Communications of the ACM
Home/Magazine Archive/September 2007 (Vol. 50, No. 9)/What Matters When Introducing New Information Technology/Full Text
Communications of the ACM
What Matters When Introducing New Information Technology
Despite more than 30 years of experience, organizations still face an uphill battle implementing information systems. A 2002 KPMG study reported that 56% of firms had to write off at least one IS project as a failure during the previous year. A 2003 Hackett Group study reported that 30% of IS projects fail. Further, according to KPMG, the average cost of failed implementations in 2002 was roughly $15 million, a significant increase from the $4.2 million average per failure in 2000 [9]. Overall, this paints an alarming picture of how much money U.S. businesses lose on failed systems and technology implementations each year.
In addition to the direct monetary costs of a failed implementation, indirect missed opportunity costs are also associated with failed or delayed implementations. Examples abound of organizations (such as Digital Equipment Corp., K-Mart, and Merrill Lynch) that missed technology-based opportunities only to find themselves playing catch-up. Decades of research has identified the issues that are essential for the successful implementation of IS, yet delays and failures continue to be reported. We examine the process of implementation here in order to determine which issues matter most during each stage of the process.
Top management support is essential for implementation success, regardless of the type of system.
IS implementation has long been an area of interest for both researchers and practitioners, with much emphasis on identifying the factors that lead to more and to less successful outcomes. In a review of the literature 1 we identified five broad factorscommitment, knowledge, communication, planning, and infrastructurerepeatedly associated with implementation success and failure. All are reasonably well known and accepted in academic and practitioner circles (see Table 1).
Whether these factors are considered individually and/or in concert, there seems to be several ways for a project to fail. Yet if we know they are so critical for a successful, on-time implementation, why aren't more projects successful and on time? Is there more to the process than simply emphasizing them? Are they equally important throughout the process? Are some more important than others? Does their importance depend on the phase of the implementation? Answering requires a longitudinal view of the implementation process. Prior research (such as [4]) indicates that different activities occur at different stages of the process. We expect that the importance of each of the five factors will change because the activities differ over the stages of implementation.
Six stages of IS implementation were identified in prior research [4]: initiation, adoption, adaptation, acceptance, routinization, and infusion. Activities during the initiation stage include looking at organizational problems and opportunities to see if there is an IS solution for them; by the end of this stage, a match between an IS solution and an organizational problem or opportunity has been identified. During the adoption stage, organizational backing for the IS solution is sought so the appropriate resources are allocated to implementing the IS identified during the initiation stage. During adaptation, the organization develops and installs the new system and modifies the organizational routines affected by it to leverage the system's benefits. At the end of this stage, the system is ready and available for use. In the acceptance stage, users begin to employ the system in their work. Routinization involves encouraging them to make use of the system as part of their daily activities. Finally, during infusion, the organization realizes increased effectiveness from the systemassuming all prior phases were successful.
It is important to note that the process typically unfolds in an iterative, rather than linear, manner. For example, it is likely that after matching a system to an organizational problem (initiation) and getting financial backing (adoption), a problem arises during development and installation that sends the process back to initiation in search of a new technology solution.
We propose that attending to the right issues at the right time improves implementation in terms of both process and outcome. To help understand the importance of each implementation issue, we conducted a case study at the headquarters of a large, multinational, multidivisional corporation that examined five recently completed IS implementations. The organization, characterized by extensive corporate and IS planning, has a corporate IS division and IS liaison personnel in its operating divisions. The liaison personnel are responsible for ongoing communications with the operating division and corporate IS. The projects were analyzed using a business process reengineering (BPR) methodology that first mapped the overall processes, then identified the disconnects, or breakdowns, in them. For the organization, the analysis aimed at understanding why the implementation of these systems had taken so long.
The five projects selected for analysis were chosen because of their diversity in terms of technology and the divisions involved. All had been completed prior to our analysis and (to varying degrees) were ultimately judged to be successful by the users. Yet, despite their successful outcomes, each project experienced significant delays during its implementation process.
To collect data regarding the projects, the organization formed a steering committee of 12 people. Each of the five projects we examined had two or three different steering committee members assigned to it. These people led the technology evaluation team and conducted, on average, seven interviews and two focus groups consisting of approximately 10 employees. Due to some overlap across projects, 52 employees were involved in the overall data-collection effort. They included corporate IS managers and staff involved in the projects, as well as the managers and other personnel from the operating divisions implementing the new technologies and selected end users.
