complex and changing phenomena as in this case.

5.2. User Involvement and Social Media It seems that the developers have benefited from particular user involvement opportunities related to social media. User created fansites have been an important source for developer inspiration, and the online user panel is an interesting way of making prolonged contact with a particular user group, which complemented both the more informal feedback that developers get in the service itself and the more formal surveys. The social networking features – user and group homepages as well as tagging – provided Habbo users with new possibilities to express themselves within the service and find other interesting users and groups. On the other hand, these social networking features made it easier for Sulake to get an overview of group activities in Habbo as well as search and aggregate data on users. Let us return to one question posed in the beginning. What happens with user feedback in a “forever beta” approach, is there more or less room for it compared to traditional product development?

8


In the Habbo case, it seems that user feedback can shape the service most when the design space is open. How open the design space is, depends on what is going on in development. During concept development and early beta testing Habbo was fairly open for feature suggestions, but it did not take many months for the developers to get more ideas than they could handle for the foreseeable future. Then the question was no longer about getting good ideas, but rather about finding projects and funding to implement the ideas and stabilize the service. And, until the service was stabilized and packaged enough for manageable duplication in the expansion phase, the design space was fairly closed. From a Finnish user point of view, the next window of opportunity to shape the service was probably the implementation of the social networking features in early 2007, which again changed the rules of the game. Then as the new features were tried out, user feedback again helped fine-tune the service. User feedback varies between Habbo hotel countries. Like with any trans-national service, Habbo developers experience the dilemmas of language regions and differently sized markets. On the one hand, one could argue that, because of Habbo’s Finnish origins, hotels with close linguistic ties to strong languages in Finland would shape the service most. However, there are two factors counteracting this simple argument. First, language skills and effective communication among Sulake employees can mediate remote user needs. Second, since some user needs are readily communicated through sales statistics, the market size of a particular hotel country might be more significant than the language in that country. In a broader perspective, ”what happens with user feedback in social media?” is also a question about governance, that is what users get to decide on and the form of their influence. Typically users have less influence on generic products and more influence on customisable products. The impact of user feedback is especially big if the users themselves take the initiative to new development. In contrast, social media is not usually developed as a contract between a developer and a client organisation, but rather on the initiative of a handful of developers. Habbo users have had a big influence on early feature development and issues like community moderation (as volunteers), online discussion about Habbo (as fansite authors), but have never been invited to decide on pricing, advertisements, or feature roadmapping. This is not to say that they should be, it is just a fact that needs to be stated when outlining the Habbo user-developer relationship. Edemocracy in Habbo has not yet reached the extent of the governance experiments in virtual worlds such as EVE Online or A Tale in the Desert.

6. Conclusions In addition to the themes discussed in research literature on user involvement, user involvement regarding social media appears to be an issue of timing, trans-nationality, and governance. It is a timing issue, because user feedback influences most the features under development at a particular point in time. It is a question of trans-nationality, because user feedback from different countries is influenced by both language and market size. It is a question of governance, because user feedback might also address issues that are not considered open for discussion by the developers. Sulake’s user involvement practices reveal a bias in common classifications of user involvement methods. Traditional formal methods and usability evaluations are only one small part of the big picture regarding learning about social media users. The 23 [!] different modes of user involvement reported in this study suggest that the usability practitioner’s ‘toolbox’ needs tuning regarding social media. This case also reveals how the nature of user involvement has changed with social media. Before social media, ‘involving users’ has been a much more explicit effort, as developers and researchers have had to go where the users are and ask permission to observe and get feedback. While some ways of involving social media users remain as indirect as before, feedback from social media users is at the same time genuinely direct. The very thing people do in and return to the service for, can be interpreted as user feedback. The users might not think that they are being ‘involved’ or formally ‘invited’ – they will just not return if the service does not meet their needs. However, previously it took longer before a decision by a user to stop using a product or service reached the producers. This ‘direct feedback’ from social media users, in terms of frequent data points of user actions, has a major consequence to user studies. The techniques commonly denoted as evaluation ‘with’ and ‘without’ users need to be supplemented with ‘evaluation with database access to users’. As both use and non-use is logged in the server database, alarms can be set to trigger if use practices change. These alarms can then be used as starting points for succeeding user studies to explain the changes in use practices. How strange, or perhaps disciplinary inconvenient, it may sound, the experience from this case study suggests that acquisition and analysis of database logs should take place before embarking on either quantitative or qualitative studies on social media use.

9


7. References [1] A.L. Friedman and D.S. Cornford, Computer Systems Development: History Organization and Implementation, John Wiley & Sons, 1989. [2] E.V. Hippel, Democratizing Innovation, The MIT Press, 2005. [3] D. Leonard-Barton, Wellsprings of Knowledge, Harvard Business School Press, 1995. [4] S. Kujala, “User Studies: A Practical Approach to User Involvement for Gathering User Needs and Requirements,” Doctoral Thesis, Helsinki University of Technology, 2002. [5] J.T. Hackos and J.C. Redish, User and Task Analysis for Interface Design, Wiley, 1998. [6] M. Kuniavsky, Observing the User Experience: A Practitioner's Guide to User Research, Morgan Kaufmann, 2003. [7] J. Hom, “The Usability Methods Toolbox,” 1998, http://jthom.best.vwh.net/usability/index.htm [8] U.S. Dept. of Health & Human Services, “Usability.gov: Your guide for developing usable & useful Web sites,” 2009, http://usability.gov [9] UsabilityNet, “UsabilityNet: usability resources for practitioners and managers,” 2003, http://usabilitynet.org [10] D. Norman, “Why doing user observations first is wrong,” Interactions, vol. 13, 2006, pp. 50-ff. [11] H. Beyer and K. Holtzblatt, Contextual Design: A Customer-Centered Approach to Systems Designs, Morgan Kaufmann, 1997. [12] P. Ehn, Work-Oriented Design of Computer Artifacts, Stockholm, Sweden: Arbetslivscentrum, 1988. [13] J. Greenbaum and M. Kyng, Design at Work: Cooperative Design of Computer Systems, Hillsdale, NJ, USA: Lawrence Erlbaum Associates, 1991. [14] D. Schuler and A. Namioka, Participatory Design: Principles and Practices, Mahwah, NJ, USA: Lawrence Erlbaum Associates, 1993. [15] R. Williams, J. Stewart, and R. Slack, Social learning in technological innovation: Experimenting with information and communication technologies, Cheltenham, UK: Edward Elgar, 2005. [16] J. Stewart and R. Williams, “The wrong trousers? Beyond the design fallacy: Social learning and the user,” Handbook of critical information systems research: Theory and application, D. Howcroft and E. Trauth, eds., Cheltenham, UK: Edward Elgar, 2005, pp. 195-221. [17] A. Crabtree, T. Rodden, P. Tolmie, and G. Button, “Ethnography considered harmful,” Proceedings of the 27th international conference on Human factors in computing systems, Boston, MA, USA: ACM, 2009, pp. 879-888. [18] S. Greenberg and B. Buxton, “Usability evaluation considered harmful (some of the time),” Proceeding of the twenty-sixth annual SIGCHI conference on

[19]

[20] [21]

[22] [23] [24] [25] [26] [27]

[28]

[29] [30]

[31] [32] [33]

Human factors in computing systems, Florence, Italy: ACM, 2008, pp. 111-120. D. Svanæs and J. Gulliksen, “Understanding the context of design: towards tactical user centered design,” Proceedings of the 5th Nordic conference on Humancomputer interaction: building bridges, Lund, Sweden: ACM, 2008, pp. 353-362. OECD, Participative Web and User-Created Content: Web 2.0, Wikis and Social Networking, 2007. A. Lehmuskallio, S. Tamminen, and M. Johnson, “Managing Privacy on Social Network Sites,” Medienamateure: Wie verändern Laien unsere visuelle Kultur?, Siegen, Germany: 2008. A.J. Kim, Community Building on the Web: Secret Strategies for Successful Online Communities, Berkeley, CA, USA: Peachpit Press, 2000. Sulake, “Quick Habbo facts,” July 2009, http://www.sulake.com/habbo M. Turpeinen and K. Kuikkaniemi, Mobile content communities final report, Helsinki, Finland: Helsinki Institute for Information Technology, 2007. M. Johnson, “Unscrambling the “Average User” of Habbo Hotel,” Human Technology, vol. 3, May 2007, pp. 127-153. M. Johnson and T. Sihvonen, “On the Dark Side: Gothic Play and Performance in a Virtual World,” Journal of Virtual Worlds Research, vol. 1, 2009. A. Salovaara, M. Johnson, K. Toiskallio, S. Tiitta, and M. Turpeinen, “Playmakers in multiplayer game communities: their importance and motivations for participation,” Proceedings of ACE 2005, Valencia, Spain: ACM, 2005, pp. 334-337. M. Johnson and K. Toiskallio, “Fansites as sources for user research: Case Habbo Hotel.,” 28th Information Systems Research Seminar in Scandinavia, Kristiansand, Norge: 2005. M. Johnson and K. Toiskallio, “Who are the Users of Habbo Hotel?,” Mobile Content Communities, HIIT Publications 2007-1, 2007, pp. 89-107. M. Johnson, “How The Teenage Invasion Changed Habbo or Age-Sensitive Design In Inclusive Society,” The Good, The Bad, The Challenging: The user and the future of information and communication technologies: a transdisciplinary conference. Organized by COST Action 298, Participation in the Broadband Society, B. Sapio, L. Haddon, E. Mante-Meijer, L. Fortunati, T. Turk, and E. Loos, eds, Copenhagen, Denmark: 2009. J. Nielsen, Usability Engineering, Academic Press Inc, 1993. Idean Research, “Series 60 Developer Platform 2.0: Usability Guidelines For Enterprise Applications.” A. Järvinen, S. Heliö, and F. Mäyrä, Communication and Community in Digital Entertainment Services, Prestudy Research Report, Tampere, Finland: Hypermedia Laboratory, University of Tampere, 2002.

10

ARTICLE

VII

The Virtuality of Virtual Worlds, or What We Can Learn from Playacting Horse Girls and Marginalized Developers

Mikael Johnson, Sampsa Hyysalo, and Sakari Tamminen (2010). "The Virtuality of Virtual Worlds, or What We Can Learn from Playacting Horse Girls and Marginalized Developers". Symbolic Interaction 33 (4), 603–633. Reprinted with permission.

