National differences and ERP implementation: issues ...

Available online at www.sciencedirect.com

Omega 32 (2004) 361 – 371 www.elsevier.com/locate/dsw

National dierences and ERP implementation: issues and challenges Chwen Sheua;∗ , Bongsug Chaea , Chen-Lung Yangb a Department

of Management, College of Business Administration, Kansas State University, Manhattan, KS 66506, USA of Industrial Management, Comprehensive Research Center, Chung-Hua University, Taiwan

b Department

Received 5 April 2003; accepted 3 February 2004

Abstract Multinational ERP implementation introduces another dimension of complexity—national dierences—into the already complex nature of ERP implementation in the context of global information management. This study reviewed several issues critical to the success of international ERP implementation. Using both case research and secondary data, we examined ERP implementation at several multinational companies in the US, Taiwan, China, and Europe. Our primary purpose was to investigate the dimensions of national dierences and how they aect ERP implementation practices across nations. Our 9ndings suggest that language, culture, politics, government regulations, management style, and labor skills impact various ERP implementation practices at dierent countries. Understanding such eects will enable companies to be more proactive in planning project budget and duration. ? 2004 Elsevier Ltd. All rights reserved. Keywords: ERP implementation; Supply chain management; International operations management; Global information technology and information management; Case research

1. Introduction An ERP is an information system that manages, through integration, all aspects of a business including production planning, purchasing, manufacturing, sales, distribution, accounting, and customer service [1]. In the past few years, ERP has become a “must have” system for many 9rms to improve competitiveness. More than 60% of US companies have installed or planned to install a packaged ERP system [2]. The popularity of ERP systems is also evidenced by its sales exceeding $30 billion in 2002, an increase of 300% since the late 1990s. Information managed by ERP systems can play an active role in international supply chain systems to gain

∗ Corresponding author. Tel.: +1-785-532-4363; fax: +1-785532-7024. E-mail addresses: [email protected] (C. Sheu), [email protected] (B. Chae), [email protected] (C.-L. Yang).

0305-0483/$ - see front matter ? 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.omega.2004.02.001

a competitive advantage [3]. Take a simple ERP function as an example. When a sales person for a multinational company enters an order from a customer in any location in the world, the transaction data can permeate the entire supply chain’s (including suppliers’) information system. The system updates the inventory of parts and supplies automatically, changing the production schedules of overseas facilities and balance sheets at headquarters as well. Thus, the employees of dierent departments in various countries quickly have the information needed to complete the processing of their jobs. Feedback is fast and eKcient. From this information the sales person can inform customers of updated delivery dates, and the managers can receive accurate inventory status immediately. ERP facilitates the enterprise-wide integration of information by tying together suppliers, distributors, and customers without geographical restrictions. To summarize, an ERP system provides multinational organizations with extensive information and coordination of supply chain functions. In this vein, Davenport [43] notes, “for the 9rst time ever, information will Now

362

C. Sheu et al. / Omega 32 (2004) 361 – 371

seamlessly across diverse business functions, business units and geographic boundaries”. While companies worldwide have made substantial investments in the installation of ERP systems, implementation has proven to be unexpectedly diKcult, and 9nal bene9ts have been uncertain. Several researchers have concluded that failures are usually the result of business problems instead of technical diKculties [4–6]. The technical aspect of installing ERP systems is undoubtedly critical, but the management aspect of ERP systems could have an even greater impact on the success of the system or use. Davenport [4], Bowersox et al. [7], and Jacobs and Whybark [8] asserted that ERP was not only a software package but also “a way of doing business”. There is no single “best process” to do business as ERP systems assume. “Mis9ts”, the gap between the functionality oered by ERP systems and that required by the adopting company, would arise from country-speci9c or company-speci9c requirements [9]. These mis9ts can be related to data format, operating procedures, and output format, which force companies to make package and/or organizational adaptation during the implementation process. Studies indicate that no ERP implementation can succeed without resolving those mis9ts [9,10]. Even worse, 9rms can lose their source of competitive advantage by adopting ERP systems that do not 9t their business strategies [4]. 2. Multinational and multi-site ERP implementation In practice, the magnitude and impact of such mis9ts will be even stronger when ERP is implemented across multiple facilities with national dierences. Multisite ERP implementation costs more, takes longer, and fails more often due to the emergent technical as well as to organizational elements (complexity of module integration among multiple sites, organizational conNict, politics, etc.). Using the furniture industry as a reference, Jacobs and Whybark [8] illustrated how ERP implementation at multiple facilities could lead to disaster unless the dierences in corporate culture, production methods, and customer demand could be reconciled. Markus et al. [11] also recognized the complexities of multisite ERP implementation. When organizations are geographically dispersed, implementing ERP systems involves unique technical and managerial challenges. If facilities are located in dierent countries, national dierences such as national culture, language, management style, politics, regulations, customs, etc. can aect “the way of doing business” [12,13]. Previous studies have already suggested that these national dierences have an impact on dierent areas of information technology such as technology transfer between dierent countries [14], global information infrastructure building [15,16], global information management [17], the role of technology [18], global technology planning, adoption and implementation

