Communications of the ACM
Home/Magazine Archive/January 2007 (Vol. 50, No. 1)/Success Factors For Implementing Global Information.../Full Text
Communications of the ACM
Success Factors For Implementing Global Information Systems
Given the importance of these systems, it is surprising that the effectiveness with which they are exploited varies so much among the firms that pay for and use them [5]. Reasons include the diversity of legacy software components, technical and social issues, lack of management support (and champion), and insufficient involvement by users and management in project definition and planning [10]. Approaches proposed for addressing these problems include software evaluation procedures, software design tools (such as XML and CASE), and communicating with software specialists, managers, and users during implementation [4].
Whereas the implementation of an enterprise resource planning system is generally regarded as difficult, the implementation of a system to support multinational company networks (global IS, or GIS) is an even greater challenge. A GIS must interface with intrafirm and interfirm software in several ways:
- Providing an IT backbone in all countries in which the firm and its supply chain partners do business;
- Ensuring the exchange of highly structured information worldwide, even as local customization is allowed in each country;
- Managing a vast amount of information, including border procedures, duties, tariffs, and exchange rates; and
- Providing optimization techniques that facilitate the global logistics network.
The complexity of implementing a GIS increases with the cultural differences in the countries involved, especially when the project manager is unfamiliar with the global context and factors that must be considered during implementation [9]. Further complications arise due to languages, business and legal environments, technology and vendor support, national infrastructure, local market size, data export controls, and the availability of local IT expertise [8]. Simple projects carried out domestically, according to [7], can turn into a nightmare if carried out globally. My aim here is to identify the factors that are most important to the success of implementing a GIS.
No significant difference existed between successful and unsuccessful projects in terms of vendor selection methods or the management level at which final approval for the project was given.
Top Factors
While academic research in the 1990s explored frameworks for undertaking research in international IS no studies directly addressed the success factors for implementing a GIS. Research on the implementation of other large-scale systems, however, may help identify possible GIS success factors (see Table 1). Top management support is the one most commonly cited when implementing complex systems, followed by capable and well-understood business processes, the use of a cross-functional team, and maintaining cross-functional cooperation and communication. Other significant factors are clear project goals and the management of affected employees. Note that employee management also relates to the training of managers and a system's future users.
Given the differences between implementing a local and a global IS, are these factors also important to the success of a GIS implementation? Other potentially important factors may exist, even though they are not discussed in the literature. For example, the complexity and difficulty of organizing supply networks would suggest a need for well-organized and well-understood processes. On the other hand, due to the international scope of the projects, the financial backing, capability of the IT staff, and ease of cross-functional and cross-cultural communication might be of special importance. That is, while the literature review outlined in the table indicates possible success factors, it is likely not exhaustive or necessarily relevant.
In order to begin to address these questions, my colleagues and I conducted a study in 20022003 on success factors for global IT implementations in 16 major IS implementation projects in eight multinational firms in Toronto, Canada. Each firm had sales exceeding U.S.$10 million in 2001, ensuring that firm size and financial resources would not be confounding factors. Eight projects had been successful, according to their managers' own estimates, the eight unsuccessful. Eight exhibited a global scope, and the other other eight covered the introduction of local IS. The structure of this sample, which is similar to that of a designed experiment, allowed me to identify the factors that influence the success vs. failure of global vs. domestic IS projects.
We used a pretested and consequently revised structured questionnaire to conduct the case research. The questions allowed project managers to self-identify the factors they found important to the success or failure of their projects and use an open-question format to explain the projects' background, planning, implementation process, and outcomes. All 16 projects had been completed (or aborted) by the time of the interviews, thus providing valuable hindsight.
I analyzed the data by contrasting the responses about successful projects with those about unsuccessful projects and international with domestic projects (pattern matching). Moreover, I contrasted the frequency of the presence or absence of factors as a rough indicator of their importance. Lastly, I coded the qualitative results and validated them through a nonparametric correlation analysis of the quantified factors.
The findings validated many, though not all, of the factors identified in the literature as making a difference in the success and failure of IS projects. That is, some factors exhibited so little correlation with success and failure that their causal relationship with success could not be validated. On the whole, all firms largely knew what they would be getting into and what would be required to start and implement their projects; as a result, these factors did not make a difference. In particular, most of the eight failed projects did, at least apparently, have the backing of top management. The firms also knew that the projects would require significant financial resources and significantly affect firm operations. No significant difference existed between successful and unsuccessful projects in terms of vendor selection methods or the management level at which its final approval was given.
