The diffusion of information systems development methods

ARTICLE IN PRESS

1

46

2

47 Journal of Strategic Information Systems xx (0000) 1–18 www.elsevier.com/locate/jsis

3 4

48 49

5

50

6

51

7

52

11 12

Paul Beynon-Davies*, Michael D. Williams

9

13

53 54 55 56 57

O O

10

The diffusion of information systems development methods

8

58

European Business Management School, University of Wales Swansea, Singleton Park, Swansea SA2 8PP, UK

59

15

Received 18 September 2001; accepted 13 November 2002

60

PR

14 16 17 18 19 20

Abstract

33

Keywords: Information systems development methods; Rapid application development; Technological diffusion

25 26 27 28 29 30 31

TE

24

EC

23

R

22

R

21

D

32

This paper describes an investigation into the diffusion of information systems development methods (ISDMs). Such methods have a critical strategic impact on organizational performance in that they are the means for constructing, adapting and renewing the IS infrastructure in line with organizational strategy. We portray method diffusion as an instance of technological diffusion. Technological diffusion is normally portrayed within the IS industry as an overtly rational process. In this paper we postulate that the external diffusion process of ISDMs has many features in common with broader social movements. We maintain that the diffusion of ISDMs is a key example of knowledge-based diffusion and use the case of the Dynamic Systems Development Method (DSDM), a public domain standard method for Rapid Application Development (RAD), to refine a model of knowledge-based diffusion. Such a framework is important for understanding how such methods shape the community of practice in the IS development domain. Specifically how organizations adapt, deploy and use such knowledge. In further work we intend to verify aspects of this knowledge-based perspective in longitudinal and further case-study work. q 2003 Published by Elsevier Science B.V.

O

34

37

C

35 36

1. Introduction

40 41 42 43 44 45

In this paper we discuss the diffusion of information systems development methods (ISDMs), and portray ISDM diffusion as an instance of technological diffusion. Technological diffusion is normally portrayed within the information systems (IS) industry as an overtly rational process. However in this paper we postulate that

U

39

N

38

* Corresponding author. Tel.: þ 44-01972-205678; fax: þ 44-01792-295618. E-mail address: [email protected] (P. Beynon-Davies). 0963-8687/02/$ - see front matter q 2003 Published by Elsevier Science B.V. PII: S 0 9 6 3 - 8 6 8 7 ( 0 2 ) 0 0 0 3 3 - 1

STRINF 1184—23/12/2002—19:11—UMASHANKAR—60123 – MODEL 1

61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90

ARTICLE IN PRESS 2

93 94 95 96 97 98 99 100 101 102 103 104 105 106

the diffusion process for ISDMs has many features in common with broader social movements and particularly knowledge-based diffusion. We utilise the case of the recent history of the Dynamic Systems Development Method (DSDM)—a public domain standard method for Rapid Application Development (RAD)—to illustrate some of the key features of method diffusion processes within a UK context. A discussion of two organizations’ experiences of adopting DSDM is used to refine a published model of knowledge-based diffusion. The structure of this paper is as follows. Firstly, we define the features of an ISDM and discuss to what degree the study of the diffusion of this ‘soft’ technology reflects ‘best practice’ in the IS industry. Secondly, we consider the issue of technological diffusion and make the case for considering method diffusion as a social movement. Thirdly, within this broad interpretive framework we describe the importance of a published knowledge-based model of technological diffusion. Fourthly, we discuss the case of the ISDM– DSDM— that we have used in a preliminary manner to test the efficacy of the model. Finally, we present our conclusions vis-a`-vis the validity of the model, and also call for some muchneeded further work in this area.

107 108

2. IS development methods

111

In this section we make the case for the study of ISDMs, define the component elements of an ISDM and consider three perspectives on the purpose of an ISDM. Information technology (IT) systems are critical elements of the modern ‘electronic economy’ and provide supporting infrastructure for the upsurge of interest in electronic business and electronic commerce. Traditionally practical knowledge concerning appropriate ways of developing IT systems has been packaged in knowledge bundles known as development methods. Development methods embody social structure and changes to methods reflect economic, political, social and technological changes in the IT systems environment. Methods are also the vehicles by which practitioners in this domain introduce changes to development practices. Methods therefore have a critical strategic impact on organizational performance in that they are the means for constructing, adapting and renewing the IT infrastructure of organizations.

117 118 119 120 121 122

TE

116

EC

115

R

114

R

113

D

109 110 112

123

128 129 130 131 132 133 134 135

O

C

127

A systematic survey of the existing literature on ISDMs conducted by Wynekoop and Russo (1997) revealed that over half of the 123 research papers examined consisted of normative research in which concept development was not based on any empirical grounding or theoretical analysis, but merely upon the authors’ speculations and opinions. Of those that did constitute empirical research, almost half were undertaken to evaluate ISDMs or parts of ISDMs. Few studies were conducted in order to identify how ISDMs are selected or adapted, or how they are used. There also appears to be little interpretive research and few practice descriptions or case studies. Our research is particularly directed at selection and adoption processes and employs an interpretive research stance (Walsham, 1993).

N

125 126

2.1. Lack of empirical literature

U

124

136 137 138 139 140 141 142 143 144 145 146 147

O O

92

PR

91

P. Beynon-Davies, M.D. Williams / Journal of Strategic Information Systems xx (0000) 1–18


148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180

ARTICLE IN PRESS P. Beynon-Davies, M.D. Williams / Journal of Strategic Information Systems xx (0000) 1–18 181

3

2.2. Component elements and types

226

182 183 184 185

227

Whilst acknowledging the difficulties involved in defining an ISDM (Beynon-Davies, 1998) for the purposes of this paper we make use of the view that an ISDM comprises the following components (Beynon-Davies, 1998):

186 188 189 190 191 192 193 194

231

† † † †

A model of the information systems development process; A set of development techniques; A documentation method associated with these techniques; Some indication of how the techniques chosen along with the documentation method fit into the model of the development process; † Some philosophy which might be defined as the set of assumptions about what constitutes information, an information system, and the place of information systems within organizations.