[19], and others [18,20,21]. In line with these studies, we contend that no universal ERP system can be implemented in dierent countries successfully without resolving mis9ts resulting from national dierences. The business models, including operating processes underlying most ERP software packages, reNect European and US industry practices. Such operating processes are likely to be dierent in Asian countries, having evolved in a dierent cultural, economic, and regulatory environment. As an example, Soh et al. [9] studied ERP implementation practices in a hospital in Singapore and cautioned about potential cultural incongruence in implementing ERP in Asia. Potential mis9ts could arise from areas including data format, operational procedures (e.g., billing and collection), and output format. Resolving such mis9ts has required extra implementation time and expense. Recently several major POM and IS journals such as Communications of the ACM (e.g., [22]), European Journal of Operational Research (e.g., [23]), The DATA BASE for Advances in Information Systems (e.g., [24]), and Information Systems Frontier (e.g., [25]) have dealt with ERP implementation issues, but none of the articles from the special issues directly addresses national dierences and cultural issues in their studies. The literature has not oered comprehensive studies of dimensions of national dierences and how they aect the ERP implementation. Without resolving national dierences, organizations are unlikely to apply ERP successfully in integrating the international supply chain systems. Understanding such eects will enable companies to be more proactive in planning for multinational ERP implementation. Speci9cally we are interested in the following research questions: 1. Which aspect of multinational ERP implementation is aected by national dierences? How and why is it aected? 2. What factors of national dierences aect multinational ERP implementation practices? How and why do they aect the implementation? The next section describes the research methodology used to collect data from six companies that have implemented ERP in multiple countries including US, Taiwan and China, Norway, and several European countries.

3. Research design We used both case study [26,27] and secondary data research method [28,29] to collect necessary data in this project. Case research involves the integration of data gained from organizations of similar nature which have implemented ERP, while the secondary data analysis uses literature review on multinational ERP implementation in Europe.

C. Sheu et al. / Omega 32 (2004) 361 – 371

A case study is “an empirical inquiry that investigates a contemporary phenomenon within its real-life context, especially when the boundaries between the phenomenon and context are not clearly evident” [27, p. 13]. Case studies are the preferred strategy when “how” and “why” questions are being posed, and the researcher has little control over events [27]. Case study research can be used to achieve various research aims: to provide descriptions of phenomenon [27], develop theory [26], and test theory [48]. It provides evidence for hypothesis generation and for exploration of areas for areas where existing knowledge is limited [49]. The case study research method has become the most widely used qualitative research method in information systems research [50], and in operations management (OM) research there is an increasing interest and application of the method among OM researchers [37,47,51–53]. For example, Ellram [53] placed cases study on the logistics scienti9c agenda. Meredith [51] points out the unfamiliarity with the nature of theory building using case and 9eld study methods in the OM 9eld. In a recent article, Voss et al. [37] claim that “case research has consistently been one of the most powerful research methods in operations management” (p. 195). Some studies [e.g., 37,52] provide guidelines for conducting case study research in OM. Our study employs multiple cases to increase generality [27]. First, we identi9ed four manufacturing companies in the United States, Taiwan, and China for the purpose of on-site interviews, observation, and data collection. Three of the four facilities interviewed are located in Taiwan, though their headquarters may be located in another country, and the fourth plant is in the United States. Three companies have implemented ERP systems across nations, while another is in the process of its overseas implementation. We collected information regarding the “cross-national” ERP implementation based on archival documentation and interviews with managers who were involved in overseas ERP implementation. Three companies are from the high-tech industry, and the remaining one is from the personal care industry. In addition to case study method we also performed secondary data analysis [28,29] based on the literature review on multinational ERP implementation in Europe. Secondary data can be de9ned as data collected by others and archived in some form, not speci9cally for the research question at hand [28]. It has been contended that secondary data analysis can oer a number of bene9ts to a researcher [30,44,45,47]. For instance, Hakim [30, p. 16] suggests that relying on secondary data rather than gathering primary data can actually bene9t the development of theory. Secondary data analysis enables researchers to focus more closely about the theoretical aims and substantive issues of the study rather than on the practical and methodological problems of collecting new data. According to Larsson [46] potentially this type of research methodology can bridge the nomothetic–idiographic research gap using prior empirical studies. Jarvenpaa [44] recognized that secondary data in the IS and POM 9eld can provide a potential ‘gold mine’ of empirical

363

research. Lewis [47] calls for greater employment of existing case studies by OM researchers. Existing case studies oer a potentially eective and eKcient means for comparing complex and disparate operations settings [47]. There are two general methodological issues for conducting a secondary analysis of a qualitative data set: (1) the degree to which the data generated by individual qualitative methods are amenable to a secondary analysis, and (2) the extent to which the research purpose of the secondary analysis can dier from that of the primary study without invalidating the eort and the 9ndings [45]. To address these two issues in this study, the ERP cases of two Norwegian 9rms, Statoil [31–33] and Hydro Agri Europe (HAE) [34–36], were identi9ed and analyzed. With six ERP cases from dierent countries and continents we attempt to use the replication logic through multiple cases [26,27] and consequently improve the generalization of the 9ndings. Table 1 summarizes company pro9les consisting of industry, products, 9rm size, time after ERP implementation, ERP software, facilities locations, etc. In general, all companies have the following characteristics: 1. they are all multinational corporations with facilities located in more than one country; 2. they all have a clear and long-term vision; and 3. they have implemented ERP for 2 years or longer. For those companies that involved case study method, a case study protocol was designed and used to guide the structured interview and data collection. The protocol was reviewed and pre-tested by a group of researchers and senior managers. On-site interviews and observations were conducted after the pilot case study was concluded. Multiple sources of evidence were used to enhance the reliability and validity of this study [27,37]. 4. Data analysis and results Once data were collected, Miles and Huberman’s [38] tabular method was also used to guide the data analysis. Patterns between national dierences and ERP implementation practices were identi9ed for each individual case. To ensure quality of research design and our 9ndings, we carefully conducted several validity tests indicated as necessary by previous studies [27,37] (Table 2). In the remaining part of this section we review the information collected in relation to the background of the six cases and present the inNuence of various national dierence variables on their ERP implementation practices. 4.1. Background information Company NCQ is a notebook manufacturer for several large US computer companies such as Dell, Hewlett