Several other management factors were also relevant in determining project success (see Table 2). Most important, top management did not perceive as urgent six of the eight unsuccessful projects. Moreover, in four of the unsuccessful projects, top management did not understand the project's intricacies and initially withheld approval when it was proposed. These factors indicate that, even though top management claimed to support the projects, that support did not really exist. That is, in these cases top management did not make the project a top priority. This conjecture was supported by the fact that most of the eight failed projects were considered not disruptive or urgent during implementation, further leading management to conclude that no special attention was needed.
Lack of top management support showed itself in various ways. In one project, management bickered about the financial resources to be devoted to the project, which was approved and begun but only after several rounds of negotiations. The lack of leadership also caused the implementation phase to stretch out until the project was finally reassigned to a different division within the firm. In another project that was also poorly understood and poorly supported by top management, implementation was halted in mid-course due to poor interim results.
Planning and implementing a system must be done with a sense of urgency, but not at the cost of proper planning.
Other factors influencing the success or failure of projects included the involvement of a large number of employees and functional areas from within the firm, along with a number of external partners. These factors indicate the need for system designers to model processes across functions and involve users in system design. Managers of four unsuccessful projects admitted that, in hindsight, they had neglected three critical factors: not using cross-functional teams; not engaging in cross-functional communication; and not bringing end users on board early enough during the project. Consequently, these implementations were painful, with only moderately successful results and low user acceptance. Most were not completed on time or within their planned budgets.
In contrast, many managers of successful projects stressed the importance of proper and detailed planning, investigating the underlying processes, and allocating appropriate human and financial resources. Despite the planning, the managers of successful projects had to address many of the same hurdles as managers of unsuccessful projects, including tight constraints on time and resources, technical glitches, and end-user training. What set them apart was that they had expected the projects would be disruptive to their firms' operations and that difficulty would result. Due to detailed up-front planning, they were able to flexibly respond to the situations and complete the projects successfully. Six of the successful projects also remained largely within their planned timeframes and budgets. These results largely coincide with the success factors most frequently discussed in the literature listed in Table 1.
Project Scope
The factors that led to the success or failure of IS projects did not apply equally to local and global projects. Top management supported (at least superficially), approved, and expected disruptions of their firms' business operations for global projects as much as they did for local ones. However, differences existed in top management's perceived need to involve more people and business functions. Almost all managers recognized that implementing a global IS would be more complex and resource-intensive than implementing a local IS. This is most clearly reflected in the fact that all managers (even of unsuccessful global projects) had, in advance, understood the necessity of involving a large number of people and many functional units in their projects. In contrast, none of the eight unsuccessful local projects involved different functional units, and most did not involve enough people. Hence, while complexity and resources were differentiating factors in the success of local IS projects, they did not, even when recognized, lead to global project success. This finding is underscored by the fact that in only one failed global project was scope a major challenge.
Even when managers recognized the need for cross-functional communication in global projects, implementation often did not occur as planned. Global systems often exchange standardized data while allowing for regional customization. These conflicting needs are driven by a combination of businesswide accounting and control, as well as by local taxation and accounting laws, customs, and languages. These mixed requirements can be met only when functional and regional units collaborate. While the need for user and cross-functional involvement were clear at the outset, two of the unsuccessful global projects (and none of the successful ones) made the lack of departmental communication and having everyone on board at least partly responsible for the lack of success. This result reflects the difficulty and complexity of implementing global IS, even when management knows what must be done. Only detailed and thorough planning helps circumvent this pitfall.
The complexity of global projects was also exemplified by the frequency with which the implementation was expected to disrupt firm operations. For global implementations, three of the successful projects had anticipated substantial disruptions, while three of the unsuccessful ones had not, leaving the unsuccessful ones exposed and unprepared to counteract the resulting disruptions.
These cases also indicated that more top managers understood the technical intricacies and organizational impact of failed local IS projects than of failed global projects. This point reflects the complexity of GIS projects. Complex projects require more time and resources and better preparation and management than their simpler counterparts. The importance of top management's understanding is more essential to the success of GIS projects, as it leads to greater support and better change management throughout implementation. In contrast, local projects may be carried out relatively quickly by small implementation teams, meaning that sustained support by top management is simply less critical.