299

The Virtuality of Virtual Worlds, or What We Can Learn from Playacting Horse Girls and Marginalized Developers Mikael Johnson National Consumer Research Centre of Finland

Sampsa Hyysalo Helsinki Collegium for Advanced Studies

Sakari Tamminen Software Business and Engineering Institute

This article discusses the emergence of Habbo Hotel as a large-scale virtual world and its changing versions through time. We consider how social events taking place within the hotel are conditioned by designed-in symbolic resources and how, in turn, with creative processes of symbolization, novel social objects emerge out of particular interactions. We then discuss how the membership trajectories of hotel members are essential in understanding how these interactions are shaped and evolve. We also look at how the digital infrastructure and its evolution lends itself to being material for actualizing particular social worlds within the hotel and how the trajectory of the infrastructure reflects the relationship between the user and the developer communities of Habbo. Finally, we discuss why, to understand more fully what virtual worlds consist of at large, analysts should look into the mutually constitutive interactions among users, developers, and surrounding business models over longer periods of time in the constantly varying actualization of any given virtual world. Keywords: virtual worlds, digital symbolization, materiality, membership trajectories, social worlds, emergence

To understand computer-mediated communication, it should be of obvious importance to decipher how computed environments come to mediate communication and hence what constitutes their “virtual” character. Most research on virtual worlds Direct all correspondence to Mikael Johnson, National Consumer Research Centre, P.O. Box 5, FI-00531 Helsinki, Finland; e-mail: [email protected]. Symbolic Interaction, Vol. 33, Issue 4, pp. 603–633, ISSN 0195-6086, electronic ISSN 1533-8665. © 2010 by the Society for the Study of Symbolic Interaction. All rights reserved. Please direct all requests for permission to photocopy or reproduce article content through the University of California Press’s Rights and Permissions website, at http://www.ucpressjournals.com/reprintinfo.asp. DOI: 10.1525/si.2010.33.4.603.

SI3304_06.indd 603

10/21/10 5:52:30 PM

604

Symbolic Interaction Volume 33, Number 4, 2010

has sought to provide answers to these and related questions by focusing on characteristic features of virtual worlds such as their architecture, digitality, communication patterns, functionalities, narrative structures, gaming qualities, and parallel places in the real world. Yet empirically speaking, any one of these characteristics alone appears to offer an important condition rather than a defining factor for the often complexly mediated action and communication that takes place in virtual worlds. This case is roughly analogous to the state of mainstream social and behavioral science concerning the “real” world some decades ago when complex social phenomena were explained by sophisticated but rather reductionist explanatory models. The pioneers of symbolic interactionism then lamented the mainstream’s neglect of social interaction and interpretation of action as the primary sources of meaning and emergence of social objects (e.g., Mead 1964; Blumer 1986). The present-day research on virtual worlds lacks consideration of the complex interplay among different actants, actors, and processes that constitute a virtual world. There are few studies—none to our knowledge—that have sought the answers to what virtual worlds are by tracing how the lines of action of different people and nonhuman actors are fitted together. This condition presents another classical concern of symbolic interactionism, namely, do we really know that the “usual suspects” for the constitution of virtuality are sole and sufficient source of their existence and to the way they mediate human action and interaction (Clarke and Star 2007)? Some actors that at first glance do not appear consequential to the construction of virtuality may on closer inquiry prove so. Take, for example, the horse aficionados in Habbo, a virtual world with 135 million registered users and visited monthly by 15 million unique users worldwide.1 These horse fans, mostly young girls ages nine to fourteen, have turned the relatively uniform humanlike avatars available in Habbo into virtual horses and riders, virtual hotel rooms into stables, and virtual rugs into grass, and they do all sorts of things that one can imaginatively enact with such arrangements. In doing so, they draw from cultural and social resources not limited to, but just as importantly augmented by, the design of Habbo, by Sulake Corporation Oy. In this, they are among a large number of individuals who log in to their rooms every day and start acting out a hotel with friends, online acquaintances, and strangers met there. The hotel of course is designed by Sulake, but it was hardly purposefully designed for playing virtual stables. In fact, what Sulake has been designing has changed dramatically over the years, as the graphical chat room the founders built for their friends as a hangout began to take a (teenie) life of its own and began to include people like teenage virtual capitalists, within-virtual-world movements, concerned parents, and youth workers, as well as corporate interests and increasingly difficult demands for means of navigation and networking. The activities of playacting horse girls and many other groups have, as we detail below, added new facets and even layers of virtuality to the hotel.2 Gaining a better grasp of these dynamics of how interactions in virtual environments contribute to their evolution is also timely on account of the transformation that their functions and design are presently undergoing. New digital environments

SI3304_06.indd 604

10/21/10 5:52:30 PM

The Virtuality of Virtual Worlds

605

such as Habbo have their historical roots in the development of technologies capable of performing action at a distance and representing different actors simultaneously in one place. The first text-based games for more than a handful of players, MUD and its descendants, were developed in the late 1970s. Computer-mediated communication was transformed in the 1980s and 1990s as chat worlds became graphical, starting with Habitat. By the turn of the century, virtual worlds such as Ultima Online, Lineage, and EverQuest had collected about a million users all told, and now visits to Second Life, World of Warcraft, Gaia Online, Habbo, and so forth have turned into mainstream activities (Bartle 2003; Castronova 2005; Kzero 2009). These digital spaces have increasingly begun to blend with other social media, and, consequently, visitors to Habbo Hotel themselves debate whether it is a game, a chat, or something else. In their press releases, the developers of Habbo have called it a virtual world, a social networking service, an online game environment, a virtual gaming community, a teen community, and a teen Web site. Such sorting out of the categories and intertwinements that constitute virtual existence like Habbo—whether it be virtual fish, flesh, or fowl, so to speak—runs a risk of falling back to technologically determinist registers (or at least results in explanations of social action stemming from the inherent designed-in characteristics).3 In contrast, in this article, we approach the constitution of Habbo, as an exemplar of the constitution of a virtual world, through a rich set of empirical materials gathered from its developer company as well as through an extended ethnographic and historical follow-up on the action and actors in the virtual hotel. Our research agenda can be condensed into the following question we seek to address in this article: What is the role of interaction in the constitution of virtuality of virtual worlds? Before we move on to the empirical analysis, however, we need to make recourse to virtual worlds research to show what is already established about these sites as well as to substantiate our claim that virtual worlds research has to date skipped too hastily over the interactions taking place in these sites or at least failed to draw the full implications of their importance. This discussion is continued by introducing our own theoretical points of departure, methodology, and research design.

Virtual Worlds—Stories, Game Structures, Places, or Process? To understand better how our analysis contributes to the analyses of digital spaces called virtual worlds, we take a look at what we already know about virtual worlds. Games studies’ two major schools of thought, narratology and ludology, have framed virtual worlds as storytelling and game structures. Narratologists argue that games should be understood as new forms of stories and that they can be studied through theories of narrative (Murray 1997). According to ludologists, however, narratologists analyze game experience with too much focus on the form of the new media, representations, and the play of meaning, but too little focus on human practice, game mechanics, and rules (Aarseth 1997; Juul 2001; Malaby 2007).

SI3304_06.indd 605

10/21/10 5:52:31 PM

606


Games scholars tend to associate virtual worlds with games. Games, in turn, are considered as play, or a form of human activity with three intrinsic features. Play is viewed as separate activity from everyday life (from “work,” for instance), safe (consequence-free or not productive), and pleasurable or fun (Caillois 2001). However, recent research has been struggling with the tight coupling of games and play. There is increasing empirical evidence that playing games is not intrinsically consequence-free or separable from everyday experience. Virtual economies, exchanging virtual goods for dollars, and virtual crime provide the prime examples (Lastowka and Hunter 2004; Castronova 2007; Dibbell 2006; Lehdonvirta 2009). Indeed, standing on solid anthropological research, Malaby (2007) argues convincingly that none of the features of play mentioned (separable, safe, pleasurable) holds as an intrinsic, universal feature of games when these virtual worlds are examined empirically. Virtual worlds have also been analyzed with the aid of typologies (Messinger et al. 2008; Elverdam and Aarseth 2007) on the basis of the features of their material infrastructure, much as in early Linnaean research in biology. It is possible to distinguish virtual worlds from social networking sites (Boyd and Ellison 2007) like Facebook, or online games with designer-provided objectives such as World of Warcraft and The Sims Online. However, from economic, legal, or social sciences standpoints, these distinctions are not as clear-cut. As Darwinism forced a change in biology, a move from typologies based on ideal types to ones based on origin, the virtual world typologies get into trouble as the “Web creatures” seem to crossbreed. For instance, Habbo added social networking tools in 2007, several Facebook applications provide role-playing possibilities, and recently Habbo became a Facebook application. This necessitates an increased sensitivity to process and contingency, how activities in virtual worlds are open-ended and not predetermined by game rules (Malaby 2007). Virtual worlds have also been analyzed with the aid of analogies such as the concept of “third places” (Oldenburg 1999). Third places—pubs, cafés, coffeehouses, barbershops, beauty salons, and so forth—are used to keep in touch with reality outside the home and the workplace. They can be described as accessible, playful, focusing on conversation, like a home away from home. Many computer-mediated contexts share several of the qualities of third places, though the limits of this analogy are still open for debate (Soukup 2006; Steinkuhler and Williams 2006). However, independently of the chosen analogy, there is a risk of the virtual world becoming naturalized as “not real” and its virtual worldliness escaping analysis. While analogies of “real” places offline help users understand and act online, thus rendering the digital spaces as socially meaningful places, they leave the virtual entity metaphysically hanging in the air and are not suitable for producing a nuanced understanding for the newness virtual spaces bring into the world. Must a virtual world be one thing? Must we be able to settle the questions about its character already in advance? A number of science and technology studies, exemplified here by Taylor’s (2006) research on EverQuest, suggest otherwise. Taylor describes the artifact of EverQuest as constantly changing and contextually rendered by different actors. The first EverQuest designers, the Live Team, and the legal

SI3304_06.indd 606

10/21/10 5:52:31 PM


607

and marketing departments all understand the object in their own ways. The players bring with them their own backgrounds as they encounter EverQuest and have made it something else than what came off the shelf in 1999. This interest in investigating how various actors mingle in constructing and constituting a virtual world is also our aim in this article. We are also drawn in the same direction as Taylor (2006:162) as she notes, “The simple punch line, you see, is that EverQuest is not just one thing nor easily contained in the object that came off the shelf in 1999. In some ways we might think of it as a boundary object” (Bowker and Star 1999). While we do fancy finding virtual worlds as (one or more instances of) boundary objects as an interesting opening for novel theoretical avenues in virtual world research, we intend to pursue this line of reasoning more fully.

Studying the Interactions of Multiple People(s) in Habbo: Methodology, Methods, and Data In drawing from the social-worlds framework in symbolic interactionism, we believe this framework is apt for more than Taylor’s above “punch line” coined through one of the concepts developed within it (Star and Griesemer 1989). The framework offers analytically robust tools to link such issues as material design, socialization, and play into organizational characteristics and onward to how different peoples and reference groups interact in Habbo. We first attend to the special character of symbolically mediated interactions in virtual worlds and its implications for the makeup of the symbolic repertoires available. We then expand our analytic focus stepwise from the copresent interactions to examine how actors and actants have ended up in the kinds of encounters and interactions we first analyze, and in so doing wish to draw attention to the several intersecting social worlds (Strauss 1978; Clarke and Star 2007) at interplay in the virtual world and how its composition as a symbolictechnological assemblage can be located as an emergent property (actualization of virtuality) of this interplay. To traverse analytically from the copresent interactions to social worlds, we turn to the symbolic interactionist concept of trajectory (e.g., Strauss 1993:53–68; Bowker and Star 1999). Following Strauss (1993:53), we understand that trajectory “refers to a course of action but also embraces the interaction of multiple actors and contingencies that may be unanticipated and not entirely manageable.” With interactions of multiple actors, trajectories tend to be shaped by multiple, intertwined lines of action and multiple social worlds as frames of reference and as conditions for action. Contingencies point to the unanticipated character of trajectory, in that trajectory is used to denote sequences of actions that involve twists and turns rather than ones amenable to predictive control (p. 53). In our use, the concept is further specified by following Bowker and Star (1999:165–94) in emphasizing (1) more long-term (weeks, months, years) depiction of actions and interactions, (2) involvement in trajectories of nonhuman elements, and (3) the need to examine the phenomena of interest through a complex interplay of intertwined trajectories that have varying durations and paces.