364

C. Sheu et al. / Omega 32 (2004) 361 – 371

Table 1 Company pro9le NCQ

CareM

NoteF

MoniP

Statoil

Hydro Agri Europe (a division of Norsk Hydro)

Industry (Products)

Computer (notebook, PCB)

Consumer personal care, cosmetics

Computer (notebook, PC)

Computer (monitor)

Oil and Gas

Petrochemicals (fertilizer)

Annual Sales

∼US$ 4.0 billions

∼US$ 0.6 billions

∼US$ 1.5 billions

∼US$ 2.2 billions

∼NOK 64 billions (approx. 7 billion USD)

∼NOK 30 billions (approx. 3.2 billion USD)

No. of employees

10,000

4300

5400

5000

16,000

7000

Facility locations

Taiwana , China

USAa (production), Taiwan (telemarking)

Taiwana , USA, China, Czech Rep.

Netherlandsa , Taiwan Taiwan

Norwaya , other European countries

Norwaya , eight other European countries

Languages used

Mandarin

English,Mandarin

Mandarin, English, Czech

Dutch, Mandarin

Norwegian & other languages

English, Norwegian & many other languages

Facilities visited

Taiwan

USA

Taiwan

Taiwan

Secondary data

Secondary data

Time after implementation

3 years

2 years

4 years

3 years

2 years

3 years

ERP software

SAP

J.D. Edwards

Magic or SAP

SAP

SAP

SAP

a Corporate

headquarters.

Packard, IBM, and Gateway. As an original equipment manufacturer, it faces frequent order changes in volume, delivery and product speci9cations. The company has an excellent R& D department, able to quickly respond to various customization orders. With a well-established supplier network, it eectively uses outsourcing and overtime to adapt to changes of production volume and delivery. As requested by its US clients, the company implemented ERP system in Taiwan headquarters in 1999 and extended the implementation to its facilities in China the following year. A project team consisting of CEOs, plant managers, and IT managers was involved in the implementation. CareM is known for high-quality personal care and cosmetics products. It has three manufacturing facilities located in the US and has more than 10 overseas telemarketing of9ces. The decision to implement ERP was driven from forces within the organization. Managers wanted reasonable access to data to make good decisions, and ERP was believed to be the tool to achieve this goal. All three manufacturing facilities implemented ERP systems prior to 2000. Several overseas telemarketing oKces such as Japan, Taiwan and Korea are currently implementing the systems.

NoteF manufactures notebook computers. Its product design and sales functions are located in Taiwan, and key components such as CPU and motherboard are manufactured in China. Depending on customer location, 9nal assembly is performed in either US or Czech plants in order to speed up delivery. The company also owns small assembly facilities in Brazil and Netherlands. Similar to NCQ, NoteF was asked by its US clients to implement ERP systems in 1998. Dierent ERP o shelf software (SAP, Oracle, and Magic) were used in three dierent facilities (US, Czech, and China). Each facility formed its own project team to implement the ERP system, while the headquarters in Taiwan coordinated the eort between facilities. MoniP is one of the largest monitor manufacturers in the world. Its headquarters is located in The Netherlands, while the facilities in Taiwan are responsible for R& D and for coordinating the purchasing and production in nearby Asian countries. MoniP initiated ERP in 1999 with the intention of improving its internal integration and reducing manufacturing cycle time. The implementation was independently performed in Taiwan facilities with minimum coordination with The Netherlands headquarters.

C. Sheu et al. / Omega 32 (2004) 361 – 371

365

Table 2 Research validity tests Test

Approach/tactic used

1. Construct validity Develop suKcient set of operational measures to reduce subjective judgment

(a) Reviewed multiple sources of evidence so that the 9ndings were based on the convergence of both qualitative and quantitative information (b) Established a chain of evidence with suKcient citations to relevant portions of the case study database (c) Asked participants and informants to review draft case study reports

2. External validity Ensure generalization of the 9ndings 3. Reliability Demonstrate that repetition of the data collection procedures will result in the same conclusions

Applied multiple-case replication. Each company was selected and studied so that it either predicted similar results or produced contrasting results (a) Used case study protocol to document the procedures (b) Created a case study database including case study notes, documents and narratives for future reviews and analysis

All tests were conducted based on Yin’s (1994) guidelines.