In one casean ERP system implemented in a firm's European and North American locationsmiddle management was largely responsible for its successful implementation. Despite occasional interference or ignorance by top management, middle managers drove not only the project's implementation but also the education of their superiors. Due to the scope of the project, this process was so frustrating and exhausting that many of these people left the firm after the system went into production. This case reflects the importance of middle management in complex IS projects, a point strongly supported by anecdotal evidence. Top management should thus support and open doors for middle managers responsible for a project's day-to-day management, understand the processes, and interface with IT personnel. This support is clearly more necessary in a GIS than for a less complex local project.
Implications
When carrying out this research there was no presumption of which factors would be important for the success of implementing either global or local systems. Rather, project managers were asked to explain what they found important and identify the issues they had to deal with. Thus, not all factors identified in the literature as important to successful implementations were mentioned by these managers and validated in the context of this research.
The findings did, however, yield a direct message for managers of global projects: Since their requirements are so complex, preparation for their implementation is crucial. Planning and implementing a system must be done with a sense of urgency, but not at the cost of proper planning. In global projects, plenty of difficulties (competing for management attention) arise even when the most thorough planning is in place. If planning is lacking, these difficulties become insurmountable. The project team must also consider how to simplify existing processes. Process reengineering is often simpler than attempting to customize already complex systems to satisfy even more complex processes.
Project managers also needed buy-in from top management, and a sense of urgency helps convince it to devote its attention to the project. Top managers must understand the project and its potential hurdles, as they determine the allocation of personnel and financial resources. When difficulties arise (inevitable in global projects) top management is then best placed to help the project team deal with them effectively.
Finally, a GIScomplex, geographically diverse, with often conflicting central and local requirementsdemands the involvement of many people from many different functional and regional units, including the firm's supply chain partners, vendors, customers, and consultants. Entrusting a project to the headquarters IT department is inadequate. Employees from the firm's various regions, departments, and partners must all have input in the design and implementation process. Unlike local IS implementations, recognizing the need for more resources and cross-departmental involvement alone does not ensure successful GIS implementation. Detailed planning, flexibility during implementation, and competent leadership are much more crucial for implementing a global IS than for implementing a local IS.
Conclusion
Following these straightforward and intuitive guidelines dramatically increases the liklihood of successfully implementing both global and local systems. While some are not new, this research highlights the factors that are most important to a GIS implementation and how the findings differ from others in the literature. It would be interesting to validate these case-based results through a large-scale survey involving a few hundred GIS implementation projects.
References
1. Akkermans, H. and van Helden, K. Vicious and virtuous cycles in ERP implementation: A case study of interrelations between critical success factors. European Journal of Information Systems 11, 1 (Mar. 2002), 3546.
2. Al-Mashari, M., Al-Mudimigh, A. and Zairi, M. Enterprise resource planning: A taxonomy of critical factors. European Journal of Operational Research 146, 2 (Apr. 2003), 352364.
3. Cox, J., Zmund, R., and Clark, S. Auditing an MRP system. Academy of Management Journal 24, 2 (June 1981), 386402.
4. De Brabander, B. and Thiers, G. Successful information system development in relation to situational factors which affect effective communication between MIS users and EDP specialists. Management Science 30, 2 (Feb. 1984), 137155.
5. Kim, B. and Oh, H. An exploratory inquiry into the perceived effectiveness of a global information system. Information Management & Computer Security 8, 3 (Aug. 2000), 144153.
6. Kydd, C. Understanding the information content in MIS management tools. MIS Quarterly 13, 3 (Sept. 1989), 277290.
7. Roche, E. Managing systems development in multinational corporations: Practical lessons from case studies. In The Global Issues of Information Technology Management, R. Zigli, Ed. Idea Group Publishing, Harrisburg, PA, 1992, 630654.
8. Selig, G. A framework for multinational information systems planning. Information & Management 5, 2 (June 1982), 95115.
9. Tractinsky, N. and Jarvenpaa, S. Information systems design decisions in a global versus domestic context. MIS Quarterly 19, 4 (Dec. 1995), 507534.
10. Umble, E., Haft, R., and Umble, M. Enterprise resource planning: Implementation procedures and critical success factors. European Journal of Operational Research 146, 2 (Apr. 2003), 241257.
11. White, E., Anderson, J., Schroeder, R., and Tupy, S. A study of the MRP implementation process. Journal of Operations Management 2, 3 (May 1982), 145153.
12. Wixom, B. and Watson, H. An empirical investigation of the factors affecting data warehousing success. MIS Quarterly 25, 1 (Mar. 2001), 1741.
©2007 ACM 0001-0782/07/0100 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2007 ACM, Inc.
No entries found