195 197 198

207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225

D

TE

EC

206

R

205

R

204

O

203

C

202

N

201

† Structured methods: Structured methods emerged during the 1980 s and initially used a linear model of the development process. Clear phases are identified with clear inputs and outputs from each phase. Techniques in areas such as data modelling and process modelling developed within structured frameworks. Structured Systems Analysis and Design Method (SSADM) is probably the most prominent example of a structured method (Weaver et al., 1998. It was created in the early 1980 s and has been particularly influential in both public and commercial sectors within the UK; † Rapid application development (RAD) methods These methods (Stapleton 1997) use an iterative model of the development process and generally specify high-level phases based around some form of prototyping. RAD methods are generally contingent in that they do not prescribe in detail the techniques to be used. A variety of techniques can be adapted to the needs of a particular project. DSDM has become the most prominent example of a RAD method currently in existence; † Object-oriented methods. OO methods tend to use a consistent object focus for the analysis, design and implementation of IT systems. Most such approaches utilise a contingent model—sometimes a linear model; sometimes an iterative model for the development process. The most influential contemporary variant of OO methods is the Unified Modelling Language (Rumbaugh et al., 1999) and the associated Unified Software Development Process (Jacobson et al., 1999).

U

199 200

232 233 234 235 236 237 238 239 240

Generally we may identify three major types of development methods (Beynon-Davies, 2002):

PR

196

230

O O

187

228 229

The three types of method above focus on the development of IT systems. A number of methods focus more on the development of human activity systems such as those in the area of participatory design (Mumford, 1983). The object-oriented (OO) paradigm has also influenced both the structured and RAD movements. Hence, both the current versions of SSADM and DSDM incorporate OO techniques within their remit.


241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270

ARTICLE IN PRESS 4 271


2.3. Best practice

316

272 273 274 275 276 277 278 279

317

The limited evidence on the adoption of ISDMs into industrial practice that we possess appears to suggest that ISDMs are perceived by practitioners as explicitly representing ‘best practice’ in the IS industry (Stolterman 1991). As such, we might argue that ISDMs represent an embodiment of social structure, and that changes to an ISDM reflect economic, political, social and technological changes in the environment of IS work. Consequently, a historical study of the diffusion of an ISDM is a cogent way of studying many of the changes that have taken place within the IS industry over the same period.

280

287 288 289 290 291

300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315

TE

EC

299

R

297 298

R

296

This of course begs the question of why ISDMs are still being created in the IS industry, and why organizations still invest considerable resources in purchasing and attempting to utilise these methods. What is perhaps interesting in this light is the finding of Button and Sharrock (1993) that developers frequently rationalise their behaviour in deviating from best practice due to project exigencies of limited time, resources, etc. We would argue that there is a necessary synthesis between these two apparently contradictory positions of best practice and organizational fetish. An ISDM should be treated as a knowledge bundle in that they act as devices for the communication and adoption of IS development practice. The creation and dissemination of ISDMs can be seen as the conventional way in which innovations in development practice are framed within the IS practitioner community. We would also argue that hype—the promotion of unsubstantiated, positive features of technologies—is a necessary and important driver of technological diffusion in the IS industry. ISDMs are also important ways of establishing the boundaries of development practice for new recruits to the industry. Fitzgerald (1997), for instance, argues that there is evidence that ISDMs are useful as a guide for inexperienced developers. Experienced developers frequently diverge from the constructs of a particular method whilst, at the same time, acknowledging the importance of methodical development. This tends to frame ISDMs in Suchman (1987) terms as being plans or guides to action, rather than deterministic rule-sets to be followed rigorously.

O

295

C

294

2.5. Knowledge bundle

N

293

U

292

323 324

328

On the other hand, there is a reasonable body of evidence to suggest that the practitioner community rarely utilise ISDMs in the way they were originally intended. The studies of Edwards et al. (1989) and Fitzgerald (1996), among others, clearly indicate that development organizations either do not use any ISDM in practice, or adapt and use existing ISDMs in their own way. Where they are used in their original form, Wastell (1996) has argued that methods frequently become used as a fetish and a way of reducing uncertainty in the development process.

PR

286

322

326 327

2.4. Organizational fetish

D

285

321

325

283 284

320

O O

281 282

318 319


329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360


5

3. Models of technological diffusion

406

362 363 364 365 366

407

In this section we examine two alternative viewpoints on the process of technological diffusion, that we have chosen to call rational and interpretive models. We particularly focus on the relevance of one particular variant of the interpretive approach—the knowledge-based model of diffusion.

