Nowhere is the growth of the services economy more apparent than at IBM, where IBM Global Services (IGS) has grown to be roughly half of IBM's overall business in little more than a decade. With the growth of IGS, we now have the opportunity to direct research toward IBM's consulting business. We see three broad areas in which research can contribute to IGS:

1. On-demand innovation services, which means matching existing researchers and their areas of expertise to consulting engagements as appropriate.
2. Design and implementation of advanced technology targeted at specific internal IGS systems, intended to make workers more effective and processes more robust.
3. Basic and applied research of people-intensive and information-intensive systems.

The third focus area represents the biggest change from IBM's existing research program, but the payoff for developing deeper understandings of consulting work and organizational practices is potentially enormous. Much of the research in organizational behavior and change, work practice studies, and practice-based technology design is done at universities in psychology, communication, sociology, management science, and business departments. We have an opportunity to contribute to this research at IBM, focusing scientific attention on the human, organizational, and technology problems of business consulting. Our goal is to leverage university research and partner with university researchers as appropriate to advance our understanding of consulting businesses. By focusing our research efforts on IBM's consulting business, we have the opportunity to address the large-scale, people-intensive and information-intensive problems of these increasingly important organizations. Our efforts will be devoted to understanding the problems of these organizations from the perspectives of people, practices, communication, and technology so that we provide new opportunities to generate revenue, save money, and improve workers' lives.

For instance, support and maintenance of large middleware software systems is a difficult job, as such systems are inherently complex and contain many interdependent components. Administrators who support these kinds of middleware systems as part of strategic outsourcing engagements run by IGS have developed various problem-diagnosis and collaborative practices to help them do their jobs effectively. Yet there has been little research on the human-computer interaction side of system administration aimed at studying administrator practices and at developing interaction techniques to support problem-diagnosis and collaboration in this context. We have begun to study the work of middleware administrators in the field. So far we have spent four weeks observing web and database administrators in several of IGS's service delivery centers in the US. We found that administrators have developed many ad hoc practices and strategies to help them do their jobs effectively. Based on our studies, collaboration among administrators with various specializations seems to be a key method to compensating for the complexity, scale, risk, and heterogeneity of large-scale, enterprise computing systems.