SI3304_06.indd 607

10/21/10 5:52:32 PM

608


The rationale behind these more specified uses is the following. Interactionist science, technology, and medicine studies have strongly emphasized giving materiality due place in interactionist analysis (e.g., Clarke and Star 2003; Clarke 2005), and this is equally important for studies of virtual worlds, where early literature celebrated virtual identity games as ephemerally free of bodily and material constraint (Turkle 1995; Robinson 2007). Bowker and Star’s work exemplifies how the concept of trajectory can be usefully expanded to analyze how matter comes to matter. Furthermore, so interpreted, it becomes roughly parallel with the “biographies of technologies and practices” approach in science and technology studies, a quest that goes beyond interactionism to stress the need to study both development and use as well as to do so with multiple timescales and grain sizes of data in order to yield more balanced accounts of the sociotechnical phenomena under study (Pollock and Williams 2008; Russell and Williams 2002; Hyysalo 2010). Indeed, even in the strongly anti-determinist and methodologically maverick field of technology studies (and pervasively in other fields examining social aspects of technologies), the vast majority of studies have, somewhat surprisingly, focused on studying either development or use (e.g., Latour 1996; McLaughlin et al. 1999; Woolgar 1991) or, if covering both, addressing one or the other side only at some distance in terms of the granularity of analysis (Berg 1997; Pinch and Oudshoorn 2003). In practice, the methodology we adopt here means several years of ethnographic involvement with both developer and user communities and their evolution. In the present study on Habbo Hotel, several bodies of data were gathered over the years, presented below chronologically so as to do justice to the gradually deepening access to both developer and user communities that could be negotiated during the study. 1. The project started in 2003 with participant observation in Habbo user communities, pilot interviews, community manager interviews (n = 4), and an explorative survey (June 2004) on the visitor profiles (n = 10,000), as Sulake’s first global youth survey was published two years later. The first author also participated in Habbo, explored the features and their affordances, and analyzed default values and users’ degrees of freedom. The visits to Habbo occurred, on average, about twice per month throughout the first three-year span of research, and every second month from 2007 onward (for more on user groups, see Johnson 2007 and Johnson and Sihvonen 2009). 2. Fan sites became an important source of knowledge about the user communities. In 2004, 173 Finnish Habbo fan sites were identified, and 23 of these that were created for a large Habbo audience were analyzed in detail (Johnson and Toiskallio 2005). The size of a fan site varied between five and fifty Web articles, and the most active sites had forums with thousands of posts. Whereas the survey findings provided background statistics, the fan sites and forum discussions provided insight into active user groups and popular activities and on Hotel history. Since its very beginning, Habbo-goers have maintained several sites devoted to what has changed in the hotel and how, including kinds and looks of furniture, characters, in-game games, and design flaws. These members’ own documentation has been carefully followed and stored throughout the six years of the research.

SI3304_06.indd 608

10/21/10 5:52:32 PM


609

3. Throughout the research project, the first author collaborated with other Finnish Habbo researchers doing thesis work on a Habbo topic. One Sulake employee wrote her humanistic master’s thesis on communication and action in Habbo (Pietiläinen 2004). In addition, two youth workers wrote bachelor’s theses based on their youth work in Habbo (Sihvola 2005; Koskinen 2006), leading to an evaluation study by Merikivi (2007). The collaboration has provided important secondary sources of Habbo data. 4. In 2005 the first author conducted thematic two- to three-hour interviews with ten Habbo developers, or about two-thirds of the Habbo game development organization at the time. Six of the developers interviewed (graphical designers and both client and server developers) had been with the organization since the beginning, five years earlier, while four developers had about one year of Habbo experience. In conjunction with these interviews, historical materials were collected (screenshots, access to previous versions, press releases, advertisements, etc.) about the development of Habbo and its predecessors, to help construct its development in the years 1999–2003. 5. Having examined the user groups, adjoined sites, and developer practices, the study turned to in-depth interviews with particular users, taking the form of two- to three-hour individual, pair, and group interviews with eleven- to sixteen-year-old users (n = 6), and users aged thirty and older (n = 6) to focus on their participation histories, their motivations, the meanings they give to Habbo, groups they participate in, and so forth. 6. Habbo has also been a topic for students in various usability research courses at Helsinki University of Technology, with which the first and second authors are affiliated. Two student efforts are noteworthy: a usability test with new ten- to twelveyear-old users (n = 8) in 2004 and a software engineering effort in 2004–2005 to develop a fan site starter kit for active Habbo users. The first author mentored the second group and conducted a pair interview in 2007 about the playability testing and usability practices at Sulake. 7. After these activities, the first author had the opportunity to take part in an intervention study with Sulake. For release 9 of Habbo, in 2006, a set of user feedback methods was explored with different stakeholders inside Sulake. This included confidential data sets: database statistics and surveys from two countries on the use of a new feature. In addition to these research activities, Sulake representatives participated in project partner seminars held every six months, two workshops on virtual economy, and many project meetings arranged by the research project. These meetings made informal discussions and a continuous dialogue with Sulake possible. Taken together, these bodies of data provide us with an excellent view of the varying forms of interchange and dialogue between the varying users and developers of this virtual world. The data analysis has proceeded in multiple waves over the years, including but not limited to quantitative and qualitative sorting of Habbo user profiles (data sets 1, 2, and 5), coding and examining of interviews in regard to participation and development histories in Habbo (data sets 4 and 5), analysis of

SI3304_06.indd 609

10/21/10 5:52:33 PM

610


fan site contents and developer perceptions of the fan sites (Johnson and Toiskallio 2005; Johnson 2007, data sets 2 and 4), construction of time lines of Habbo’s various kinds of development (data sets 2, 5, and 6), and construction of the changes in Habbo’s user–developer relations over the years from multiple sources of data (Johnson 2010). Let us now dive into the material, first examining how interaction takes place in this virtual world and then moving on to (re)construct several intertwined trajectories (Bowker and Star 1999; Strauss 1993; Timmermans 1999) on multiple scales of analysis, proceeding from individual members’ participation trajectories and their reference groups to trajectories of material facets of Habbo and to longer time frames of the evolution of Habbo and its developer community.

Social Interaction in a Virtual World Virtual Social Interaction Habbo is a virtual world where children and teenagers meet, socialize, and play many types of games. In the overall design of Habbo, clear winning conditions and game-play rules have been avoided. The hotel consists of public and private rooms, where the virtual hotel visitors, called Habbos, chat, buy virtual furniture, decorate rooms, play mini-games, and arrange social events. Most of the teenage players log on after school and spend, on average, around forty to forty-five minutes per day there. Instead of an entrance fee or monthly fee, the profit model is based on micropayments in the hotel. Virtual furniture, mini-games, and membership in the Habbo Club are bought with Habbo credits. These credits can be purchased with real-world money. At the time of writing, there are Habbo hotels in more than thirty countries, and 15 million players visit Habbo each month (Sulake 2010). Virtual worlds, such as Habbo Hotel, are complicated settings for analyzing symbolic interaction and its constitutive character in the formation of coordinated action for (at least) two reasons. First, as in any given social community, interaction depends on the use of shared symbolic repertoires (Mead 1964; Blumer 1986), but in virtual worlds this has some peculiar features. Because symbols can consist of any given gesture, object, or relation between the two, they must be recognized as such by the participants in the interaction. This recognition happens only if the actants use symbols consciously, which, in turn, requires that they be able to assume the role of the interactional counterpart so as to anticipate the reaction to the action. It is only the fitting together of these interpretations of action (now symbolically mediated) that gives also the subsequent, resulting event a symbolic meaning.4 However, social acts performed in virtual worlds like Habbo are not exactly like face-to-face interaction situations, where vocal language is the primacy medium and repository of meaningful symbols. Instead, virtual worlds have predefined, heavily limited, and a priori designed-in communication modes that are not primarily vocal and give out predefined, graphically represented symbolic resources on top of “ordinary” written

SI3304_06.indd 610

10/21/10 5:52:33 PM


611

symbolic language (understood as coherent discourse—be it English, Arabic, or teenage slang) that are immediately visible to all users. All stages of interaction are mediated by computers, Web interfaces, and the technologies of infrastructure that make this all possible. This is the second significant aspect of symbolic communication in virtual worlds: it requires a constant interpretation of the graphically represented objects it consists of, and that these really be “objects” in their interactionist sense— everything that can be indicated and referred to by the users themselves (Mead 1964; Blumer 1986), which, as we show below, can take a myriad of forms in the multifarious interactions that take place within, for example, Habbo. Users must learn these and not other modes or symbols to establish meaningful social interaction, and, indeed, for many people entering Habbo the first time, the first reaction is a frequent loss of sense and frame of reference—what is going on here? Regular Habbo-goers also frequently face similar difficulties in understanding what is going on in areas where actors and activities are unfamiliar to them, and are at an utter loss to decipher the meaning of conversations carried out there. This points to the third important aspect of in-virtual-world interaction, that the designed-in symbolic resources are complemented and resymbolized by the interactants and that this can, and often does, have dramatic consequences for the very “makeup” of the symbolic setting. Let us approach these aspects now one step at a time to elaborate how they play out in Habbo and what their implications are for understanding virtual worlds. To begin with, the interaction within Habbo Hotel draws both on predefined modes of communication and on a limited set of given symbolic resources and mechanisms from which the users can draw in their interactions. Table 1 summarizes the interactional mechanisms available in Habbo.