367 368

378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393

398 399 400 401 402 403 404 405

O

C

397

Iacono and Kling (1996) criticise the overtly rational account of technological diffusion for ignoring the contribution of other actors—including colleagues, trade associations in the computer industry, professional societies, regulatory agencies, and the media—in the process of the promotion of computerisation. This exhibits similarities with the account of technological diffusion found in the literature (see for instance the work of (Bijker et al., 1987)) on the social construction of technology (SCOT). SCOT accounts of technological diffusion emphasise the way in which complex networks of actors (both human and nonhuman) ‘configure’ particular technologies. For instance, Pinch and Bijker (1987) discuss how historically, the development process of the bicycle was characterised by conflicting requirements such as speed and safety, conflicting solutions to the same requirements, and

N

395 396

3.2. Interpretive models

U

394

415 416 417

O O

PR

377

D

376

TE

375

Iacono and Kling (1996) characterise the history of computing in terms of computerisation movements, and employ this idea to critique conventional notions of the genesis and utilisation of computer-based technologies. They argue that conventional answers to the question of why societies computerise fall either into the camp of arguing that computer-based technologies are adopted because they are efficient substitutes for labour or older technologies, or that the adoption of computer-based technologies is a reflection of the change from an industrial to an information society. These answers rest on the interaction between two kinds of social actors: vendors of computer hardware and software, and consumers who purchase computer systems and services to meet a perceived need. This rational account of the genesis and utilisation of computer-based technologies has much in common with models proposed in the general literature on the diffusion of innovations. Diffusion is a term originally used in chemistry and biology to describe the processes by which gases move from areas of high concentration into areas of low concentration. Rogers (1995) defines the diffusion of innovations as being the process by which an innovation is communicated through certain channels over time among the members of a social system. Diffusion studies in this vein seek to trace and explain the path an innovation’s acceptance follows through a particular social system over time. Generally, such studies tend to describe diffusion in terms of an S-curve in which a few agents adopt an innovation early on, passing on their experience and influence to the later majority. Factors such as social structures and organizational cultures may facilitate or impede the process.

EC

374

414

R

373

411 413

3.1. Rational models

R

371 372

410 412

369 370

408 409


418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450

ARTICLE IN PRESS 6

452 453 454 455 456 457 458 459 460 461 462

various problems such as those of a ‘moral’ nature (such as Victorian women wearing skirts or trousers on high-wheeled bicycles). This critique leads Iacono and Kling to offer an alternative account of the rise of computer-based technologies. They argue that the rise of computing technologies is “due to a collective action similar to that of other social movements, such as the environmental movement, the anti-tobacco movement, the movement against drinking and driving, or the women’s movement…advocates of computerization focus on the creation of a new revolutionary world order where people and organizations use state-of-the-art computing equipment and the physical limitations of time and space are overcome”. Iacono and Kling proceed to focus upon two key processes critical to computerization, the ways in which movements persist over time through computerization movement organizations (CMOs), and the ways in which CMOs recruit participants.

463

467 468 469 470 471 472 473

PR

466

D

465

474

482 483 484 485 486 487 488 489 490 491 492 493 494 495

EC

R

481

R

480

O

479

C

477 478

N

476

They also argue for the importance of members of CMOs and the social network it encourages. “The primary resources of all social movements are members, leaders and some form of social network. Shared beliefs are communicated along these networks and lines of action are advocated.” In this sense, CMOs take on the role of framing technologies (Orlikowski and Gash 1994) for their membership by amplifying current problems, interpreting events, and emphasising the benefits of the transformed social order. A key rhetorical form common to many CMOs is the idea of technological utopianism—the renewal of society through technology. Iacono and Kling (1996) tend to focus upon ‘hard’ technologies such as IT networking to illustrate the social nature of computerization movements. We argue that this perspective also has much to contribute as a framework for the analysis of the diffusion of ‘soft’ technologies such as ISDMs amongst the IS community itself. We would argue that ISDMs are knowledge bundles that are diffused between organizations through social networks in which CMOs are significant actors. 3.3. Knowledge-based models

U

475

497 498 499 500 501 502 503 504 505 506 507 508

In order for any social movement to persist over time, pioneering activists must create more enduring organizational structures than those embodied in emergent and informal groups. These organizations are entities capable of taking social action. They can raise money, mobilise resources, hold meetings and formulate positions…computerisation movements persist over time with the help of computerisation movement organizations. These are organizations, or groups of organizations…which act as advocacy groups for the computerisation movement. They generate resources, structure membership expectations, educate the public, and ensure the presence of recognised leaders who can lend their prestige and interorganizational connections to the movement.

TE

464

496

O O

451


There is a clear synergy between Iacono and Kling’s focus on social movements and Newell et al’s recent published account of a knowledge-based model of diffusion. Newell et al. (2000) focus upon the diffusion of knowledge between and within organizations in relation to technology. Diffusion of knowledge involves a process of


509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540

ARTICLE IN PRESS P. Beynon-Davies, M.D. Williams / Journal of Strategic Information Systems xx (0000) 1–18

546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565

EC

566 567 568

4. Method diffusion

569

573 574 575 576 577

R

R

572

O

571

Veryard (1987) equates an ISDM to an IS, and the implementation of an ISDM as both a technological as well as an organizational innovation. He argues that, “the implementation of a new methodology is much more difficult and more critical to the success of the methodology within an organization than the selection”. We wish to portray the adoption of an ISDM as a diffusion process, as a process in which development approaches are promoted, taken up, adapted and used amongst development organizations.

C

570

4.1. Rational models of method diffusion

581 582 583 584 585

In a rational model of this diffusion process, IS development organizations are the consumers of this ‘technology’. Producers of this ‘technology’ may be either promulgators of the method or vendors of method-based products and services such as training. Hence, in the case of SSADM a body such as the Central Computer and Telecommunications Agency (CCTA) of the UK civil service promoted the method, whilst other organizations such as the National Computer Centre (NCC) continues to sell products and services associated with this method.

