Labor is by far the largest cost in any business intelligence project: nearly 60% according to some estimates. But the actual "keyboard time" to configure a system is minimal. All the work goes into deciding what that configuration should be.
This work is the time spent understanding user needs and available data, designing solutions that combine the two, testing and reviewing results with users, and then repeating the process until the users are satisfied with the results. Developers have long recognized that the key to keeping down project costs is minimizing the changes made after delivering the initial result. Modifying, retesting, and re-reviewing a fully elaborated existing system can be much more effort than building it the first time.
Yet changes are inevitable. Users and analysts often don't understand their true requirements until they see the first concrete results of the project. Moreover, those requirements will change over time in response to new information - some provided by the BI system itself - and to changes in the underlying business. In other words: required changes pile up at a steady or accelerating rate, while the effort to make each change grows with system complexity. The inevitable result is a growing backlog of unmet business needs. That build-up is the BI Log Jam. Read this analyst report to learn how to break the cycle.
SQL Server Research Library
