This article is an excerpt from the book Data Strategy (Addison-Wesley, 2004) by Sid Adelman, Larissa T. Moss and Majid Abai. If you ask ten knowledgeable people for their definition of integration, you are likely to get ten different answers. Most of the answers will be an interpretation of Webster’s dictionary definition for integration, which is “to make whole or complete by adding or bringing together parts.” This could mean anything from two systems passing data back and forth (loosely coupled) to a shared data environment where all data elements are unique and non-redundant and are being reused by multiple applications (tightly coupled).

In reading many articles about the various data governance initiatives that have been started by organizations of all sizes and across all industries, one may sense that all such programs are immediately effective. Unfortunately, this is not the case, as Gartner predicts that less than 10% of initial data governance programs succeed (Gartner, 2008). The reasons for these failures are as varied as the programs they represent, but there are some themes that can be identified to serve as warnings for new or existing programs.

Data warehouses, marts and all analytical solutions are often thought of as technical structures that are defined, built, and managed by the IT department. By now I think we all know that the solutions must be business driven, having a real business problem to solve, and IT needs to have some documented requirements of what the business wants to do with the data.