TABLE 1. Summary of the Interactional Resources Available in Habbo Interactional Resources “Verbal” Paraverbal (gestures) Affordances of space for solitude Personal space Fidelity Mediation control Identity Fine-grained audience control Navigation Space customization Persistent text Moderation

SI3304_06.indd 611

Habbo Hotel Examples Visualizations of speech: saying, shouting, whispering Dancing, waving, standing, sitting, lying down, virtual Post-it notes, Respect Hideaways, Hotel View, “invisible spots” Guest rooms Pixel graphics Voluntary logon Avatars, their appearances, nicknames Friends List and Private Messages, Shared Groups, Friend Categories Hotel Navigator, teleporting, room categories, location of friends, event list, homepage friends lists, profile tags Virtual Interior Design User Profile Web pages, Discussion forums, Tags Kick, ban, shut-up, chat filter, furni trading automat, furni ownership history

10/21/10 5:52:34 PM

612


Figure 1: Speech bubbles in Habbo (Sulake 2004). The avatar’s name is displayed in boldface before the message. Such limited means of interaction in Habbo have direct consequences for how interaction can and does unfold. The symbolic interavatar communication takes place predominantly through speech bubbles (Fig. 1)—the way we have been accustomed to in the case of cartoon characters’ communication—and comes in three different modes to choose from. These modes are “say,” “shout,” and “whisper.” Saying something in the room makes other avatars near you hear what you are saying. If you want to draw the attention of all the Habbos in the room, you can shout. Or if you want to be secretive and share your thoughts only with a particular Habbo near you, you can whisper. The possibility of changing the modality of textual communication imitates the “real-life” vocal modalities, but is represented visually through the bubbles. The visual representation is important because the temporal aspects of conversation are reworked with this modality of symbolic expression compared with vocality. While vocal conversation takes place in linear sequential order, the “visual speech” gives the interaction and, for instance, the expected turn taking in it (Sacks, Schegloff, and Jefferson 1974) a sense of spatiality and conversational simultaneity— conversational encounters happen predominantly in space rather than in time and also act as particular symbols and resources for marking space and space-claiming activities. Another interavatar communication resource is a private communication device called the Habbo Console.5 It allows for sending direct personal messages through the space-time of the Habbo world to your avatar friends. It also acts as material evidence of one’s friendship networks through the list of avatar names one has. These different mechanisms and the modalities of communication they afford, as should

SI3304_06.indd 612

10/21/10 5:53:02 PM


613

Figure 2: Instant messaging through the Habbo Console, and personal voice control.6 be analytically evident, are used for different purposes—in Habbo, as elsewhere, the choice of the medium in itself constitutes an important part of the message. Besides communication in different modes, the creation of avatars and of rooms are important ongoing exercises in symbolic identity constitution. An avatar’s bodily form, color, gender, and clothing style create a subtle set of distinctions, since “Habbos” (as the avatars are called) can be composed only from a set of predefined characteristics. Members of established groups and frequent Habbo visitors can easily read and interpret a large variety of signs about these material/symbolic bodily articulations of the “Habbos”: how old are they? are they newbies? is that one a member of a group already or not? are they “cool” or not? who are the “scammers” or moderators? and so on. What materializes in the body of avatars (in clothing, skin color, gender, and interaction styles) is the whole symbolic universe of Habbo, the way distinctions between good taste and bad taste in a certain world within the hotel are made and constantly remade through social negotiations. Similarly, various material things that are used to build the room environments provide a method of nontextual communication by allowing a certain variety of distinctive elements to be bought (“furni”) and used to decorate one’s room. The

Figure 3: Changes in the color and form of the most visible elements of a Habbo avatar: hair, shirt, and trousers.

SI3304_06.indd 613

10/21/10 5:53:03 PM

614


rooms can be used to convey different tastes in decoration by buying and trading furniture judged to belong to a certain desired style. These are central elements in the world making within Habbo, and this is how a priori designed symbolization is actualized within. However, as we stressed above, new and emergent symbolization is frequently achieved. This happens through creative communication practices and skillful interplay between material and textual, between symbolic and relational elements within the rooms. To get a better empirical grounding in how exactly these social worlds emerge out of novel symbolization processes, let us examine more closely the “equine” activities inside the hotel. This happens in the play between avatar/ users, objects, and their relations to spaces—a process where the equine activity becomes actual in Habbo through the emergence of significant and shared symbols that tie different actors together and mediate action. Playful social relationships here are framed by configuring both human (avatar) and material (digital) elements in imaginative ways that still bear some resemblance to horse activities outside the digital world. Horse-riding activities (like any other, such as formula racing, found within Habbo) require a quasi-bounded digital space, or a setting that allows the play to be focused on commonly shared digital objects. The spatial infrastructure is arranged in particular ways to enable interpretative flexibility about already named objects provided by Sulake as basic components of the rooms—tables, chairs, decorations, and such. A new “placeness” can be symbolized through their spatial rearrangement, the result of which is that any object residing within this social space becomes resymbolized. In horse rooms (e.g., “stables” and “riding grounds”), for example, the spatial rearrangement indicates that tables and chairs have become new social objects (e.g., stalls or hurdles) that are constitutive ingredients of virtual horse activities.

Figure 4: The riding grounds for the horses. On the left, the manège, with Habbo tables as “hurdles.” On the right, the room where the horses are “put out to pasture,” with three green Habbo rugs as “virtual grass” and the blue rug as a “pond of water.”

SI3304_06.indd 614

10/21/10 5:53:13 PM


615

The riding activity itself is socially coordinated by assigning roles to the avatars within the rooms by their location or by active textual communication. Within rooms indicated as settings for equine activities, an avatar can become a “horse owner” or a “horse.” Horse ownership is claimed for an undefined period of time by asking the horse for its approval. The role of “owner” is self-claimed. One can just start to perform linguistically most of the care work one would do on an organic horse outside Habbo—grooming and patting, shoeing, and harnessing the horse all happen by letting the others know that this is happening to the horse at the moment. Becoming a horse, on the other hand, is indicated to others by standing within the bounds of the marked stable—a spatial bodily “gesture” that has been turned into a sign of particular role-taking initiative. The role of being a horse is further symbolically indicated by selecting a brown skin color for the avatar and by typing the breed, temperament, and age of the role-played horse in the avatar’s description (an identity indicator accessible to others through clicking on the avatar). The role of “being a horse” is often reinforced by making “horse sounds” and communicating the feelings (via undecipherable uttering or descriptions of the sounds and feelings of the horse marked with an asterisk at the beginning and end of the description to separate it from normal conversation), but this kind of communication is not necessarily involved in the case of all “horses”—central is that the interavatar communication is a noncommunicative act in its symbolic sense. The horse sounds and silence both act as noncommunicative markers in the intentionally framed human–horse conversation of gestures, or the staged nonsymbolic interaction that paradoxically occurs through symbols.7 A horse has various ways of accepting a new owner, most often indicated by walking close to the avatar that asked for the permission and responding to the commands the avatar gives to the horse. Here the textual and the nontextual communication intertwine seamlessly to perform a virtual trans-species agreement on the central relationship needed for “riding.” Horse owners can then command the horses for riding sessions on dedicated riding grounds—these are usually other rooms connected to the main stables by teleports allowing for a spatial bridge between them. The riding grounds are enacted by decorating the connected rooms as different kinds of terrain. As the avatars cannot really ride on each other, the riding activity itself in Habbo happens by walking the avatars side by side. The self-given descriptions of action are also differentiated from normal directed talk by asterisks. These communicational frames, as Goffman (1974) would call them, give indication and sense for the actors in the riding world in three senses: first, they indicate the position (the “self”) of the actor (e.g., “owner”); second, they indicate the desired position of the other self (“horse”); and, finally, they indicate the correct interpretation of the joint action to be formed by this coordinated action. The frequent switching of the role positions also reinforces this equine activity—new entrants learn what is to be expected, how to use the different symbols and bodily gestures turned into agreements to communicate correctly, and finally how to make sense of the various emergent possibilities of action and interaction that the virtualworld-turned-into-new-activity offers.

SI3304_06.indd 615

10/21/10 5:53:14 PM

616


Figure 5: A horse stable together with horses, horse owners, and horse girls. Notice the different ways of framing direct and indirect communication with asterisks by horses and horse girls.

The Actualization of Social Activity in Habbo: Groups and Social Worlds as Virtual Resources for Action The horse stable leads us to the issues of outside symbolizations brought into Habbo. By following the fan sites (Johnson and Toiskallio 2005), we can identify gang members (people who dress similarly and hang out only with each other), Habbo supervisors, cheaters, quiz makers and players, and celebrities, as well as more loosely framed groups such as goths, punks, emos, skaters, and hip-hoppers who differentiate themselves through clothing and particular communication styles. This is a crude generalization, since we continuously learn about more and less peculiar ways to perform being in Habbo. For instance, one group of friends uses Habbo a few times a year to gather there and “catch up.” Also not unheard of is parents leaving messages to and communicating with their children through Habbo. Strange events may take place, such as avatars clad in all-gray clothes repeating “follow the path,” gaining a following, and then disappearing—only later to turn out to be art students staging a performance. What should be clear by now is that not only are there some generic

SI3304_06.indd 616

10/21/10 5:53:25 PM


617

orientations within the hotel, but, just as importantly, there are generic orientations on the “outside” that are brought in and actualized there. Let us return to our stable owners. As imaginative as their solution was, they created only a small proportion of the rather impressive number of rooms related to horse riding and grooming activities. The equine world has a tangible presence in Habbo—one can find hundreds, or thousands (no one, including Sulake, really knows how many), of “horse settings.” This is all the more interesting for having taken place without any special design support to assist the actualization of equestrian activities, be it graphical design or providing the users with special modes of communication between horse- and humanlike avatars. Importantly, however, the actualization of the virtual equine activities draws from not only the quasi-bounded digital space but entire outside “equine worlds” (cf. Becker 1982) consisting of interrelated activities and subworlds in horse breeding, riding, stable keeping, veterinary medicine, international horse shows, Olympic games, and wide publicity in newspapers and TV—mingled with smells of autumn forests and so forth. Just as importantly, the virtual horse subworld in Habbo draws from other virtual pet worlds (significant in popularity, such as www.neopets.com) and “real-life” pet worlds and hence comes into being as a mixed set of imaginative, communicative, and material resources.8 All together, these constitute a host of practices, activities, and cultural representations from which the horse girls in Habbo draw—both reflexively and unwittingly—in their creative actualization of the equine activities within Habbo (Becker 1982; Strauss 1993). As already stated, the equine world is just one of the multitude of social worlds brought in with people visiting Habbo. At the more formal end of the scale we encounter youth workers moving about as they try to find and stay in touch with teens. On the more fleeting end we find Habbo being used as a chat arena for grannies exchanging recipes and new mothers exchanging experiences. There are two important implications we hope to have established through the above discussion: (1) new kinds of virtualities emerge in virtual worlds through joint coordination of action that shapes its digitally represented material properties, conditions their reinterpretation, and thus undermines the suggested status of communicative symbols “designed” to be available in the virtual world; and (2) the actualization of these virtualities—emergent properties and designations of new objects and their interpretations through joint action—owes much to symbolic resources brought into the virtual world from the outside and from parallel and previous experience.

Trajectories of Participation and Materiality We find it important, however, to deepen both of these observations by moving from temporally confined depiction of a virtual world’s interaction to increasingly longer time frames of its existence. This helps us understand in more depth how, exactly, these emergent properties emerge in the history of particular people and materialities and in the Habbo setting as a whole. In doing so, we move away from a setting that is

SI3304_06.indd 617

10/21/10 5:53:25 PM

618


typical computer-mediated communication analysis to “contextualizing” it through descriptions that are more commonly found in constructivist studies of technology (Clarke and Star 2007). This also exemplifies the potential of symbolic interactionism for bridging these two common research perspectives.

