Title
Effectively Managing Information Systems Architecture
Author
Wai Fong Boh, Graduate School of Industrial Administration Carnegie Mellon University, Daniel Yellin, IBM Research, Bob Dill, IBM Application Innovation Services, and James Herbsleb, School of Computer Science Carnegie Mellon University
Date
3/14/2008
(Original Publish Date: 6/1/2003)
(Original Publish Date: 6/1/2003)
Abstract
Most standards research focus on standardization of Information and Communication Technology (ICT) standards across an industry. Prior research has not focused on standards management issues within organizations. It is important for research on ICT standards to consider the issue of how organizations should effectively manage their internal standards. Internal ICT standards are manifested as information systems (IS) architecture standards and frameworks. We differentiate between infrastructure architecture standards and integration architecture standards. We argue that it is important to differentiate between these two types of architecture standards, because of the differences in the focus, scope, and benefits of infrastructure and integration architecture. We make use of the information processing theory to make hypotheses about how the structure and organization of the architecture team and inter-unit coordination and control mechanisms are expected to differ for effective management of integration and infrastructure architecture. For infrastructure architecture standards, the goals and benefits are obvious to the IT department, but not necessarily so for the individual business units. While using the standards will provide long-term benefits to the organization as a whole, each business unit may not observe a direct benefit from using the standards in the short term. We hypothesize that to effectively manage infrastructure architecture standards, projects should be IT driven, architecture teams should be managed centrally, and the necessary inter-unit coordination and control processes should be in place to govern the interactions of architects and IT operations personnel. On the other hand, integration architecture standards provide business-focused benefits, but are more costly, more complex to manage and require more business involvement. We hypothesize that to effectively manage integration architecture standards, projects should be driven by the business goals of one or more lines of business, architecture teams should have a matrix structure, and the necessary coordination mechanisms should be in place to govern the interactions of architects, IT development personnel and line management. In both cases, organizations should ensure that their architects have the necessary experience working on projects that their architecture standards have an impact on. To test our hypotheses, we are conducting a multi-method study that includes both the qualitative case-study method and the quantitative survey method.
Link