U

580

N

578 579

586 587 588 589 590 591 592 593 594 595 596 597

O O

545

PR

543 544

complex ideas relating to organizations and technical processes that exist in the public domain being bundled together and packaged in particular ways by technology suppliers. These knowledge bundles are designed to create solutions that can be presented by technology suppliers as relatively simple best-practice fixes for organizational problems in a wide-variety of contexts. Such knowledge becomes de-contextualised or commodified in black-boxes that hide its complexity. Black-boxed solutions are communicated directly to potential users via communication channels linking technology suppliers and users—a process of supply push. However, users do not wait passively for new ideas to be broadcast from the supply side, but actively search out new ideas through boundary-spanning activity. This involves connections with other firms, government bodies and educational institutions and is described as being a process of user pull. Newell et al. (2000) lay particular emphasis on the importance of professional associations as a network that boundary spanners use in the search process. This is because knowledge acquired through professional associations is seen to be in some way more neutral than that supplied directly by technology suppliers. However, technology suppliers also use the professional association network to push the same packaged solutions. Knowledge arrives on the organizational agenda in a black box, by supplier push and/or user pull. However, when knowledge arrives in the organization it has to be unbundled and contextualised. This process may prove problematic for a number of reasons. For example, the organization may lack an established knowledge-base of people and skills for understanding and applying the ‘technology’. Alternatively, the organization may have people with the requisite knowledge and skills but preclude them from the implementation process because of difficulties of internal networking. It is at this point that consultants may frequently be called in to manage the knowledge implementation process.

D

542

TE

541

7


598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630

ARTICLE IN PRESS 8

633 634 635 636 637 638 639 640 641 642 643 644 645 646

The selection of ISDMs by organizations is assumed to be grounded in a rational selection process based upon organizational need. The adoption of a given ISDM is assumed to be related to notions of organizational utility. Kautz and McMaster (1994) elaborate on a range of factors that enable the diffusion of ISDMs. These factors include prior experience of ISDM introduction, management support and commitment, a clear mission, organizational culture, method usability, education and training, monitoring and evaluation, and good change management. Whilst acknowledging the importance of all of these factors, we wish to emphasise here the importance of the role of CMOs in this diffusion process, and also the way in which members of such CMOs ‘champion’ particular ISDMs within their organizations. We may distinguish between internal diffusion of methods within organizations, and external diffusion of methods between organizations. The former has been studied to a limited extent in relation to the way in which methods become adapted and used (Edwards et al., 1989; Button and Sharrock, 1993; Wynekoop and Russo, 1997). The latter is a topic area which appears to have attracted a much lower level of research activity and has a correspondingly lesser body of knowledge available.

647

653 654 655 656 657 658 659 660

D

652

CMOs as we have portrayed them occupy an interesting position between that of technology supplier and professional association. It is for this reason that we have expanded Newell et al’s model highlighting the position of CMOs as adjuncts to professional associations (see Fig. 1). CMOs take the knowledge bundles produced by technology suppliers and actively promote the use of such technologies in organizations, frequently in concert with other agencies such as professional associations. Boundary spanners from organizations will look to such CMOs as the means of learning quickly about the characteristics of and benefits of technological innovations technologies and as convenient means for promoting the uptake of such technologies within their own organizations.

TE

651

4.2. Position of the CMO in method diffusion

EC

649 650

R

648

661

R

662 663

5. DSDM

In this section we use DSDM and its diffusion as case material to help refine the knowledge-based model of diffusion described above. DSDM is chosen for a number of reasons:

C

667

O

664 665 666

671 672 673 674 675

† DSDM is an ISDM. It is a published, ‘open’, standard for RAD work; † DSDM is both a technology supplier and a CMO. It constitutes both an organization which creates and maintains an ISDM and an agency for the promotion of RAD practice in organizations; † Successful example of technological diffusion. DSDM, because of its contingent nature, has been highly successful at diffusing amongst European organizations and organizations world-wide.

U

670

N

668 669

676 677 678 679 680 681 682 683 684 685 686 687

O O

632

PR

631



688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720

ARTICLE IN PRESS P. Beynon-Davies, M.D. Williams / Journal of Strategic Information Systems xx (0000) 1–18

9 766

722

767

723 724

768 769

725

770

726

771

727

772

728

773

729

774

730

775

731 732

776 777

O O

721

733

778

734

779

735

PR

780

736 737 738

D

739 740 741 743

TE

742

Fig. 1. A knowledge-based model of diffusion.

744 745

5.1. DSDM as ISDM

753 754 755 756 757 758 759 760 761 762 763 764 765

R

R

752

O

751

† Model of the development process. DSDM utilises an iterative or incremental model of the development process. This model defines four key phases with iteration both within and between phases, as illustrated in Fig. 2; † Set of techniques. DSDM prescribes the use of a number of new development techniques such as Joint Requirements Planning workshops, Joint Application Design workshops and timeboxing, but generally adopts traditional systems; analysis techniques such as Entity Relationship diagrams in a contingent way † Documentation method. The method suggests a loose set of documentation approaches and generally expects that documentation is kept to a minimum within IS projects;

C

750

N

749

DSDM was created in 1994 by a number of representatives from a group of UK companies. The DSDM consortium produced a first version of its manual in 1995, a second version of the manual in 1996, and a third and less radical revision of the manual was published in 1997. The 1997 version of the manual was extended with a number of socalled ‘white papers’ and has now been published as version 4 of the method (Consortium, 2002). DSDM can be characterized as an ISDM in that it provides elements in each of the five areas used to define an ISDM:

U

747 748

EC

746


781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810

ARTICLE IN PRESS 10

P. Beynon-Davies, M.D. Williams / Journal of Strategic Information Systems xx (0000) 1–18 856

812

857

813 814

858 859

815

860

816

861

817

862

818

863

819

864

820

865

821 822

866 867

O O

811

823

868

824

869

825

PR

870

826 827 828

831 832

TE

Fig. 2. Phases of DSDM.