Membership Trajectories—Changes in Orientation and Multiple Memberships We shall now detail how the emergent symbolizations in Habbo, not only for horse girls but also for the material properties of Habbo, such as key interaction resources for navigation, become established historically. In other words, we examine how the interplay between designed-in, reinterpreted, emergent, and redesigned symbolization takes place through time. With a risk of appearing utterly descriptive, we first revisit some of the actors’ own sociology on how people come to do the kind of imaginative things they do in Habbo. According to the development company’s market research (Sulake 2006, 2008), there is no single reason users play Habbo. Sulake has segmented Habbo users into five equal-sized categories: achievers, traditionals, creatives, rebels, and loners. This segmentation is based on their values, attitudes, subculture memberships, and interests. The differing user-base segments highlight a diversity of motivations among users (hinted at by their descriptive labels), but these are not linked with actions online nor do they explain how user motivations change over time. Besides demographics, a good deal is also known about what kinds of activities take place in Habbo (Johnson and Toiskallio 2007). Players engage in various collective activities: beauty contests, furniture trading, maze solving, dating, role-play in hospitals, travel agencies, gangs, schools, TV shows, stables, and so forth. With regard to how such engagements develop, players themselves present and publicly represent and debate categories of engagement involved, on some of the fan sites tightly linked with the Habbo community. On the fan site Nerokala, a “Habbo journalist” called Andna explained that Habbo characters could be classified into novices, amateurs, pros, and superhabbos according to their experience. In a Habbo magazine (Paratiiskala, no. 8), fan site host Toivo differentiated among four kinds of Habbo characters—furniture traders, competitive players, VIPs, and chatters— and suggested that people move from furniture and competitions to chatting. Our survey of ten thousand Finnish community visitors by and large confirms the above member typifications of people’s activities. But to dig deeper into these matters, we depart from Strauss’s (1978:124) assertion that “rather than looking simply at organizational mechanisms such as recruitment, we should also be looking at how people get contacted by, encounter, rub up against, are introduced to, drawn into, and hooked on social worlds.” Accordingly, we next focus on how people find their way to Habbo in the first place, what they do there, and how their presence there changes. To illustrate the change in presence through time, let us outline two trajectories that are more instructive of the different key issues of life in Habbo (in a

SI3304_06.indd 618

10/21/10 5:53:25 PM


619

sense, what Flick [1998:68–73] would call “critical cases”). In presentation of these trajectories, we condense their path into five stages to aid in comparison. The first trajectory is one of an eleven-year-old boy, whose path from “novice” to “pro” illustrates the economic, status, and furniture-related aspects of Habbo. He himself characterizes some of his doings in Habbo as follows: J: Then I kept a kind of a hotel. It made damn good profit. M: Oh? J: They all bought rooms and got, . . . M: So you had a kind of [your own] hotel there inside the [Habbo] hotel? J: Yes. M: Okay, cool. J: And they rented a room for a day and . . . M: So you divided your own room into parts or? J: No, I made those really small rooms and then I put a bed there and stuff. M: Yes. J: Then I ren–, I provided them with teleports and then a big lobby room and made a lot of [rooms] and then everybody rented them all around. M: Oh? J: Then I had workers there, and they had to be given salary every now and then. M: Yes, yes, what did the workers do then? J: Well, they sort of showed them where the rooms are and stuff . . . M: What other kinds of those systems have you had? J: At one point, I kept a café . . . M: What salary do you then give to the workers there? J: Well, I give some of the profit, all the lousy furni. M: Okay, the lousy furni? J: Yes, ’cause some want to work when they don’t have anything.

A closer look at his enterprises reveals the following participation trajectory. First encounters: The boy learned about Habbo and visited it on occasion. However, his father forbade him spending money on Habbo, so he could not buy any furniture. Early position: He discovered that one can make a profit by running a bazaar in a hotel guest room. In a bazaar, the room guests line up to trade with the room owner. The bazaar keeper/room owner can make a profit, because he can decline bad furniture-trading offers and accept good ones. Establishing presence: Together with a friend, he decided to keep a café, since they knew that one can share ownership rights in the room, and one person doesn’t have to be there all the time to keep the café open. Continued involvement: The greatest profit came from a hotel within the hotel. Then he created many small rooms and connected them with teleport hub rooms. After this, he “employed” other Habbos to show the way to the rooms. The suite was very popular, and the formerly broke kid became a Habbo millionaire. Waning participation: Two years later, the father informed us that his son does not go to Habbo that regularly anymore but is still running an in-game formula course for a friend of his there.

SI3304_06.indd 619

10/21/10 5:53:26 PM

620


This membership trajectory is very economically oriented: all the actions the boy performed are directed toward capitalization on the possibilities provided by the digital infrastructure. In a sense, this is exactly what the Sulake corporation would want people to do in Habbo. However, the father’s decision not to allow his son to spend any “real” money on Habbo led the boy to capitalize on his own imagination— the spaces he created for other Habbo-goers were based on imitation of some culturally relevant spaces outside Habbo, “offline.” It is also noteworthy that the end of his three-or-so-year involvement is a slow let-go that continues to entertain some of his friendships and skills. We elaborate in greater depth on the significant shifts in orientation and status with the second trajectory, that of two of the “horse girls” we encountered above and who played Habbo mostly together. First encounters: The girls encountered Habbo with friends and siblings; one of them inherited furniture from her brother, who was active in Habbo two years earlier. Early position: Both tried various Habbo activities at first, participating in various user-created game and furniture-trading events. Establishing presence: The girls began to run a “horse stable” with others, where they held horse-riding competitions. In contrast to most stables in Habbo, their rooms were listed not in the pet category but in the “fun” and “other” categories. Continued involvement: The horse stable activity ended at some point, both because its upkeep became too laborious and because it was no longer considered socially “cool” as stables became more common in Habbo. At the time of the interview, the main activity of the girls was to meet their friends, for example, in a room with multiple “kissing corners” where one engages in dating. Their motivation to participate in a particular room is also to “keep it on the [top] list” of Habbo rooms—a self-reflection toward the Habbo “fashion.” Waning participation: The girls were still active in Habbo but also are beginning to have a presence in IRC-Galleria, a Finnish site targeted to older teens. The trajectory of these two friends begins and ends in a manner that is relatively common for Habbo presences. The participation in the hotel is relatively transient (Kazmer 2007:114) and includes shifts in orientation, or, as the girls put it, “At first stuff was important, but it no longer matters.” Just as the generic orientations and positions within Habbo are not mutually exclusive, multiple and partial memberships flourish in terms of outside commitments. As with the other trajectories, these people come from somewhere and go somewhere after their time in Habbo is over. Both girls had at some point ridden also in real life, they enjoyed other pets, and they were aware of the “wolf pack”—another animal theme group in Habbo— fighting other packs. The somewheres people come from affect how people engage with other people and activities in the hotel. Newcomers may have a particular reason for exploring the hotel, or they may orbit Habbo and other attractive games and environments for a while, trying some of them a few times before getting more involved.

SI3304_06.indd 620

10/21/10 5:53:26 PM


621

So, the initial orientations tend to shift, once one is in contact with Habbo and its other visitors, even as one or several activities come to define the communications and interactions people have. These participation histories lay conditions for symbolic interactions with the designed-in symbols as well as to any emergent symbolization that may occur. This is particularly evident as the participation histories are also anchored to offline world participations of the members, and because membership in one community acts as a resource in the other on those occasions when participation paths converge. In short, the patterns and contingencies in participation trajectories strongly shape what kind of emergent symbolization may occur and hence what kind of emergent virtualities may actualize. To underscore this latter point further, we now examine the trajectory consisting of a constant dialogue between designappropriation-resymbolization-redesign (and so on) in the actualization of Habbo’s navigational means.

Material Trajectories, or How Matter Comes to Matter The materials, just as people do, have varying trajectories—in fact, so much so that Habbo aficionados have devoted entire pages to tracking how different avatar properties, furniture, decorative elements, and room spaces have evolved over the years.9 The material trajectories can, on the one hand, be seen as responses to player preferences and wishes (we further substantiate this claim below). At the same time, they always depend on Sulake Corporation Oy and the worlds of social media design, advertisement, and technology management that are drawn to Habbo via the company. Let us follow in detail one trajectory that Habbo members themselves tend to forget—that of navigation mechanisms. There are several strategies for addressing the problem of finding similar-minded persons or friends within the vast Habbo. These strategies have developed over time as the number of users and rooms in Habbo has increased. At first, people only walked from room to room, but as rooms multiplied this became impractical. Over the first few years, the developers developed three important devices: the Navigator, the Console (friend list), and furniture with teleporting capabilities. The room-listing device is called Navigator. In the early days, it was a searchable list of all the rooms by name in a particular hotel, both meeting points and attractions created by Sulake and rooms created by Habbo users. The list enabled all room owners to brand and to advertise their room publicly by writing public descriptions for a room. But, as user-created rooms in Habbo were found in excess, the list also began to be as long as the Great Wall of China. This is why these user-created rooms became organized into categories in October 2004, around ten to twenty subcategories, depending on hotel size. The Habbo Console, in contrast, featured instant messaging among friends in the hotel. Acquaintances in Habbo could agree to become “Habbo friends,” which implied that, with the aid of the Console, they could mutually see each other’s whereabouts in the hotel and send instant messages to each other.

SI3304_06.indd 621

10/21/10 5:53:27 PM

622


At this point, Habbo Hotel had changed from multiple village-like places into one resembling a town, action moving from public places increasingly to private areas, creating a need to better hook people up to where all this action was. For instance, for the horse girls then, besides checking their friends on their Console, one strategy to find other Habbo-goers interested in horses was to use the room descriptions as indicators of a room turned into a stable. But room names took, for instance, our horse girls and their quest for finding a similarly interested avatar only so far. The problem is that if the horse girls were not previously acquainted and in some other room in Habbo, they could not easily recognize other horse girls among the ever-growing hordes of avatars. An ingenious self-identification technique emerged to address this problem. The horse girls started to add a small textual identification before their avatar’s one-line description, which indicates, for example, the name of the stable room. This small indicator then worked as a sign of membership of a particular horse group and made the identification processes both for the girls themselves and for outsiders easier. Hence users’ innovative actions again added to the developer actions. The Navigator device, coupled with the possibility to name user-created rooms and to link avatars to that name, provided our horse girls with a powerful tool to find potential equestrian rooms and members belonging to these. But, as users started creating multiple rooms with the same theme, the Sulake developers were inspired to create direct links between two rooms. So the teleport was born. Teleports, of which some aesthetically resemble British phone booths (a result of being developed for the British Habbo in 2001), are devices that will transport an avatar from one room to another in a predetermined way. Teleports always come

Figure 6: The Navigator device, subcategories of user-created rooms (left) and rooms from the category Hair Salons and Modeling Rooms.

