Of course, creating cross-functional strategy and planning teams is important for any enterprise IT project, says John Hagerty, an analyst at AMR Research. But it's especially true for enterprise data warehouse and business intelligence projects because their success depends on broad user support and because consequential business decisions are made on the faith that information is accurate.

When it came time to deploy the BI tools at Rensselaer, that top-down and cross-functional support was crucial. For example, Jackson made clear that she only wanted to see numbers that came from the data warehouse. Strict data governance was enabled through multi-committee support. And creating new processes for data reporting—such as how to divvy financial credit in multi-disciplinary research efforts—was aided by the cross-functional relationships and understanding that had been built up during the development phase.

Think big, start small, deliver quickly.

Once Rensselaer decided to deploy BI, the first six months of work focused on laying a strong foundation. The grand vision for the project assumed that eventually, all business data would be filtered using the BI tools. And so, the university developed an overarching data policy and procedures that could be used by any groups created to define, cleanse and manage information on an ongoing basis. At the same time, Kolb's group created a systems architecture based on an Oracle data warehouse, Informatica's PowerCenter data integration platform and Hyperion business intelligence technology.

For its first set of reports, the university chose financial information, unveiling a data mart (a collection of data about a specific subject) and reporting tools in November 2002. "We wanted to have a success at a very fundamental level," says Kolb, "and finance touches everything." That pervasiveness is not just in terms of the data itself—everybody has a budget—it's also about the people involved. The finance group, like IT, works with everyone. And its success with the application made the group a powerful advocate. "Finance became a huge partner with us," says Kolb.

Hagerty of AMR Research says focusing on quick gains is a key to success in any BI project. "Start small, deliver value and get people bought into the value along the way," he says.

This first project exposed a lot of dirty data contained within the ERP system, which provided a powerful reality check for users. Those mistakes—say a missing zero—may have originated with an inattentive employee. With the improved reporting system, the finance manager was able to suss out those mistakes more quickly. As a result, finance became a vocal advocate of clean data, and helped to enforce the new enterprise data policies.