For each technology, the evaluation team relied on retrospective analysis by the interviewees. The interviews and focus groups were conducted over a period of eight weeks. Individual interviews took about 60 minutes; focus groups typically took 120 minutes. The result was the creation of five individual process maps describing the complex temporal pattern of events leading to the implementation of each of the systems, as well as to identification of the disconnects (such as missing connections, failed hand-offs, and unnecessary exchanges) that occurred during each implementation process. For each project, the verification and final evaluation of the process maps and associated disconnects was performed by the original project participants, regardless of their role in the BPR project.
A total of 185 disconnects was identified across the five projects, with the number per project ranging from 36 to 39, for a fairly even distribution across projects. Further, the number of disconnects within each stage ranged from 43 to 49, also representing a fairly even distribution. Once the disconnects were identified, they were mapped onto the implementation process. We then translated the organization's implementation process maps to coincide with the stages discussed in [4].
This mapping enabled us to code each disconnect based on the five key implementation issues and the six process stages. For the purpose of the study, we considered only initiation, adoption, adaptation, and acceptance; the other two stagesroutinization and infusionoccurred outside the scope of the study. Table 2 includes sample disconnects associated with each issue; Figure 1 is a graphical representation of the relative importance of each factor by stage.
What Matters Most?
Overall, the results supported our initial expectationsthat while all of the typically identified issues are important for project success, all are not equally important at every stage of a project's development. A number of themes emerged from our analysis of the implementation process disconnects: First, as outlined in both figures, there is a definite pattern of ebbs and flows in terms of issue importance. Table 2 identifies communication and commitment as the two most frequently cited issues overall. Figure 2 reflects the finding that knowledge is most important during initiation, communication most important during adoption, and infrastructure most important during adaptation. These three issues even out in importance during acceptance, with commitment gaining in importance throughout the entire process. It is not surprising that commitment would emerge as one of the two most frequently cited factors; it is among the most cited in the literature and is at the heart of implementation success and failure. Research indicates that top management support is essential for implementation success, regardless of the type of system. The results of this study are in line with prior research, but also add an interesting twistthat the importance of commitment increases over time.
Since the process maps for the organization end at initiation, extrapolation is necessary to being able to discuss infusion and routinization. Extending the current pattern, infrastructure would likely continue its downward trend, as the technology would be substantially in place at this point. Knowledge would likely continue its slight upward trend, though the emphasis here would shift from learning about the technology to learning about how to use the technology. Likewise, we expect that commitment would continue its upward trend as the technology is infused, dropping off significantly once it becomes part of the everyday routine, or routinization. Communication will also likely continue its upward trend and be important in infusing the technology in the organization.
What should organizations do to be successful? These findings lead to a number of suggestions for how organizations can succeed at each stage of the implementation process (see Table 3). The first is that they must emphasize different activities at each stage of the project's life cycle. While each is important at every stage, some are more important within a particular stage. It is entirely possible for an organization to excel at any one of the issues listed in Table 3, but unless they are at least competent in each of them, implementations can run over budget or fail. Alternatively, organizations could excel at several of the issues and still not complete a project on time and on budget because they emphasize the wrong thing at the wrong time.
Organizations able to adjust how they conduct their employees and skills to support IS projects are best able to enhance their performance.
The initiation process relies on overcoming knowledge disconnects to find technology solutions to business problems. A central knowledge base is critical to helping employees identify and recognize the value of new knowledge. Absorptive capacity [3] is predicated on the need for existing stores of knowledge to facilitate assimilation of new knowledge. Managers with a good grasp of technology are better equipped to identify opportunities offered by new technology. Similarly, members of the IT organization with a good understanding of the business are better able to identify and implement technologies that assist the business. This suggests that, even during times of financial trouble, organizations should invest in educating their employees, including sending them to classes, trade fairs, and conferences. Lacking such knowledge propagation, organizations may find themselves stuck in initiation, unable to identify the new technologies best able to support their business strategies and long-term interests.
During the adoption process, communication is critical. This is where user-IS forums are particularly useful. Their goal is to let users talk about their business needs and IS personnel about systems in order to facilitate knowledge sharing. Another technique is IS-liaison positions in which IS employees are physically placed in business units. This technique helps them develop domain knowledge that then facilitates communication among IS and business employees [10]. Organizations unable to exploit the potential of new technologies may be stuck here.
