THE IMPLEMENTATION PUZZLE OF CRM SYSTEMS IN KNOWLEDGE BASED ORGANIZATIONS Bendik Bygstad The Norwegian School of Information Technology Hans Burums vei 30 1357 Bekkestua Norway bendik.bygstad@nith.no ABSTRACT Many companies have large expectations to the use of CRM systems, expecting to harvest benefits from dialogue marketing and internal knowledge synergies. How should these systems be implemented? And how easy do the benefits come? The research approach is a longitudinal, 6 years, case study of a company implementing CRM both as a marketing principle and as an information system. The implementation was from the outset regarded as an organizational experiment, and the case is laid out in some detail to provide a somewhat "thick description" of the social setting and actors' behavior. The high failure rate of CRM projects illustrates the gap between our intentions and outcomes. Interpreting a longitudinal case study and the research literature we find two options to improve our practice. From a managerial view we should treat CRM projects as complex challenges, needing tight project control and the application of change management techniques, focusing on the marketing process and data quality. In contrast, we could accept that the mechanisms at work at the micro level are only partly controllable by management techniques, and we would let the infrastructure grow organically. INTRODUCTION Increasingly, companies' ability to implement new IT solutions is of crucial importance for the company's ability to change (Applegate 1999). Both strategically and economically, it is therefore vital that a company has the skills to implement information systems fast. As documented in IS research, implementing information systems into an organization is hard, and often unsuccessful (Markus 1997). This is also the case with CRM systems, and perhaps even more so. While there are well-known and impressive success stories (GMAC 2002, Sprint 2002), failure rates of CRM projects may be as high as 70 % (Tafti 2002). Comparing the large expectations regarding CRM with the actual results in companies, a picture of sobering consideration and sometimes downright disappointment emerges. Why is it so difficult? It is documented that most problems in CRM implementation are not technical. Instead, common problems include organizational change and fluctuation, different views on customer information and changes in the business, for example mergers (Schwartz 2002). Argyris and Schön (1996) describe the gap between intent and realization, between what our strategists have recommended us to do and what we are actually capable of doing, as a major challenge for strategic management theory. The core of the challenge is not related to the strategic analysis itself, but to the level of real-time microactions - i.e. all the small decisions managers and employees take during the implementation of change and creation of knowledge. Examples may include problem solving, communication with employees, inter-departmental conflicts and single customer relationships; situations where the battle of change and learning really happens. We suggest that studying these microactions, and the defensive routines at work in an organization, is relevant for understanding the implementation puzzle of CRM. This paper tells the story of a 6 year 1