Statoil is an integrated oil and gas company with business operations in 25 countries. The company is the largest producer in Norway and is the second largest exporter of oil worldwide. The company launched the SAP ERP project in 1996 and 9nished in 2001. The goal was to develop a comprehensive, versatile, and well-aligned global IT infrastructure through ERP. The project team consisted of over 300 people from multibusiness units on the Norwegian continental shelf. The team was delegated the responsibility by corporate management for the success of the project. HAE [34–36] is the largest division in a diversi9ed Norwegian company Norsk Hydro. In the 1990s it was decided that more centralized control and governance were required, and the development of a global corporate information infrastructure was needed. As part of this strategic IT initiation, Hydro has implemented SAP applications at multinational business units, representing nine European countries. The project started in early 1995 at 29 locations, representing nine dierent countries and was planned to be 9nished by mid-1999. The ERP project was managed by a multicultural group of project managers and team leaders, representing seven dierent nationalities. The project team visited several sites and produced a set of best practice processes. 4.2. National di7erences and implementation practices Six categories (culture and language, government/ corporate politics, management style, government regulations, time zone, and labor skills) of national dierences emerged from the data. Those six categories are presented with the framework developed by Ein-Dor et al. [39]. The

National Differences

Multi-National ERP Implementation Practices

Socio-Psychological Group

ERP Adaptation Culture&Language

Management Style

Centralization of Implementation Decisions

Economic/Political Group

Information Sharing Government/Corporate Politics

Regulations/Legal Requirements

Implementation Approach

Demographic Group

Internal Technical Personnel Resource/Labor Skills

Project Duration

Geography/Time Zone

Training Program (Content & Time)

Fig. 1. National dierences and multinational ERP implementation practices.

framework contains a list of national cultural variables aecting global information technology. Ein-Dor et al. [39] suggest that these variables are diKcult to change in global IT implementation. Those variables can be classi9ed into three main groups, socio-psychological, economic and demographic. Fig. 1 assigns the six categories discovered in our study into the three groups of national cultural variables.

366

C. Sheu et al. / Omega 32 (2004) 361 – 371

Table 3 Summary of major 9ndings Category of national dierences

Impacts on implementation practices

1. Culture/ language

• Technical problems in entering data • Cultural resistance • Communication barriers between facilities due to dierent languages • Localized implementations • Alteration of training programs in dierent site

2. Management style

• Dierences in priority setting • Implementation style: either “big bang” or “piecemeal” approach to the implementation • Project duration

3. Government/ corporate politics

• Trans-border information sharing and money transaction • Problems in information sharing between facilities; information hiding • Political conNicts between facilities; mistrust between facilities (host and subsidiaries)

4. Regulations/legal requirements

• DiKculty in developing standardization or universalization due to dierence in forms, tax policy, procedures and others • Substantial customization of ERP packages

5. Internal technical personnel resource/ labour skills

• DiKculty in exercising decentralized ERP implementation due to lack of local personnel resources • Alteration of training programs in dierent sites • Complexity into training and use support

6. Geography/time zone

• Complexity and technical problems in ERP adaptation due to dierent currencies and exchange rates

Additionally, the speci9c relationships between the six categories and various implementation practice items are displayed. The next section discusses the major 9ndings of such relationships. Table 3 summarizes the inNuences of the six categories of national dierences on ERP implementation practices. 5. Major "ndings 5.1. Socio-psychological factors The socio-psychological group includes several variables such as values, beliefs, symbols, tradition, life style, and attitudes toward technological progress, to name a few. Two categories from our study belong to this group of variables. 5.1.1. Culture and language (in8uence adaptation, centralization, and training) We found that language dierence aects the implementation practices in both technical and managerial aspects. First, some ERP packages only support one single bite language

(e.g., English) and one double bite language (e.g., Japanese or Mandarin). Therefore, if English is in the ERP system, then only one double bite language can be used at a time. This technical limitation forced the data entry into the ERP system to use primarily English, so all overseas facilities must know English based on the system. The use of English increased employee resistance to ERP implementation. While technical problems were relatively easier to resolve, communication barriers resulting from language differences were far more diKcult to overcome. In this study, since the facilities that NCQ has are located in either Taiwan or China, there is little diKculty communicating with facilities in China because language and culture are similar. Therefore, it 11 was easy for NCQ to centralize its ERP implementation decisions without getting undue resistance from overseas facilities. In contrast, since the headquarters of NoteF (located in Taiwan) and MoniP (located in the Netherlands) do not speak the same languages as their overseas facilities, communication becomes very challenging. As a result, many implementation decisions, including the selection of ERP software and other implementation practices, were largely decentralized to local facilities. Each site

C. Sheu et al. / Omega 32 (2004) 361 – 371

purchased, installed and maintained its own ERP software without consulting the other. Language dierence also has a compounding eect with internal personnel resources on a training program. In the HAE case, all user documentation and training material delivered by the project was prepared in English only. Local instructors were educated to train employees from their own countries, and local resources had to be developed. The project team had to make sure that the instructors had the country-speci9c knowledge necessary to tailor the training courses to local needs, using local master data and site-speci9c SAP structures. The diKculties of centralizing ERP implementation were due not only to language and communication barriers but also to cultural dierence. Operating processes in Taiwanese organizations are likely to be dierent from their US facilities, each having evolved in a dierent cultural and regulatory environment. As expected, resistance to adopt the same operating processes was very high at overseas facilities. For instance, the US facilities of NoteF refused to adopt the same ERP software packages and con9gurations the headquarters requested, claiming that this software was inconsistent with its current operating processes. Several US employees threatened to resign if they were forced to adjust their culture to the business practices of NoteF. The inNuence of dierent culture on ERP implementation was also found signi9cant in Europe. In the case of Norsk Hydro, for example, initially the company and the project team sought a high degree of centrally de9ned, standardized solutions. Neither decentralization nor locally adopted solution would be tolerated. It turned out that all business units at dierent countries had their own way of doing things because of dierent business processes and local requirements generated by national and local dierences. Thus, the initial plan had to be altered by allowing localized solutions and decentralized ERP implementations. HAE faced similar problems due to language dierences and cultural conNicts among business units. 5.1.2. Management style (in8uences implementation approach and project duration) We refer to management style as the attitude toward setting a priority for implementing an ERP system. This style directly aects the approach and duration of implementation. The three Taiwanese 9rms (NCQ, MoniP, and NoteF) took approximately 8 months to complete the implementation in Taiwanese and Chinese facilities. The length was signi9cantly shorter than that of CareM, a US 9rm that took almost 23 months, and the two European companies that took over 3 years. Mabert et al. [2] found that most US 9rms took more than 13 months to implement their ERP systems. Several possible factors contribute to shorter implementation time in Taiwan and China. First, managers in these countries were highly pressured by severe international competition, and they were extremely committed to implementing an ERP