In adaptation, it appears that limitations or concerns with an existing technology infrastructure "catch up" with the new technology solution. Given that adaptation is the point at which the solution is physically installed, it is not surprising that the importance of infrastructure issues increases during this stage. It is important to evaluate the infrastructure early in the technology-introduction process to head off this problem. In addition, the organizationwide architecture must be clearly articulated and communicated throughout the organization, as communication of the standards could head off the problem of incompatible systems at an early stage. Organizations unable to adapt new technologies to their needs may be stuck here.
As the implementation process continues, the importance of commitment increases. The momentum associated with the project kick-off must be maintained throughout the project's life cycle. Clear signals must be sent by management, in words and deed, that the project is important and will succeed. These signals must increase in magnitude as the project continues. Thus, continuing initial project excitement through acceptance is particularly beneficial.
Planning is also significant throughout the process, with ebbs and flows of its own. Planning issues are more significant during initiation when they fail to provide a roadmap for a particular technology solution or the plan is so rigid it hinders the organization's ability to evaluate alternative technologies. The importance of planning increases again during adaptation. Current technology architectures or plans are tightly connected to the existing technology infrastructure. Issues involving the technology infrastructure reflect on the IS plan. Thus, while a coherent plan is necessary, a certain amount of flexibility must be built into the architecture, the infrastructure, and the planning process to facilitate the integration of new technologies.
This research paints a slightly more complicated picture of what organizations must do to ensure the success of their IS implementations. Rather than suggest that there are simply five things organizations must do to succeed, we suggest they aim for balance and harmony among these factors in order to maximize the odds of success. Telling all the musicians in an orchestra to play at top volume for the entire musical score is unlikely to create beautiful music. An orchestra must include not only various instruments and skilled players, its conductor must know when to have each instrument enter and exit the score to produce the desired music. Organizations able to adjust how they conduct their employees and skills to support IS projects are best able to enhance their performance.
Conclusion
By examining the factor and process elements across implementation stages, we can begin to understand why projects with "successful" outcomes still suffer lengthy delays and cost overruns. Additionally, we can begin to structure interventions that address the issues associated with delays. In light of today's inevitable cost sensitivity, it is especially important for IT implementations to stick to budgets and deadlines. The findings of this study provide insights and recommendations for organizations stalled in one or another of the stages.
References
1. Basselier, G., Benbasat, I., and Reich, B. The influence of business managers' IT competence. Information Systems Research 14, 4 (Dec. 2003), 317336.
2. Boynton, A., Zmud, R., and Jacobs, G. The influence of IT management practice on IT use in large organizations. MIS Quarterly 18, 3 (Sept. 1994), 299318.
3. Cohen, W. and Levinthal, D. Absorptive capacity: A new perspective on learning and innovation. Administrative Sciences Quarterly 35, 1 (Mar. 1990), 128152.
4. Cooper, R. and Zmud, R. Information technology implementation research: A technological diffusion approach. Management Science 36, 2 (Feb. 1990), 123139.
5. Ginzberg, M. Early diagnosis of MIS implementation failure: Promising results and unanswered questions. Management Science 27, 4 (1981), 459479.
6. Larsen, K. A taxonomy of antecedents of information systems success: Variable analysis studies. Journal of Management Information Systems 20, 2 (Fall 2003), 169246.
7. Lederer, A. and Sethi, V. Critical dimensions of strategic information systems planning. Decision Sciences 2, 1 (Winter 1991), 104119.
8. Massey, A., Montoya-Weiss, M., and Brown, S. Reaping the benefits of innovative IT: The long and winding road. IEEE Transactions on Engineering Management 48, 3 (Aug. 2001), 348357.
9. Nash, K. Companies don't learn from previous IT snafus. Computerworld 16, 21 (Oct. 30, 2000), 3233.
10. Reich, B. and Benbasat, I. Factors that influence the social dimension of alignment between business and information technology objectives. MIS Quarterly 24, 1 (Mar. 2000), 81113.
11. Ross, J., Beath, C., and Goodhue, D. Develop long-term competitiveness through IT assets. Sloan Management Review 38, 1 (Fall 1996), 3142.
12. Weill, P. and Broadbent, M. Leveraging the New Infrastructure. Harvard Business School Press, Boston, MA, 1998.
Footnotes
1A complete list of the articles used to generate this list is available at www.u.arizona.edu/~suebrown/What_Matters_When.htm.
Figures
Figure 1. Importance of implementation issues across stages.
Figure 2. Relative importance of implementation issues across implementation process stages.
Tables
Table 1. Implementation issues, definitions, and sources.
Table 2. Sample disconnects and numbers of times identified.
Table 3. Implementation process problems and recommended solutions.
©2007 ACM 0001-0782/07/0900 $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