833 834

837 838 839 840 841

EC

836

† Fit between documentation method and techniques. Some indication is provided in the DSDM manual as to how various techniques and documentation standards may be used contingently in relation to a project; † Philosophy. DSDM utilises a standard philosophy founded in rational businessoriented performance. However, unusually for an ISDM in this tradition, there is also some acknowledgement of cultural issues and of organizational learning within its description of the method.

R

835

5.2. DSDM as CMO

849 850 851 852 853 854 855

C

N

848

Many organizations within the UK have subscribed to the DSDM approach. Currently the consortium lists over 200 full members including BT, Computer Associates, Oracle, Norwich Union, Coopers and Lybrand, Lloyds Bank, the Post Office etc. DSDM also has a presence in Eire, Benelux, Denmark and France. It has been used also by US development organizations. Participating organizations have adopted and adapted the DSDM framework, utilised DSDM as their RAD route in corporate methodologies, and trained personnel in DSDM. DSDM can be conceived of as a CMO since it has enabled RAD as a ‘technology’ to persist over time. Note that we use the term technology here in the broadest sense of the term (familiar in SCOT analyses). In particular, DSDM have:

U

847

O

844 845 846

R

842 843

D

829 830


871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900

ARTICLE IN PRESS P. Beynon-Davies, M.D. Williams / Journal of Strategic Information Systems xx (0000) 1–18 901 902 903 904

† † † †

11

Bundled and De-contextualised RAD Knowledge; Organized a Supply Push for RAD; Developed Institutional Links; Supported Boundary Spanning Activities.

946 947 948 949

905

950

906

951

907 908

952

6. Cases of method diffusion

953

909 911 912 913

954

We have preliminary evidence of the way in which DSDM has diffused internally within two organizations. The case material described here was gathered using ethnographic approaches as part of a larger study of RAD. A brief description of the context in each organization and the process by which DSDM was diffused is given below.

914

924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945

PR

D

TE

923

EC

922

R

921

R

919 920

Organization 1 (O1) maintains a highly distributed IS organization. It currently has three major development centres in the UK, although many development staff are seconded to particular business units. There is something of an internal market for IS within this organization in that IS/IT projects need to be driven by strong business cases and indeed need business sponsorship to proceed. One might expect that RAD, because of its emphasis on business need, would be likely to thrive in this environment. O1 has maintained a waterfall-based internal ISDM, resembling SSADM, for a number of years. The IS division of O1 established a RAD route within their corporate methodology a number of years before the emergence of DSDM which they named formula 1 (to emphasise the rapidity of the development approach). A number of people within O1 were sent to the United States to collate collective experience on RAD which led to the development of the internal RAD route. Following this, several RAD teams were set up at the various development centres to promote RAD internally within O1. O1 were an early subscriber to DSDM. A number of RAD consultants within the company who were originally promoting formula 1 had input into the working of the consortium. A number of projects had been undertaken in O1 using a RAD approach and prior to the organization’s membership of DSDM. O1 is now a key member of the DSDM consortium. RAD consultants within O1 run both one-day RAD awareness courses and three-day RAD practitioner courses both for internal members of the organization and for external persons. The organization promotes a rational selection process (a suitability filter) for applying RAD to development work. This filter is applied to all new development projects to assess their suitability for RAD. There is little evidence of the actual number of RAD projects being conducted. However, we have been told that as much as 20% of projects are expected to follow a RAD approach. Currently, RAD projects seem to be centred in particular business areas where good relationships have been established with the RAD teams and in which the various issues of the ‘culture change’ needed for successful RAD have been addressed.

O

918

C

917

958 960

6.1. Organization 1

N

916

956 957 959

U

915

955

O O

910


961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990

ARTICLE IN PRESS 12 991


6.2. Organization 2

1036

992

999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019

R

1020 1021

6.3. Bundling knowledge

R

1022 1023

1028 1029 1030 1031 1032 1033 1034 1035

O

C

1027

N

1025 1026

RAD can be seen to be a set of old and new ideas from the ISDM area. DSDM can be seen to be an attempt to package or bundle RAD - making it easier for consultants to market RAD, and for persons to become accredited suppliers of RAD. One might argue that most ISDMs are attempts to commodify system development knowledge in this way. RAD in general, and DSDM in particular, may be viewed as being a response to increasing levels of uncertainty both in the business and development domain (Beynon-Davies et al., 1999. DSDM has also served as a focus for amplifying current problems with the software development process such as the contrast between business and engineering quality associated with IT systems. They have also interpreted current events in the ISD arena and have offered a vision of the benefits of a more iterative approach to systems development such as more ownership and greater commitment from end-users as well as the production of more effective business systems.