SI3304_06.indd 622

10/21/10 5:53:36 PM


623

Figure 7: Three forms of the teleport device: a phone booth, a toilet cubicle, and a closet. in pairs—once a teleport set is acquired, either by purchasing or by receiving it as a gift from someone else—the user has two ends of a teleporting route. Teleports usually connect places that have something in common—they might connect a set of rooms owned by one person or a vast room network owned by a group of friends. The biggest networks might contain more than one hundred connected rooms. It is the teleport device then that made it possible to create various sub-virtual-worlds within Habbo Hotel itself by making it easily malleable in spatial terms and, as such, supports the enactment of subworlds, such as the stables our horse aficionados are so fond of. As the numbers of users and user-created rooms increased, Sulake developers introduced a number of new features in 2006–2007 to manage the complexity of finding like-minded people in Habbo. The biggest change was probably the addition of user homepages, which made the Habbo inhabitants browsable and searchable. In addition, the Habbo users could display their own rooms and list of friends publicly, which made social navigation in Habbo even easier. Inspired by the group formation among users in Habbo, the developers added features to support group communication: a group identifier (badge), group homepage, and discussion forum for group members. A few months after the user homepages were launched, Sulake introduced a feature called tags, which again changed social navigation in Habbo. Habbo users could now tag themselves with their hobbies, interests, personality, and status in Habbo. It is easy to see the resonance with users’ self-identifiers here. The tags are searchable, and by clicking on a tag the user gets a listing of all users and groups with that tag (a reverse index). Then, by clicking on another Habbo user, the first user can view the homepage of that user. And on the homepage, the rooms of that particular user normally are listed, and the first user can enter that room in the hotel directly, without using the Navigator. Now, at the time of writing, Sulake is changing the Navigator again. The interaction between users and developers, as argued above, happens both through a creative user appropriation of the digital symbols and mechanisms the

SI3304_06.indd 623

10/21/10 5:53:36 PM

624


developers are offering at any given time and through the introduction of novel interaction mechanisms and symbols to Habbo in the hope of “bettering” the virtual world. In fact, a constant “mattering” of the infrastructure comes about only through this dialogue, showing again nicely how virtuality is actualized through the digitality of the virtual world itself. With these developments and the growth in size of the hotel, Sulake, together with its users, has facilitated the emergence of increasingly differentiated subworlds and events.

Overall Development of a Virtual World The Development Trajectory of Habbo The material trajectories take us squarely to the more long-term trajectory of Habbo’s overall material and social development. In the early Habbo days, the hotel was developed by a handful of game developers with core competencies in graphic design, Macromedia Shockwave Flash clients, and Java server programming. At first they developed the hotel called Kultakala (“goldfish” in Finnish) for themselves and their friends, but just a year after launch it became popular among teenagers. Through the internationalization of the hotel, the organization grew, and every country got a local office with a few employees working on moderation, community management, customer relations, and marketing. More administration and business personnel have joined, and game development now involves more than a dozen game developers. At the time of writing, Sulake employs about 270 people, in the Helsinki headquates and eleven other offices around the world.” Should read: At the time of writing, Sulake employs about 270 people in the Helsinki headquarters and eleven other offices around the world. Habbo was based on two previous launches, Mobiles Disco (October 1999) and a snowball game called Lumisota (February 2000). Mobiles Disco provided the basic hotel infrastructure in the pixelated “retro” style: rooms, easy navigation from room to room, the chat interface with speech bubbles, and avatar customization and movement inside the rooms. For Lumisota, the developers implemented paying through cell-phone messages, which at the time was a practical solution for the Finnish market. This worked well for the Finnish Hotel Kultakala (August 2000) but wasn’t that successful in the UK version (February 2001). At the time, British teens didn’t use cellular phones as widely as Finnish teens, which meant that Sulake had to develop new ways to buy furniture. Phone call systems, youth cash cards made of paper, and check systems came first, but soon payment by credit card, use of PayPal, and bank transfers followed. With the UK hotel also followed the teen invasion of Habbo. It was not designed for teens at first, as illustrated by the developers making furniture sets with bar desks as integral elements and one of the new public rooms being greatly influenced by stereotypical English pubs. The large number of teen visitors also meant a large number of concerned parents. To keep all this together, community management

SI3304_06.indd 624

10/21/10 5:53:36 PM


625

received a lot of attention. A set of guidelines for good behavior was established, called Habbo Way, and tools for governing it were developed. The moderators got their own interface to Habbo, which meant that they didn’t have to be in the same room to follow a discussion, and a text filter was developed to screen for swearwords. The customer service got an automated response system to reduce the amount of personal service needed. In 2002–2003 Sulake developed its business, the hotel, and the community simultaneously. Strategic partnerships were made, and brands such as Mountain Dew and Britney Spears entered Habbo, the former as a vending machine serving Mountain Dew bottles and the latter as a large poster to be hung on walls. Later these brands disappeared as Sulake tried another approach by developing Coke Studios for Coca-Cola and Walt Disney Magic Kingdom for Warner Cinema. These were customized versions of the technical game engine behind Habbo, which helped fund the development of the hotel. The hotel’s technical architecture was stabilized, and security improvements were made to package the hotel into a product that could be more easily rolled out in new countries. The community manager in each country started writing weekly newsletters, sending “Welcome!” messages to new visitors, celebrating birthdays, arranging competitions, arranging activities and events, and so forth. While, in the beginning, Habbo didn’t follow any temporal rhythms, seasonal themes now were brought in to give Habbo a particular rhythm. Competitions and new furniture emerged to celebrate Halloween, Christmas, Easter, the summer holidays, and other themes or topics important to teenagers. Right from the start, the Habbo visitors started writing and debating Habbo on their own homepages, and some created totally Habbo-themed Web sites. While these fan sites complemented Sulake’s community management, they also helped Sulake in regulating sites that were out to scam the users. In 2006 Sulake launched its own forum and later group support features, which again changed the power balance in the Habbo “mediaworld.” Habbo’s profit model allows a group of friends to meet in Habbo for free, while those interested can pay for the hotel services they want. Sulake offers a standard set of furniture items in a catalog that is updated now and then with new furniture items. Some furniture items are given monthly to members of the Habbo Club, and collector’s furniture items at 20 credits become available for two weeks only, which increases their future value. The furni items do not wear out, which means that the same piece of furniture can be passed along to new hotel visitors as older ones grow tired of furniture trading. Most of the Habbo visitors do not pay, but those who do spend about €10 a month, on average. To secure revenue streams, in 2003 Sulake introduced the Habbo Club, which—in exchange for a few euros per month—provides its members with priority access to rooms and club-only hotel public spaces; exclusive hairstyles, clothes, dance styles, and guest room layouts; and some special features. About 85–90 percent of Sulake’s income related to Habbo comes from such small payments by many users (cf. the “long tail”), and the rest from advertisements in Habbo.

SI3304_06.indd 625

10/21/10 5:53:37 PM

626


Marginalization of Developers, or The Distributed Agency in the Development of Habbo The development trajectory of Habbo Hotel has been significantly shaped by its visitors. As noted, Sulake met the teen invasion to Habbo Hotels by crafting a social order appealing to teens and parents. By 2006, thirteen-year-old girls considered themselves old in Finnish Habbo, and fifteen-year-olds hid their Habbo visits at school in order not to be ridiculed by their peers. Long before this, the original user set—young adults consisting of Habbo designers’ professional and leisure peer groups—found themselves in more and more awkward conversations with other visitors. Typical avatar names showing one’s year of birth, such as “Cosmo85,” made them either recognized celebrities or old freaks no one wanted to converse with—both equally undesired positions in terms of enjoying their visits. The invasion of kids pushed older teens and adults to more grown-up and less frequently visited areas and hours of the hotel, such as after pub closing time (in the real world). Various legitimate places also remained within Habbo, such as a table in “Café Heaven,” reserved only for certain VIP regulars. Even as the face of the hotel changed drastically and rapidly, there was remarkably little open conflict. Older visitors simply emigrated from the niches invaded by kids to more secluded parts of the hotel and to other digital environments. While the rejuvenation of Habbo was actively facilitated by the design and business decisions of Sulake Corporation Oy, it led the company into a set of quandaries. Many teen hangouts attract younger kids, who follow in the wake of “cool” older kids, who in turn leave the place that has been turned suddenly “uncool” by what they regard as weenie brats; Habbo runs a risk also of losing its target group of thirteento eighteen-year-olds. As youth workers know, this common pattern has long been known to desolate places designed to keep, for example, teens off the streets and is notoriously hard to reverse. An interesting side effect is that the developers of Sulake, once insiders, are no longer confident that they understand what is cool (let alone why) even when it is pointed out to them by kids. They have thus become outsiders—first big brothers and then uncles—to the communities of most visitors. Originally unheard of, techniques such as user research had to be introduced by Sulake to follow the development of Habbo. Such measures also stem from the increasing complexity and specialization of the development organization, where not everyone can afford to invest the time to follow events in all Habbo hotels around the world. The final part of the quandary is that the rejuvenation of Habbo differentiates hotels in different countries. In Japan, for instance, kids do not have PC access at home in the same fashion as in Finland, and Habbo visitors there remained in the twenty to forty age group. In some later releases such as for Spain and Portugal, Habbo was originally branded for the thirteen to eighteen age group and was immediately appropriated by the younger teens. Moderating and developing a “chain of Habbo Hotels” face new challenges, with the visitor bases having increasingly different tastes

SI3304_06.indd 626

10/21/10 5:53:37 PM


627

and interests, arguably adding to difficulties in taking features developed in one hotel to others. While virtual worlds may allow greater possibilities for moderation and channeling action by the developer company than more tangible products, Habbo makes the limits of this control clear. Who the visitors are and what they do affects the behavior of other visitors—even when most of them never actually meet each other. The people and patterns of interaction shape the evolution of the place. Sulake lost omnipotence over all rooms and action some years ago because of the technical complexity and the number of rooms in Habbo. However, it decided not to worry about the loss of control but instead to stay on alert as to what interesting and potentially fruitful developments might occur. This is not least because the virtual world experiences versions of realworld phenomena such as dynamics between relatively young and older kids and the establishment of communities having ties and building on representations to outside social worlds—all of which are hard, if not impossible, for the company to control. The clear lesson to draw from the overall development trajectory of Habbo is that virtual worlds research, and social studies of technology and media more generally, must start addressing both development and usages of social media within the same study. Just following action and interaction in a given moment in a virtual world like Habbo would give an extremely truncated view of the sociability and interaction there. Moreover, not extending the study to both development and usage activities would, as it appears to us, lead to almost inevitably mistaken guesswork over what, for instance, developers have intended with specific features and, conversely, how the design of a feature affects play in the virtual world (just think of the virtual stables that would be virtually impossible within the limits of Habbo’s design!). The digital materialities of Habbo are highly relevant to how Sulake can manage, channel, stir up, and block actions of the vastly different people and interest groups that move around in Habbo. However, to view Habbo’s materiality as “Sulake’s creation” could emerge only from research (or opinion) ignorant of the developers’ close interaction with, and responses to, player wishes and interests (Johnson 2007, 2010). Over a tenyear span, in total, twenty-three purposeful and formidable approaches to mediating developers and users can be observed (Johnson 2010)! The virtuality of a virtual world cannot be reduced to any one of its elements or group of actors affecting it. Instead, many facets of it are virtual also in the sense of being emergent phenomena arising from the interplay and conjoining actions of a multitude of actors.