367

system to increase their competitiveness in the global marketplace. The pressure also came from their US customers such as Dell, Compaq, IBM, Hewlett-Packard, etc. Other than the pressure factor, management personality is another possible factor aecting the implementation duration. Yen et al. [40] asserted that managers in Taiwan have the personality type that takes extra care with technical details and are more determined to complete short-term projects. In NCQ, the ERP project was lead by the CEOs, and all project members were completely removed from their primary responsibilities in order to dedicate themselves to the project. Additionally, all three Taiwanese 9rms took a “big bang” approach and implemented all ERP modules at one time. In contrast, CareM, like many US 9rms, took the “phase by module” approach in its implementation [6]. Its management adopted the attitude of “playing by the rules”, “following the manual” and made no compromise with all required preparation during the implementation. While the eectiveness of these two management styles remains to be studied, 9rms in Taiwan apparently completed the implementation in less time. The dierence in implementation approach can also be explained by Hofstede’s [12] study in cultural dimensions. Hofstede characterized China and Taiwan as low uncertainty avoidance and high collectivism and US as medium uncertainty and high individualism. These national characteristics imply a more aggressive approach to the ERP implementation in China and Taiwan, as observed in this study. In both European cases, management style was more like “piecemeal” than “big bang”. Thus, both projects took much longer than that of three Taiwanese 9rms. This style was the compounding eect of several factors such as the inNuential role of unions in Europe [31], the size of the projects in both cases and local resistance. 5.2. Economic/political factors In Ein-Dor et al.’s [39] framework the economic group includes variables such as scope of internal trade, encouragement of economic ties and balance of trade, to name a few. We modify this group by combining political factors into the economic group and identify two major categories below: government/corporate politics and regulatory requirements. 5.2.1. Government/corporate politics (in8uence adaptation and information sharing) We discovered a unique variable aecting the ERP implementation involving both Taiwan and China, namely their current diplomatic relationship. These two governments have imposed strict regulations on transporting goods, exchanging information, and transferring money between each other. Without formal diplomatic relationships, many business operations between these two countries are currently conducted in a way that packaged ERP software does not permit. Consequently, companies are forced to

368

C. Sheu et al. / Omega 32 (2004) 361 – 371

make substantial eorts to modify ERP software according to their use. This 9nding is consistent with Palvia et al.’s study [20]. Palvia et al. pointed out that political/economic/regulatory factors directly or indirectly inNuence dierent IT-related areas such as transborder data Now restrictions and trade and customs regulations. These political forces are rather diKcult to overcome. Government politics may be inNuential in global ERP implementations including countries in dierent economic systems (e.g., socialism and capitalism) and/or political and governance philosophy (e.g., dictatorship and democracy). Another example of how the political factor aects ERP implementation was found at the corporate level. Political concerns at this level seemed to attribute to current low information sharing between facilities in all companies. All companies expressed concerns over sharing too much information, fearing that individual facilities would compare and argue for getting better or equal treatment (in materials, customer orders or any form of resource support) than other facilities. If so, the corporate headquarters would have more diKculties allocating resources to the whole organization considering the global bene9ts. As a result, companies have purposely limited information sharing between facilities, resulting in diminished bene9ts of ERP systems. 5.2.2. Government regulations/legal requirements (in8uence adaptation) The process of importing/exporting materials and goods diers from country to country. Countries that have mutual tari preferential treatments would have a dierent set of custom forms, import/export tax rates, and custom clearing procedures than those countries that do not. No ERP package handles these variations, and the resolution of such “mis9ts” requires substantial modi9cation in ERP packages. Both NCQ and NoteF had to rewrite the ERP software source code to ensure a 9t with their businesses, an eort unfortunately never anticipated. We also found that tax policies dramatically increased the complexities of multinational ERP implementation due to more complicated operating procedures. Dierent tax rates produced extra work in entering sales orders from dierent countries into the ERP system. Furthermore, tax rate dierentials encouraged companies to develop complex business procedures for saving tax expenses. For instance, NoteF invested in a company located in a country (e.g., British Virgin Islands) where the tax rate is lower than that of Taiwan. NoteF resold its sales orders to this company at a price low enough to avoid paying income taxes. The invested company then found a country with low production cost (e.g., China), which resulted in high pro9t but low income taxes. This tax-saving approach created unique operating procedures and required complicated accounting procedures to deal with transactions between the three regions. A substantial amount of adaptation was made for NCQ to customize ERP packages to its use.