U

1024

1038 1039 1040 1041 1042 1043 1044 1045 1046 1047

O O

998

PR

997

D

996

TE

995

1037

In contrast to O1, organization 2 (O2) maintains a highly centralised IS organization. The IS service is situated in a large purpose built site, providing system development services to the rest of the organization. However, the internal IS service of O2 is not the sole provider of such services. Business sectors are able to purchase IT services from outside the organization. O2 maintains a highly specified development process with two levels of corporate ISDM—the first level describes some overall corporate standards; the second level specifies a number of distinct development routes in fine detail. Overall, the internal ISDM forms a waterfall-based approach, again similar in form to SSADM. In our discussion with organization members, the main rationale expressed for adopting RAD was the increased competition from outside agencies and developers tendering for development work which was traditionally the domain of the internal IS service. Such outside agencies were seen as not being fettered with needing to follow the in-house method. O2 had not developed any RAD route before DSDM. It became a DSDM member in 1996 following a developing interest from a small group of practitioners within O2. Explicitly, RAD and DSDM were promoted by the New Technology group within O2 and an internal methods expert (a DSDM member). The New Technology group comprised 3 – 5 persons tasked with investigating the potentialities of innovative technologies for O2. As part of this remit a so-called ‘Flagship’ project was chosen to test the utility of RAD/DSDM for the organization. This project completed successfully and a RAD/DSDM route has now been written into the formal corporate method. Besides the formal RAD route we have little evidence that RAD has diffused effectively within O2. This may have something to do with the highly bureaucratic structure of the development centre at O2— something that Jones and King (Jones and King 1998) propose as being anathema to the adoption of RAD. Another problem may be the difficulty in gaining commitment from business users to this change of approach within O2. On the flagship project, for instance, gaining appropriate levels of user involvement proved problematic.

EC

993 994


1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080


13

6.4. Supply push

1126

1082 1083 1084

1127

DSDM has carried out a number of activities which have provided a supply push for RAD as a soft technology. It has:

1085

1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109

6.5. Institutional links

1118 1119 1120 1121 1122 1123 1124 1125

O

C

1117

N

1115 1116

DSDM has been adopted by a number of IS/IT consultancies and proffered as a service. Close links have been developed between DSDM members and members of professional associations, particularly the BCS, with the BCS eventually lending its weight to accrediting DSDM practitioners. Links have also been forged with other agencies - for instance, a number of UK universities joined DSDM and have started to offer DSDM accreditation to their students. 6.6. Boundary spanning

U

1114

R

1112 1113

R

1110 1111

1131 1132 1133 1134 1135 1136 1137

O O

1090

PR

1089

D

1088

TE

1087

1130

† Raised funds through organizational membership both of an individual and corporate kind. Many organizations within the UK have subscribed to the DSDM approach. Two levels of subscription are provided to the consortium. The first level entitles individuals and groups access to the publications and events of the consortium. Level two subscription entitles the organization to participation in the continuing formulation of the method; † Mobilized resources. DSDM now administers an accreditation programme (linked to British Computer Society (BCS) professional accreditation). A key part of this accreditation involves vetting training courses run by commercial vendors. There is also a route for consultants to become accredited in the use of DSDM; † Held meetings throughout the UK to proselytize the benefits of the approach. Members of the consortium have promoted the method in venues such as BCS specialist group meetings and other industrial seminars. DSDM members have also been particularly active in ‘evangelizing’ the benefits of this RAD approach at one-day awareness events throughout the UK. An annual DSDM conference has become an established event, having run for a number of years; † Formulated positions. The DSDM consortium produces a regular newsletter for its members. The consortium also regularly produces what it calls ‘white’ papers on key extensions to the method. A key member of the consortium, Jennifer Stapleton, produced a book in the late 1990s which documents the ISDM (Stapleton 1997). This literature is employed particularly to formulate positions in opposition to ‘traditional’ development approaches such as those philosophically aligned to the waterfall model. The literature also promotes the opposition of ‘true’ or 2nd generation RAD to ‘hacking’ or 1st generation RAD. DSDM portrays itself as a disciplined development approach interested in the business quality of systems;

EC

1086

1128 1129

Boundary spanning activities and user pull were certainly evident in the two organizations studied. DSDM has been extremely successful in enlisting and supporting champions of the method within particular organizations. In organization O1 a number of business analysts


1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170

ARTICLE IN PRESS 14

1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199

7. Conclusions

R

1200 1201

7.1. Summary

R

1202 1203

1208 1209 1210 1211 1212 1213 1214 1215

O

C

1207

N

1205 1206

Method diffusion has many characteristics in common with social movements. The idea of computerization movements provides a valuable counter to the rational model of diffusion characteristic of portrayals of technological diffusion. ISDMs are an important ‘soft’ technology in that, at least overtly, they are meant to embody ‘best’ practice in IS development. This idea of best practice supports a rational accounting of method diffusion. The explanation of low take-up of methods and considerable adaptation of methods within organizations is usually accounted for in rational terms such as the cumbersome nature of ISDMs and the difficulties of implementation (Chatzoglou and Macauley 1996). In this paper we have provided a social constructivist interpretation of the diffusion of ISDMs. In particular, we have emphasized the importance of computerisation movement organizations to the take-up of technologies and considered the way in which organizations such as DSDM can be regarded as CMOs. We have also considered

U

1204

1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227

O O

1176

PR

1175

D

1173 1174

were originally promoting an in-house development approach with RAD features. Members of this consultant group were early participants in the setting up of the DSDM consortium and consequently the organization became one of the first large-scale adopters of the development approach. RAD consultants within this organization now run both oneday RAD awareness courses and three-day RAD practitioner courses for both internal members of the organization and external persons. In O2 RAD and DSDM were promoted by the New Technology group within the organization and an internal methods expert (a DSDM member). A RAD/DSDM route has now been written into the formal corporate ISDM. O1 already had a base of experience of RAD prior to the creation of the DSDM consortium. Organizational members skilled in this technology were quick to jump on the DSDM bandwagon, although interestingly their adoption of DSDM as a corporate standard does not seem to have changed their in-house RAD approach to any significant degree. Within O2 the New Technology Group were looking for new development approaches to help solve a severe image problem at the development center. Two members of the new technology group attended commercial seminars on RAD and DSDM as well as those run by professional associations such as the BCS. Because of their prior base of knowledge, skills and personnel, O1 was far more able to de-contextualise DSDM as a technology for the organization. The major effort appears to have been one of translation between the prescriptions of DSDM and that of the existing in-house RAD method. The second organization had no prior experience of RAD approaches. DSDM was selected on the basis of championing by members of the New Technology Group but experienced a number of difficulties of adaptation into the accepted practices of the organization. Although a DSDM route was written into the corporate ISDM standards, the flagship project experienced a number of problems in adapting DSDM prescriptions to the expectations of both developers and users within the organization.