DISCUSSION Taken together, these findings foreground several issues for virtual worlds research. First, as Robinson (2007) stresses, we need to pay attention to how virtual worlds are related to real-world activities and social worlds. People come to virtual worlds from various somewheres that greatly affect what they engage with in a virtual world and how their computer-mediated communication takes place. We need to focus on both the users and the corporation behind the virtual world, as well as look at the

SI3304_06.indd 627

10/21/10 5:53:38 PM

628


shifting activities, orientations, and motivations that are negotiated in their various interactions. The interactionist concept of trajectory is further conducive to analyses at multiple “grain sizes” of data and, accordingly, to keeping in sight how trajectories of members, material characteristics, and groups intertwine and how they constitute the longer time frame trajectory of the virtual world and its developer company. Second, while these observations may seem trivial, they are significant for debates about what virtual worlds are: no one element such as specific functions, narrative structures, particular architecture, placeness, or digital environment turns a site into a virtual existence. Following Taylor 2006, we argue that the multiplicity of resources rooted in various background social worlds of participants is likely to feature as important in any given virtual world. It may be less visible in such virtual worlds as do not so explicitly encourage people to externalize their affiliations, interpretations, and desires concerning the real world. More tightly scripted role-playing environments may also draw only in a more limited way from the participants’ experience. While such differences are likely, they do not necessarily follow from, or primarily from, a virtual/real division, or game/hanging out, or particular narrative structures, or gaming qualities. Instead, the differences come in the form of how different actors negotiate their positions, which can affect others and are affected by others. To us, these likely differences are a further reason to foreground the analysis of the interaction between multiple peoples in the study of virtual worlds over the explanations held dominant to date. A third noteworthy facet of Habbo is the dramatic changes the virtual site has gone through in the few years of its existence. The teen invasion and consequent changes in the design and commercial aspects of Habbo have added entirely new kinds of virtualities to it, as our example of horse girls bears witness to (along with goths and many other subgroups; see Johnson and Sihvonen 2009). Drawing from the equine world and children’s playmaking practices, these teen girls have managed to alter (one could say subvert) the materialities of Habbo and effectively added a layer of imaginative existence to Habbo that never was designed into it. The virtuality of virtual worlds comes into being in multiple ways, some more digital, some relying more on imaginative add-ons to designed components. Fourth, Habbo’s meandering development trajectory—consisting of intertwined trajectories of a large number of people and different material objects, such as the changing means of navigation and communication—calls into question any temporally rigid conceptualization of what it is. This observation leads us to depart from Taylor. Virtual worlds present interesting conceptual terrain for interactionist concepts of objects and materialities we should attend with care. The earliest Habbo, where developers and their friends interacted and most newcomers belonged to roughly the same reference group, would have been aptly characterized as their site for within-group interaction. This would, however, no longer aptly describe Habbo after the teen invasion. Indeed, Taylor’s (2006) insistence concerning EverQuest being a boundary object would have suited Habbo well at the point when it was limited to a few thousand, or tens of thousands of, players who gave it robust and generic

SI3304_06.indd 628

10/21/10 5:53:38 PM


629

meanings in their communications with others in and out of the hotel, and more specific meanings among their key reference groups, whether designing, playing the economy, or just chatting with friends (cf. Star 1989; Bowker and Star 1999). But Habbo’s more complex shape is no longer easily amenable to this conceptualization either. The reasons are similar to ones that led Bowker and Star (1999) to coin the term “boundary infrastructure” to denote complex infrastructures that act much in the manner of boundary objects between diverse social worlds but have such a complex and large structure that calling them “objects” would seriously misrepresent their materialities as well as the social interactions involved.10 Yet Habbo’s material makeup (even as it is spread into about thirty countryspecific and somewhat varying versions) appears too uniform to be well conceptualized in terms of a “boundary infrastructure,” as it lacks many key characteristics of an infrastructure (cf. Star and Ruhleder 1996). It should also by now be clear that we second Taylor’s insistence that virtual worlds are too diverse and their existence too rooted in multiple outside social worlds for equating a large virtual world with “a social world” (in the sense of it being one social world per se), even as other immersive digital environments have been discussed as social worlds, as in the work of Kazmer (2007). It is an interesting question whether all the various massively multiplayer digital environments together should be seen (or could be seen in the future) as constituting such a cluster of activities and subworlds that they would aptly be conceptualized as a distinct social world of “virtual worlds.” It would then be of further interest to examine the role of Habbo’s emergent virtuality and developer responses to it in the shaping of this broader setting. However, as the question is mostly empirical and we have not conducted systematic research on other virtual worlds, we wish to refrain from any further speculation. The point here, quite obviously, is not in trying to present some offhand fitting of different stages of a growing virtual world to interactionist concepts but to illustrate that, just like “real-world” social formations, digital ones have careers wherein their shape—even their nature, as in what kinds of virtualities can be actualized in the virtual world—can vary importantly because of the interactions taking place there.11

FINAL REMARKS The study presented in this article draws heavily on the idea that, to understand virtual worlds, one needs to study “both ends” of the user–producer relationship, here translated as Habbo visitors and the Sulake corporation, and see their various coproductive entanglements through time, as an inherently mutable process. Our claim is substantiated by the illustrative examples above of how the constitution of the “worldliness” of virtual worlds is coproduced by their constant interaction in different modalities and mediations. In many ways, the study of virtual worlds can lean on an idea presented by many in interactionist research, perhaps most aptly phrased by Becker (1982:35): “The world exists in the cooperative activity . . . not as a structure or organization . . . keeping in mind that such things change.” What we

SI3304_06.indd 629

10/21/10 5:53:39 PM

630


see as the most important lesson learned in the case of Habbo is that virtual worlds are not reducible to any one description or metaphor, because they are made out of emergent interactions within and between many intersecting social worlds, some of which are highly transient and others more stable, materializing in and through shared, cocreated digital infrastructures. The future challenges of internet research lie therefore not only in the study of particular typologies made possible by the stable worlds made out of generative relations provided by the human and nonhuman actors (individual and collective) but also in the temporal processes that make possible the different articulations between these elements of constant permutations of social life. This implies future development of sensitizing concepts complementary to the ones derived from “offline” interaction as well as a research strategy capable of addressing the changing trajectories of social settings that we today call virtual worlds.

Acknowledgments: We wish to thank Sulake Corporation Oy and the Mobile Content Communities research project for enabling this study. Hyysalo’s work has been funded by Academy of Finland, grant 1122706.

notes 1. By one way of counting (that of www.kzero.co.uk), Habbo is the world’s most populous virtual world and definitely very popular as a (web of) teenage site(s). 2. Virtuality, here, should be understood as the not-yet-actualized events that technoscientific assemblages such as virtual worlds make real, or actualize, in digital environments (see Deleuze 1994; cf. Mol 2002). We give empirical examples of this actualization of virtuality in-action below through the tools provided by symbolic interactionism—that is, by asking how symbolization here becomes a social fact. 3. This would amount to claiming that there is no virtuality, in the sense of emergent events, in social communication per se (that all communication would be totally predictable beforehand) or, even worse, that “virtuality” of a (communicational) event would be constituted solely by the technologies at hand. 4. Of interest here is that the meaning of action does not require the actions to be mediated through symbols, as Mead (1964) pointed out—the turns taken in and through the conditioned “conversation of gestures” do have “objective” meaning, as does the outcome of this kind of interaction. Interaction only becomes symbolically significant if the action and the turns taken in action become anticipated, controlled, and temporally arranged with regard to a prior idea of a meaning of an event, human, thing, or generalizable relationship between them (Mead 1964, esp. 75–82 and 117–25). Blumer (1986:8) later translates these as nonsymbolic and symbolic forms of action. Virtual worlds problematize this heuristic differentiation, as we show below. However, we do not address this issue further here except by indicating in the footnotes when questions about the empirical soundness of this dichotomy become relevant. 5. The functionality of the Habbo Console was split into Friend management, Minimail, and Instant Messaging in 2008 with subsequent user interface changes. 6. These screenshots are not an exact representation of Habbo’s contemporary appearance, because they and the data are tightly coupled with a specific time and Habbo version. 7. And it is in instances like these where the analytic distinction between nonsymbolic and symbolic interaction breaks down—empirically taken, staged playacting and layered symbolic interaction can include elements of conversation that are symbolically intended and indicated to be nonsymbolic. For a good analysis of this communicational paradox, see Bateson and Bateson 1987:80.

SI3304_06.indd 630

10/21/10 5:53:39 PM


631

8. Sulake also provides ready pets in Habbo, bots that have rigid ownerships and very limited interactive capacities, and hence horse interactions are in striking contrast to “having a pet” in Habbo. 9. For old features and feature evolution, see www.habbox.com => Guides + Content => Habbo Knowledge => Old Habbo Features or, for country-specific variations, http://www.habborator. org/furniture/list.html, and for release details http://www.habborator.org/facts/index.html. 10. Surely, Habbo can be considered a social object as in something that can be referred to, but this would be an utterly superficial way to sensitize ourselves to it. Let us remember that, after all, for instance, in Finland, as much as 25 percent of the early-teen cohort has, or has had, a presence there. 11. Some of the other interesting questions we did not have space to explore within this article concern the specifics and intersections of teen and digital cultures. One example is the transience of Habbo for its participants, most designers included, as one’s trajectory in Habbo lasts two to three years and typically moves to other social media environments after this.

REFERENCES Aarseth, Espen J. 1997. Cybertext: Perspectives on Ergodic Literature. Baltimore, MD: Johns Hopkins University Press. Bartle, Richard. 2003. “Designing Virtual Worlds.” New Riders Games, July 25. Bateson, Gregory and Mary Catherine Bateson. 1987. Angels Fear: Towards an Epistemology of the Sacred. New York: Macmillan. Becker, Howard S. 1982. Art Worlds. Berkeley: University of California Press. Berg, Marc. 1997. Rationalizing Medical Work. Cambridge, MA: MIT Press. Blumer, Herbert. 1986. Symbolic Interactionism: Perspective and Method. Berkeley: University of California Press. Bowker, Geoffrey C. and Susan L. Star. 1999. Sorting Things Out: Classification and Its Consequences. Cambridge, MA: MIT Press. Boyd, Danah M. and Nicole B. Ellison. 2007. “Social Network Sites: Definition, History, and Scholarship.” Journal of Computer-Mediated Communication 13(1). Retrieved from http://jcmc. indiana.edu/vol13/issue1/boyd.ellison.html. Caillois, Roger. 2001. Man, Play, and Games. Urbana: University of Illinois Press. Castronova, Edward. 2005. Synthetic Worlds: The Business and Culture of Online Games. Chicago: University Of Chicago Press. ———. 2007. Exodus to the Virtual World: How Online Fun Is Changing Reality. New York: Palgrave Macmillan. Clarke, Adele E. 2005. Situational Analysis—Grounded Theory after the Postmodern Turn. Thousand Oaks, CA: Sage. Clarke, Adele E. and Susan L. Star. 2003. “Symbolic Interactionist Science, Technology, Information, and Biomedicine Studies.” Pp. 539–74 in Handbook of Symbolic Interaction, edited by L. T. Reynolds and N. J. Herman. Walnut Creek, CA: Alta Mira. ———. 2007. “The Social Worlds Framework as a Theory/Methods Package.” Pp. 113–37. in The Handbook of Science and Technology Studies, 3rd ed., edited by E. J. Hackett, O. Amsterdamska, M. Lynch, and J. Wajcman. Cambridge, MA: MIT Press. Deleuze, Gilles. 1994. Difference and Repetition. New York: Columbia University Press. Dibbell, Julian. 2006. Play Money, or How I Quit My Day Job and Made Millions Trading Virtual Loot. New York: Basic Books. Elverdam, Christian and Espen Aarseth. 2007. “Game Classification and Game Design: Construction through Critical Analysis.” Games and Culture 2:3–22. Flick, Owe. 1998. An Introduction to Qualitative Research. London: Sage. Goffman, Erving. 1974. Frame Analysis: An Essay on the Organization of Experience. Cambridge, MA: Harvard University Press. Hyysalo, Sampsa I. 2010. Health Technology Development and Use: Practice Bound Imaginaries and Biographies, or How Technological Configurations Evolve. New York: Routledge.