Dierent legal requirements among dierent countries were a signi9cant challenge for European cases. In the HAE case, a seemingly simple factor like the diversity in the local letter formats was an important issue, since the project team had to ensure that the common and legal standards for each of the countries were met. Dierences in legal documents, such as the need for continuous number ranges for invoices belonging to the same company in Italy and Spain, also led to exceptions in how the system could be implemented for the dierent countries. 5.3. Demographic factors In Ein-Dor et al.’s [39] framework, the demographic group includes variables such as geography, internal technical personnel resources, and average educational level. In this group, we identify two major categories: internal technical personnel resources/labor skills and geography/time zone. 5.3.1. Internal technical personnel resources/labor skills (in8uence training and centralization) A typical ERP project requires much internal technical personnel resources for the project completion. Lack of such resources often emerges as an important issue. Nonetheless, levels of labor skill vary from country to country, which could disrupt ERP implementation in dierent countries. This disruption was quite noticeable in two Asian countries. For example, most employees in Taiwan have at least a technical college degree, while the majority of workers in China are high school graduates. When NCQ implemented an ERP system in its Chinese facilities, it discovered that the same training program took twice as much time to complete as at its headquarters in Taiwan. The lack of IT knowledge and skills in Chinese facilities forced NCQ to centralize its implementation decisions. The headquarters had to send a project team to the Chinese facilities to help during the entire implementation period. In the HAE case, the company concluded that it did not have the resources to take responsibility for the required data processing and operation services. Several tasks such as the required network services, site management and support were outsourced to a company in UK. The company organized its business in independent national subsidiaries and was not able to carry out the required coordination across national borders. In addition, some problems were related to the fact that users had to call the help desk in UK when they needed support. The threshold for doing this was quite high for large user groups not speaking English. 5.3.2. Geography/time zone (in8uence adaptation) Depending on speci9c geographical locations, there can be as many as 14 h time dierence between two nations in this study. This time dierence could aect international trade operations from the aspect of exchange rates for currency. Even though many of the European countries have

C. Sheu et al. / Omega 32 (2004) 361 – 371

joined the European Union, some (such as Denmark) do not use the Euro dollar. Denmark’s reluctance requires extra consideration for exchange rates. This example is ampli9ed when considering involvement with the rest of the world. Anticipated rates used to calculate transactions can be greatly dierent than expected once the “next day” rate is provided for the transaction. 6. Suggestions and conclusions Our 9ndings con9rm that national dierences aect multinational ERP implementation practices. Six speci9c categories of national dierences that can aect the implementation were identi9ed and presented with Ein-Dor et al.’s framework [39]. National dierences aect multinational ERP implementation with regard to the type and amount of ERP adaptation, centralization of implementation decisions, information sharing, project duration, project approach, and training program. The relationships shown in Fig. 1 are important to both managers and researchers. From a management point of view, there is a need to recognize each unique national context when adopting an ERP system in dierent countries. Early identi9cation of the relationship between ERP implementation and relevant national dierence variables provides a more accurate basis to budget for contingency implementation funds and allows adequate planning for related implementation practices. Realizing the potential eects of national dierences can also be carefully thought through and is likely to be less reactive. Our results suggest that management should consult and negotiate with ERP software vendors prior to implementation to obtain additional help or to reduce expenses. NoteF clearly learned an expensive lesson by having a 40% overrun on its original budget to complete the implementation. CareM did work with its vendor in the early stages of implementation to reduce cost and risk in its US companies. The company, however, feels it has learned enough to implement ERP in other countries without the need for further assistance from the ERP vendor. Based on our research 9ndings, this could become a costly mistake. Failing to recognize the impact of national dierences on multinational ERP implementation was proven costly by the other 9ve companies. Another suggestion is that of the organization of project teams. We suggest that global ERP implementation projects should be managed by a multi-cultural group of project managers and team leaders. It is important to maintain this diversity within the project team. Cases in this study show that communication between dierent nationalities is restricted due to national dierences such as languages, cultural conNicts, and politics. Lack of communication between host and subsidiaries often results in mistrust, project delay and over budget. For example, in the HAE case [34] the two countries—Spain and Italy—were not represented among the project team leaders and managers. As the result, these two countries had a greater share of change requests and

369

errors reported than their size and business activities. This information is in accordance with a general impression in the project that people from local units would rather contact people from their own country in the central team than the correct person from an organizational point of view. Informal communication among nationalities, critical for the success of ERP projects, was limited, and people tended to use formal documentation channels when dealing with people of other nationalities. Both European cases suggest that ERP implementation in Europe is a very challenging task due to national dierences between countries in Europe. Speci9cally, ERP implementation in Europe is more complex because companies have diverse national cultures that inNuence corporate culture, thereby making the implementation of multinational ERP solutions diKcult. This study supports previous evidence that ERP implementation in the US has been more eective that in Europe because of the more complex European corporate and national cultures [41,42]. Finally, some earlier studies (e.g., [54,55]) suggest geographical dispersion of manufacturing facilities and partners as the source of international OM problems including culture and language dierences, dierences in local policies and regulations and lack of communication. This study shows that while Taiwan and China have the geographical proximity, manufacturing facilities in two countries exhibit very dierent ERP implementation patterns. In other words, geographical distance may not always account for dierences in various cultural dimensions. Further studies are necessary to investigate this issue. 6.1. Future research and limitations For future multinational ERP research we suggest revising and testing the relationships displayed in Fig. 1. This study examined only four multinational companies from two industries, high tech and personal care products, and two secondary cases from two industries, oil and gas and petrochemicals (fertilizers). Although the eects of national dierences on ERP implementation are clearly con9rmed, other countries and industries need to be examined in the future to revise the model in Fig. 1. In line with our suggestion for an international team for a multinational ERP project, another natural extension of this study is investigating what and how strategies can be applied to resolve mis9ts in multinational ERP environment. Knowing the eectiveness of various resolution strategies in dealing with national dierences should help multinationals manage their international supply chain systems more eectively. A factor that does not fall into the theme of national dierences but could also aect global ERP implementation is a 9rm’s competitive environment. Shore and Venkatachalam [14] suggested that a 9rm’s global business strategy and organizational structures are the result of its competitive environment. They also argued that 9rms in a highly competitive market would have little room to accommodate local and