TE

1172

EC

1171



1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260

ARTICLE IN PRESS P. Beynon-Davies, M.D. Williams / Journal of Strategic Information Systems xx (0000) 1–18 1261 1262 1263 1264 1265

15

the diffusion of ISDMs in the light of Newell et al. (2000) knowledge-based focus on technological diffusion. This model provides a possible new perspective on the take-up of ISDMs. Methods being knowledge bundles frequently experience difficulties with the de-contextualisation process, particularly if there is an established knowledge-base of development practice.

1266 1267

1314

1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305

1316 1317

O O

PR

D

1282

TE

1281

EC

1279 1280

R

1278

R

1277

O

1276

C

1275

1315

† Methods are particularly useful as knowledge bundles for diffusion. DSDM can be seen to be an attempt to package or bundle RAD, thus making it easier for its transmission and promotion amongst organizations, consultants and professional organizations; † Hype is important to the process of method diffusion. The evangelizing of the held benefits of method innovations is a natural consequence of the knowledge-based nature of the method diffusion process. The evangelizing process is a critical part of supply push. A new development method has to be seen as beneficial to be adopted by users; † CMOs are important to the effective external diffusion of ISDMs. In terms of external diffusion CMOs are important to ISDM diffusion in providing a coherent discourse and rationale for method innovation. In our case, DSDM has proven to provide a significant focus in organizing the debate about the applicability of RAD to information systems development, at least within the UK context. CMOs therefore seem to play, along with professional associations, an important mediating role between suppliers and users of technology. They are critical both to the processes of supplier push and user pull in technological diffusion; † ISDM diffusion occurs through processes of user pull, supply push and boundary spanning. ISDM knowledge bundles are promoted by consultancies in professional associations and supported by other institutions such as universities. Organizations are attracted to the ‘best practice’ promoted by ISDMs and their CMOs. Individuals within such organizations initiate the diffusion process by promoting the ideology of an ISDM within the organization; † Significant external diffusion does not necessarily need to occur prior to internal diffusion. Our case material suggests that some organizations are early adopters of innovative approaches and may significantly contribute to the formation of CMOs within the domain; † Internal diffusion is enabled by the recruitment of organizational members to such CMOs. In this process they become internal champions for the ISDM and communicate the shared beliefs of the CMO internally within the organization. Both of the organizations studied maintained large development centers, an in-house monolithic method, and RAD was consequently a relatively new innovation in both organizational environments. In one organization a small group of champions were in the early stages of promoting the ISDM approach. In the other, champions of the ISDM had successfully developed a formal role for themselves in promoting the innovation; † Links to external CMOs via internal and external consultants are important vehicles for supporting the diffusion process. The existence of external CMOs lends weight to

N

1274

1310

1313

The main lessons to be drawn are:

U

1273

1308 1309

1312

1270 1271 1272

1307

1311

7.2. Lessons

1268 1269

1306


1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350

ARTICLE IN PRESS 16 1351 1352 1353 1354 1355 1356 1357 1358


the need for organizational adoption. However, the success of ISDM adoption is also influenced by the interplay between the activities specified in the method and the structure and culture of an organization. For instance, we have evidence of the difficulties inherent in utilizing RAD within a highly centralized and bureaucratic organization such as O2, a conclusion supported by the work of Jones and King (1998). For instance, in such an organization, relatively low degrees of user empowerment are commonplace amongst projects. This makes it difficult to introduce changes to the expected levels of such involvement on projects—a key principle of RAD work.

1359

1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379

R

1380

8. Uncited reference

1383

Wynekoop and Russo, 1995.

O

1384

C

1385 1386 1387

Acknowledgements

1391 1392 1393 1394 1395

Thanks to the reviewers of this paper for their comments. An early version of this paper was published in the IFIP TC8/WG8.2 Working Conference on Realigning Research and Practice in Information Systems Development: the social and organizational perspective. Boise, Idaho. The work on this project was funded by the UK Economic and Social Research Council (Grant No: R000 23 5505). We would like to thank all those at participating organizations who agreed to talk to us about, as well as letting us observe, their RAD work. Thanks also

U

1390

N

1388 1389

R

1381 1382

1400 1401 1402 1403

1406 1407

O O

1365

PR

1364

We acknowledge that while we do not yet have sufficient data to provide substantive evidence to support the propositions we have outlined, the paper does provide preliminary evidence that validates a number of aspects of the knowledge-focused perspective on diffusion described. The model clearly needs further refinement. For instance, the extended model as described does not satisfactorily cover the internal diffusion processes for ISDMs. It does not provide us with sufficient understanding of the ways in which independent methods are adapted by particular organizations or explain the way in which methods are developed in-house. Hence, the model clearly needs further verification and extension in some larger case-study work. We are particularly interested in extending such a study amongst a range of organizations. Two types of longitudinal research are also required. We are interested in re-visiting some of the organizations to which we gained initial access some years ago with the aim of studying the continuing diffusion of DSDM as a method within these companies. This will provide valuable data on internal method diffusion processes and will enable connection to be made to the existing literature on method adaptation and use. There is also much potential for historical studies of method diffusion. A key instance here is to develop a case study of some other method such as SSADM or an OO method to refine the knowledgebased model of diffusion discussed.