SI3304_06.indd 631

10/21/10 5:53:40 PM

632


Johnson, Mikael. 2007. “Unscrambling the ‘Average User’ of Habbo Hotel.” Human Technology 3(2):127–53. ———. 2010. “User Involvement, Social Media, and Service Evolution: The Case of Habbo.” Proceedings of the Forty-third Annual Hawaii International Conference on System Sciences. Retrieved August 31, 2010, from http://doi.ieeecomputersociety.org/10.1109/HICSS.2010.425. Johnson, Mikael and Tanja Sihvonen. 2009. “On the Dark Side: Gothic Play and Performance in a Virtual World.” Journal of Virtual Worlds Research 1(3). Retrieved August 31, 2010, from http://journals.tdl.org/jvwr/article/viewArticle/368. Johnson, Mikael and Kalle Toiskallio. 2005. “Fansites as Sources for User Research: Case Habbo Hotel.” Paper presented at the Twenty-eighth Information Systems Research Seminar in Scandinavia, August 6–9, Kristiansand, Norway. Retrieved from http://www.hiit.fi/u/johnson. ———. 2007. “Who Are the Users of Habbo Hotel?” Pp. 89–107 in Mobile Content Communities. HIIT Publications 2007-1, April. Retrieved from http://www.hiit.fi/publications. Juul, Jesper. 2001. “Games Telling Stories? A Brief Note on Games and Narratives.” Game Studies 1. Retrieved August 31, 2010, from http://www.gamestudies.org/0101/juul-gts/. Kazmer, Michelle M. 2007. “Beyond C U L8R: Disengaging from Online Social Worlds.” New Media and Society 9:111–38. Koskinen, Aleksi. 2006. “Elämistä, olemista, pelaamista—keitä Habbo Hotelin käyttäjät ovat ja kuinka heidän kanssaan tehdään nuorisotyötä?” (“Living, Being, Playing—Who Are the Habbo Users and How Can One Do Youth Work with Them?”). Unpublished thesis, Mikkeli Polytechnic, Finland. KZero. 2009. Virtual Worlds Registered Accounts Q3 2009. Retrieved August 31, 2010, from http:// www.kzero.co.uk/universe.php. Lastowka, F. Gregory and Dan Hunter. 2004. “The Laws of the Virtual Worlds.” California Law Review 92:3–73. Latour, Bruno. 1996. Aramis, or The Love of Technology. Cambridge, MA: Harvard University Press. Lehdonvirta, Vili. 2009. “Virtual Item Sales as a Revenue Model: Identifying Attributes That Drive Purchase Decisions.” Electronic Commerce Research 9:97–113. Malaby, Thomas M. 2007. “Beyond Play: A New Approach to Games.” Games and Culture 2:95– 113. McLaughlin, J., P. Rosen, D. Skinner, and A. Webster. 1999. Valuing Technology: Organisations, Culture, and Change. London: Routledge. Mead, George Herbert. 1964. Selected Writings. Indianapolis, IN: Bobbs-Merrill. Merikivi, Jani. 2007. “Netari.fi—nuorisotyön arvo” (“Netari.fi—the Value of Youth Work”), verkkojulkaisuja 14. Nuorisotutkimusverkosto/Nuorisotutkimusseura (Youth Finnish Youth Research Society). Retrieved August 31, 2010, from http://www.nuorisotutkimusseura.fi/. Messinger, Paul R., Elini Stroulia, and Kelly Lyons. 2008. A Typology of Virtual Worlds: Historical Overview and Future Directions. Journal of Virtual Worlds Research 1. Retrieved August 31, 2010, from http://jvwresearch.org/. Mol, Annemarie. 2002. The Body Multiple: Ontology in Medical Practice. Science and Cultural Theory. Durham, NC: Duke University Press. Murray, Janet Horowitz. 1997. Hamlet on the Holodeck: The Future of Narrative in Cyberspace. New York: Free Press. Oldenburg, Ray. 1999. The Great Good Place: Cafes, Coffee Shops, Bookstores, Bars, Hair Salons, and Other Hangouts at the Heart of a Community, 3rd ed. New York: Marlowe. Pietiläinen, Reetta. 2004. “Kommunikointi ja toiminta Habbo Hotel Kultakalassa” (“Communication and Action in Habbo Hotel Kultakala”). Unpublished master’s thesis, University of Helsinki. Pinch, Trevor J. and Nelly Oudshoorn, eds. 2003. How Users Matter: The Co-Construction of Users and Technologies. Cambridge, MA: MIT Press. Pollock, Neil and Robin A. Williams. 2008. Software and Organizations: The Biography of the Packaged Enterprise System, or How SAP Conquered the World. London: Routledge.

SI3304_06.indd 632

10/21/10 5:53:41 PM


633

Robinson, Laura. 2007. The Cyberself: The Self-ing Project Goes Online, Symbolic Interaction in the Digital Age. New Media and Society 9(1):93–110. Russell, Stewart and Robin Williams. 2002. “Social Shaping of Technology: Frameworks, Findings and Implications for Policy with Glossary of Social Shaping Concepts.” Pp. 19–35 in Shaping Technology, Guiding Policy: Concepts, Spaces, and Tools, ed. K. H. Sørensen and R. Williams. Cheltenham, UK: Elgar. Sacks, Harvey, Emanuel A. Schegloff, and Gail Jefferson. 1974. “A Simplest Systematics for the Organization of Turn-Taking for Conversation.” Language 50:696–735. Sihvola, Sebastian. 2005. “Netari.fi Nuorisotyö Habbo Hotelissa” (“Netari.fi –Youth Work in Habbo Hotelissa”). Unpublished bachelor’s thesis, HUMAK, University of Applied Sciences, Finland. Soukup, Charles. 2006. “Computer-Mediated Communication as a Virtual Third Place: Building Oldenburg’s Great Good Places on the World Wide Web.” New Media and Society 8:421–40. Star, Susan Leigh. 1989. Regions of the Mind: Brain Research and the Quest for Scientific Certainty. Stanford, CA: Stanford University Press. Star, Susan Leigh and James Griesemer. 1989. “Institutional Ecology, ‘Translations’ and Boundary Objects: Amateurs and Professionals in Berkeley’s Museum of Vertebrate Zoology, 1907– 1939.” Social Studies of Science 19:387–420. Star, Susan Leigh and Karen Ruhleder. 1996. “Steps toward an Ecology of Infrastructure: Design and Access for Large Information Spaces.” Information Systems Research 7(1):111–34. Steinkuhler, Constance and Dmitri Williams. 2006. “Where Everybody Knows Your (Screen) Name: Online Games as ‘Third Places.’” Journal of Computer-Mediated Communication 11. Retrieved August 31, 2010, from http://jcmc.indiana.edu/vol11/issue4/steinkuehler.html. Strauss, Anselm. 1978. “A Social World Perspective.” Pp. 119–28 in Studies in Social Interaction 1, edited by N. Denzin. Greenwich, CT: JAI. ———. 1993. “Continual Permutations of Action.” Communication and Social Order. New York: Aldine de Gruyter. ———. 2006. “Global Habbo Youth Survey.” Lahti, Finland: Sulake Corporation. ———. 2008. “Global Habbo Youth Survey.” Lahti, Finland: Sulake Corporation. ________. 2010. Quick Habbo facts, July. Retrieved from http://www.sulake.com/habbo. Taylor, T. L. 2006. Play between Worlds: Exploring Online Game Culture. Cambridge, MA: MIT Press. Timmermans, Stefan. 1999. Sudden Death and the Myth of CPR. Philadelphia: Temple University Press. Turkle, Sherry. 1995. Life on the Screen: Identity in the Age of the Internet. New York: Simon and Schuster. Woolgar, Steve. 1991. “Configuring the User: The Case of Usability Trials.” Pp. 57–102 in A Sociology of Monsters: Essays on Power, Technology, and Domination, edited by J. Law. London: Routledge.

SI3304_06.indd 633

10/21/10 5:53:41 PM

A al t o D D4 6 / 2 0 1 3

So c ialme dia c h ange st h egamefo r use r invo l ve me ntin se rvic ede sign.Ac t iveuse r c o mmunit ie s, fastpac e d it e rat ivede ve l o pme nt , digit alt rail s, pe e r pro duc t io n, and l o w c o stso ft w aredist ribut io n arew e l lkno w n fac e t st h atbring subst ant ialc h ange s.T h e que st io n is, h avet h euse rc e nt re d, part ic ipat o ry, and l e aduse r de sign appro ac h e s ke pt up w it ht h ec h ange s, o r dot h e y pre se rvean o ut dat e d pe rspe c t ivet h ath o l ds us bac k? I nt h is t h e sis, t h eaut h o r dist il sl e sso ns fo r de signe rs, manage rs, and re se arc h e rs fro m an inde pt hc asest udy o f a pio ne e ring digit al se rvic e ,H abboH o t e lby Sul akeCo rpo rat io n. T h eanal ysis de mo nst rat e sh o wc o l l abo rat io n and fe e dbac kl o o ps be t w e e n de ve l o pe rs and use rs c h angeo ve rt ime .T h eke y me ssage ,w h ic hc o nc e rns st rat e gicuse r invo l ve me nt , is t ofo rge tpro c e ss guidanc et h at re l ie so n a st abl eand o rde rl y so ft w are l ife c yc l eand re c o nside rt h eimpl ic at io ns fo r t iming and me t h o dc h o ic e , fo r inst anc e , in t e rms o f de ve l o pe r– use r so c ialdist anc e , c umul at io no f use r kno w l e dge , and re l e vant rh yt h ms in de ve l o pme nt .

9HSTFMG*afahca+

I S BN9 7 89 5 2 6 0 5 0 7 2 0 I S BN9 7 89 5 2 6 0 5 0 7 37( p d f ) I S S N L1 7 9 9 4 9 34 I S S N1 7 9 9 4 9 34 I S S N1 7 9 9 4 9 4 2( p d f ) A a l t oU ni v e r s i t y S c h o o lo fS c i e nc e D e p a r t me nto fC o mp ut e rS c i e nc ea ndE ng i ne e r i ng w w w . a a l t o . f i

BU S I N E S S+ E C O N O M Y A R T+ D E S I G N+ A R C H I T E C T U R E S C I E N C E+ T E C H N O L O G Y C R O S S O V E R D O C T O R A L D I S S E R T A T I O N S