370

C. Sheu et al. / Omega 32 (2004) 361 – 371

national culture. There is a possible relationship between competitive environment, business strategy, and global ERP implementation. For example, 9rms in a highly competitive market are likely to and able to pursue more “centralized” ERP implementation, and such radical approaches would be more acceptable to subsidiaries and users than those in a less competitive market. Future research should conduct more in-depth analysis of the eects of competitive environment and business strategy. To ensure the quality of this case study, we carefully followed previous prescriptions in case study research. For example, we selected theoretically useful cases, used multiple investigators and respondents, conducted reliability and validity tests, collected multiple sources of evidence and systematically searched for cross-case patterns. However, in order to generalize the 9ndings, more case data need to be collected. Finally, due to logistical issues in the companies we studied, we did not interview their overseas locations. Even though we collected necessary information and interviewed those managers who participated in their overseas ERP implementation projects, there still may be possible bias. Regardless of these potential shortcomings, this exploratory study raises several issues that should be considered in future ERP, global information management and technology, and supply chain management research. At the same time, along with others [3] this study has responded to Gallupe and Tan’s call to investigate the impact of IT on the global supply chain [18]. Acknowledgements The authors wish to thank Mr. C.T. Yang for his valuable suggestions and assistance in preparing this paper. References [1] Scalle CX, Cotteleer MJ. Enterprise resources planning (ERP). Boston, MA: Harvard Business School Publishing; 1999. [2] Mabert VA, Soni A, Venkataramanan MA. Enterprise resource planning survey of US manufacturing 9rms. Production and Inventory Management Journal 2000;41(2):52–8. [3] Akkermans H, Bogerd P, Yucesan E, Wassenhove L. The impact of ERP on supply chain management: exploratory 9ndings from a European Delphi Study. European Journal of Operational Research 2003;284–301. [4] Davenport TH. Putting the enterprise into the enterprise system. Harvard Business Review July—August, 1998; 121–31. [5] Kumar L, Hillegersberg J. ERP experiences and evolution. Communications of the ACM 2000;43(3):22–6. [6] Robey D, Ross JW, Boudreau M. Learning to implement enterprise systems: an exploratory study of the dialectics of change. Journal of Management Information Systems, 2002; 17–46. [7] Bowersox DJ, Closes DJ, Hall CT. Beyond ERP—the storm before. Supply Chain Management Review, Winter 1998; 28–36.

[8] Jacobs FR, Whybark DC. Why ERP? A primer on SAP implementation. New York: Irwin/McGraw-Hill; 2000. [9] Soh C, Kien SS, Tay-Yap J. Cultural 9ts and mis9ts: is ERP a universal solution? Communications of the ACM 2000;43(3):47–51. [10] Hong HK, Kim YG. The critical success factors for ERP implementation: an organizational 9t perspective. Journal of Information and Management 2002;40(1):25–40. [11] Markus ML, Tanis C, van Fenema PC. Multisite ERP implementations. Communications of the ACM 2000;43(4): 42–6. [12] Hofstede G. The cultural relativity of organizational practices and theories. Journal of International Business Studies 1983; 75–89. [13] Simchi-Levi D, Kaminsky P, Simchi-Levi E. Designing and managing the supply chain: concepts, strategies and case studies. Boston, MA: Irwin/McGraw-Hill; 2000. [14] Shore B, Venkatachalam AR. Role of national culture in the transfer of information technology. Journal of Strategic Information Systems 1996;5:19–35. [15] Gar9eld MJ, Watson RT. Dierences in national information infrastructures: the reNection of national cultures. Journal of Strategic Information Systems 1998;6:313–37. [16] Png I, Tan B, Wee K. Dimensions of national culture and corporate adoption of IT infrastructure. IEEE Transactions on Engineering Management 2001;48(1):36–45. [17] Lai V. Issues of international information systems management: a perspective of aKliates. Information and Management 2001;38:253–64. [18] Gallupe RB, Tan FB. A research manifesto for global information management. Journal of Global Information Management 1999;7(3):5–18. [19] Jelassi T, Dutta S. Integrating global commercial operations with information technology at BP Chemicals. Journal of Strategic Information Systems 1993;2(1):77–95. [20] Palvia PC, Palvia S, Whitworth J. Global information technology: a meta analysis of key issues. Information and Management 2002;39:403–14. [21] Holland CP, Light B. Global enterprise resource planning implementation. Proceedings of the 32nd Hawaii International Conferences on System Sciences, 1999. [22] Kumar L, Hillegersberg J. ERP experiences and evolution. Communications of the ACM 2000;43(3):22–6. [23] Jacobs FR, Bendoly E. Enterprise resource planning: developments and directions for operations management research. European Journal of Operational Research 2003;146: 233–40. [24] Besson P, Rowe F. ERP Project dynamics and enacted dialogue: perceived understanding, perceived leeway, and the nature of task-related conNicts. The Data Base for Advances in Information System 2001;32(4):47–66. [25] Ross JW, Vitale MR. The ERP revolution, surviving vs. thriving information systems frontiers; special issue of on the future of enterprise resource planning systems 2000;2(2): 233–41. [26] Eisenhardt KM. Building theories from case study research. Academy of Management Review 1989;14(4):532–50. [27] Yin RK. Case study research: design and methods, 2nd ed. Thousand Oaks, CA: Sage Publications; 1994. [28] Stewart D. Secondary research: information sources and methods. Beverly Hills: Sage Publications; 1984.