D

1363

1398 1399

1405

TE

1361 1362

1397

1404

7.3. Further work

EC

1360

1396


1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440

ARTICLE IN PRESS P. Beynon-Davies, M.D. Williams / Journal of Strategic Information Systems xx (0000) 1–18 1441 1442

17

to colleagues on this research project: Douglas Tudhope, Hugh Mackay, Roger Slack and Chris Carne.

1443 1444 1445

1490

References

1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485

1492 1493 1494 1495 1496 1497

O O

PR

1458

D

1457

TE

1456

EC

1455

R

1454

R

1453

O

1451 1452

C

1450

N

1449

1491

Beynon-Davies, P., 1998. Information Systems Development, third ed., An Introduction to Information Systems Engineering, Macmillan, London. Beynon-Davies, P., 2002. Information Systems, An Introduction To Informatics in Organizations, Houndmills, Basingstoke. Beynon-Davies, P., Carne, C., Mackay, H., Tudhope, D., 1999. Rapid application development: an empirical review. European Journal of Information Systems 8 (2), 211– 223. Bijker, W.E., Hughes, T.P., Pinch, T.J. (Eds.), 1987. The Social Construction Of Technological Systems: New Directions in The Sociology and History of Technology, MIT Press, Cambridge, Mass. Button, G., Sharrock, W., 1993. In: Firth, A., (Ed.), Practices in the Work of Ordering Software Development, The Discourse of Negotiation, Oxford, Pergamon, pp. 159–180. Chatzoglou, P.D., Macauley, L.A., 1996. Requirements capture and IS methodologies. Information Systems Journal 6 (2), 209– 225. Consortium, 2002. Dynamic Systems Development Method. Version 4, DSDM Consortium, 4. Edwards, H.M., Thompson, J.B., Smith, P., 1989. Experiences in the use of SSADM: series of case studies. Part 1: first-time users. Information and Software Technology 20 (2), 411– 426. Fitzgerald, B., 1996. An Investigation into the Use of Systems Development Methodologies in Practice, European Conference on Information Systems,. Fitzgerald, B., 1997. Systems Development Methodologies, A Need for New Canons,. Iacono, S., Kling, R., 1996. In: Kling, R., (Ed.), Computerisation Movements and Tales of Technological Utopianism, Computerisation and Controversy: Value Conflicts and Social Choices, Wiley, Chichester. Jacobson, I., Booch, G., Rumbaugh, J., 1999. The Unified Software Development Process, AddisonWesley, Reading, Mass. Jones, T., King, S.F., 1998. Flexible systems for changing organizations: implementing RAD. European Journal of Information Systems 7 (1), 61 –73. Kautz, K., McMaster, T., 1994. The failure to introduce systems development methods: a factorbased analysis. In: Levine, L., (Ed.), Diffusion, Transfer and Implementation of Information Technology, Elsevier Science BV, Amsterdam, pp. 275– 287. Mumford, E., 1983. Designing Participatively, Manchester Business School Press, Manchester. Newell, S., Swan, J.A., Galliers, R.D., 2000. A knowledge-focussed perspective on the diffusion and adaptation of complex information technologies: the BPR example. Information Systems Journal 10 (1), 239– 259. Orlikowski, W.T., Gash, T.C., 1994. Technological frames: making sense of information technology in organizations. ACM Transactions on Information Systems 12 (2), 17 – 207. Pinch, T.J., Bijker, W.B., 1987. In: Bijker, W.J., Hughes, T.P., Pinch, T. (Eds.), The Social Construction of Facts and Artifacts: or How The Sociology Of Science and The Sociology of Technology Might Benefit Each Other, The Social Construction of Technological Systems, MIT Press, Cambrdge, Mass. Rogers, E.M., 1995. Diffusion of Innovations, fourth ed., Free Press, New York. Rumbaugh, J., Jacobson, I., Booch, G., 1999. The Unified Modelling Language Reference Manual, Addison-Wesley, Reading, Mass.

U

1448

1487 1488 1489

1446 1447

1486


1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530

ARTICLE IN PRESS 18

1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545

Stapleton, J., 1997. DSDM—Dynamic Systems Development Method: the Method in Practice, Addison-Wesley, Harlow, England. Stolterman, E., 1991. How system designers think about design and methods: some reflections based on an interview study. Information and Software Technology 30 (1), 137– 150. Suchman, L.A., 1987. Plans and Situated Actions: the Problem of Human – Machine Communication, Cambridge University Press, Cambridge. Veryard, R., 1987. Implementing a methodology. Information and Software Technology 29 (9), 469– 474. Walsham, G., 1993. Interpreting Information Systems in Organizations, Willey, Chichester. Wastell, D., 1996. The fetish of technique: methodology as a social defence. Information Systems Journal 6 (2), 25 – 40. Weaver, P.L., Lambrou, N., Walkley, M., 1998. Practical SSADM 4 þ , Pitman, London. Wynekoop, J.L., Russo, N.L., 1995. Systems development methodologies: unanswered questions. Journal of Information Technology 10 (1), 65 – 73. Wynekoop, J.L., Russo, N.L., 1997. Studying system development methodologies: an examination of research methods. Information Systems Journal 7 (1), 47 – 65.

1546 1547 1548

D

1549 1550 1551

TE

1552 1553 1554 1555

EC

1556 1557 1558 1559

R

1560 1561

R

1562 1563

O

1564

C

1565 1566 1567

1571 1572

U

1570

N

1568 1569

1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587

O O

1532

PR

1531


1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617

1573 1574

1618 1619

1575

1620
