Evolving an Organizational Information Infrastructure:
Transitions to Client-Server in the Transportation Industry

A thesis presented by
Espen Andersen

in partial completion of the requirements
for the degree of
Doctor of Business Administration

Harvard University
Graduate School of Business Administration
George F. Baker Foundation

Abstract

This thesis is a longitudinal case study of IS management in two large and technologically advanced companies in the U.S. transportation industry, during a period where both organizations undertook large projects that together amounted to a transition from a centralized to a distributed IT architecture. The resulting IT architectures were centered around three components: A centrally managed, standardized set of intelligent workstations, a core of mainframe-oriented operational software, and decision support oriented data repositories on parallel processing servers. The study identifies, describes and analyzes the reasons for, experiences through, and outcomes of these transitions.

The findings indicate that the management of distributed information technology can be seen as the management of infrastructure evolution, where the challenge to IS management is to maintain a technologically up-to-date, yet standardized, widely available information processing platform, where the use of the platform is determined by the individual user or small organization unit. For the IS organization, managing the delivery, diffusion and maintenance process of the technology is as important as picking the right technology in itself. A model of this process is developed, seeing management of distributed information technology as a process of repeated decisions, where the strategy implementor must balance five influencing factors (prior technology experience, corporate strategic direction, technology evolution, organizational context and industrial context) in the choice of technology, organizational structures, and specific applications of the technology.

The IS organizations in both companies underwent significant changes during the period the distributed client-server technology was deployed. In both companies, the IS organization placed more emphasis on customer interaction, setting up specific organizational structures to act as internal consultants and technology transfer agents.

The study identifies several important issues in managing distributed technology. The first is that technology commonality and coverage seems to be more important for the success of the project than the features of the technology per se -- in other words, better a mediocre technology which everyone has, rather than a superb technology in the hands of a smaller portion on the company. Second, a key to success is the management of perceptions, achieved through careful creation and selection of successful examples of technology use, explained and made available by the IS management for independent replication and implementation by the rest of the organization. Third, a key problem in managing the evolution and distribution of distributed (as opposed to centralized) information technology is the inability to dynamically allocate computing capability and functionality in a distributed architecture.

This page at http://www.espen.com/papers/thesabst.htm.
Contact information at http://www.espen.com/contact.html.