C. Sheu et al. / Omega 32 (2004) 361 – 371 [29] Hyman H. Secondary analysis of sample surveys: principles, procedures, and potentialities. New York: John Wiley and Sons, Inc.; 1972. [30] Hakim C. Secondary analysis in social research: a guide to data sources and methods with examples. London: George Allen & Unwin; 1982. [31] HepsH V, Monteiro E, SchieNoe P. Implementing multi-site ERP projects: centralization and decentralization revisited. Norway and the Annual Sociology Conference, Norway, 2000. [32] Monteiro E. Actor network theory and information infrastructure. In: Ciborra C, editor. From control to drift. Oxford University Press: Oxford; 2000. [33] Monteiro E, HepsH V. Diusion of infrastructure: mobilization and improvisation. In: Larsen TJ, Levine L, DeGross JI, editors. Information systems: current issues and future changes. FIP: Laxenburg; 1999. p. 255–74. [34] Gulla JA, Mollan T. Implementing SAP R/3 in a multi-cultural organization. Proceedings of the First International Workshop on Enterprise Management and Resource Planning Systems (EMRPS’99), Venice, 1999. p. 127–34. [35] Hanseth O, Braa K. Technology as traitor: emergent SAP infrastructure in a global organization. International Conference on Information Systems 1998;188–96. [36] Hanseth O, Braa K. Hunting for the treasure at the end of the rainbow. Standardizing corporate IT infrastructure. Computer Supported Cooperative Work (CSCW). The Journal of Collaborative Computing 2001;10(3/4):261–92. [37] Voss C, Tsikriktsis N, Frohlich M. Case research: case research in operations management. International Journal of Operations and Production Management 2002;22(2):195–215. [38] Miles MB, Huberman AM. Qualitative data analysis. Thousands Oaks, CA: Sage Publications; 1984. [39] Ein-Dor P, Segev E, Orgad M. The eect of national culture on IS: implications for international information systems. Journal of Global Information Management 1992;1(1):33–44. [40] Yen HJ, Krumwiede DW, Sheu C. A cross-cultural comparison of top management personality for TQM implementation. Total Quality Management 2002;13(3):335–46. [41] Krumbholz M, Maiden NA. How culture might impact on the implementation of enterprise resource planning packages. CAiSE 2000; 279–93. [42] Vernon M. ERP endangered species? Computer Weekly 1999; 32–5.

371

[43] Davenport TH. Mission critical: realizing the promise of enterprise systems. Boston, MA: Harvard Business School Press; 2000. [44] Jarvenpaa SL. Planning for gold in information systems research: “second-hand” data. In: Nissen H, Klein H, Hirschheim R, editors. Information systems research: contemporary approaches and emergent traditions. New York: North-Holland; 1991. [45] Hinds P, Vogel R, Clarke-Steen L. The possibilities and pitfalls of doing a secondary analysis of a qualitative data set. Qualitative Health Research 1997;7(3):408–24. [46] Larsson R. Case survey methodology: quantitative analysis of patterns across case studies. Academy of Management Journal 1993;36(6):1515–46. [47] Lewis MW. Iterative triangulation: a theory development process using existing case studies. Journal of Operations Management 1998;16:455–69. [48] Sarker S, Lee AS. Using a positivist case research methodology to test three competing theories-in-use of business process redesign. Journal of Association for Information Systems 2002;2(7):1–72. [49] Cavaye ALM. Case study research: a multi-faced research approach for IS. Information Systems Journal 1996;6:227–42. [50] Orlikowski WJ, Baroudi JJ. Studying Information Technology in Organizations: Research Approaches and Assumptions. Information Systems Research 1991;2(1):1–28. [51] Meredith J. Building operations management theory through case and 9eld research. Journal of Operations Management 1998;16:441–54. [52] Barnes D. Research methods for the empirical investigation of the process of formation of operations strategy. International Journal of Operations and Production Management 2001;21(8):1076–95. [53] Ellram L. The use of the case study method in logistics research. Journal of Business Logistics 1996;17(2): 93–138. [54] Klassen RD, Whybark DC. Barriers to the management of international operations. Journal of Operations Management 1994;11:385–96. [55] Akkermans H, Bogerd P, Vos B. Virtuous and vicious cycles on the road towards international supply chain management. International Journal of Operations and Production Management 1999;19(5/6):565–81.