Building data warehouses and their associated business intelligence (BI) applications requires integrating data from multiple sources—often using multiple formats—to create a platform for running analytical applications. Data warehouse implementations generally involve the full range of new hardware, including servers, storage, and a raft of software from databases to data extraction, transformation, and loading (ETL) tools to end-user applications. Because a data warehouse mirrors the company's entire information infrastructure, creating one is a multifaceted challenge.
Given the complexity of data warehouse projects, to whom do IT professionals turn for advice? According to an exclusive analysis of data from Survey.com ( http://www.survey.com ), project teams often turn first to business consultants when starting a data warehouse/BI project. In the long run, however, the vendors of the database management system (DBMS), software applications, and tools make the most important contributions to the project. Why? Because the vendors provide the services IT groups need to successfully reach their goals.
Graph 1 shows the top sources of assistance to enterprises initiating data warehouse/BI projects. That folks seek the advice of business consultants first isn't surprising, because successful data warehouse/BI projects involve much more than technology. Project champions must create a business case for the effort and, theoretically, demonstrate a healthy return on investment (ROI). Then the project team must cajole many of the organization's constituencies to buy into the project. For instance, line-of-business managers must help develop the specifications for the project. Enterprise processes often undergo close scrutiny and reworking. Business consultants serve as invaluable guides through those thickets.
Ironically, however, enterprises don't rank business consultants as making the most valuable contributions to successful data warehouse/BI projects. Graph 2 shows the percentage of respondents who rated vendor contributions as somewhat or very important. From this angle, application, tool, and DBMS vendors comprise the key outside players in a data warehouse/BI project. The ETL and data-integration tools directly address the complexities of setting up a data warehouse, and enterprise professionals feel they need these vendors' services.
Graph 3 reveals the importance of different services that vendors offer to data warehouse/BI projects. Vendors' technical services take five of the seven top spots, suggesting that project teams focus on overcoming technical hurdles.
And that technology focus might be why so many data warehouse projects fail. As reported earlier (Research UPDATE, March 14, 2001), according to Survey.com data, only one-third of data warehouse projects meet expectations, and only 13 percent exceed those expectations. Some industry pundits place the failure rate as high as 85 percent.
Data warehouse represents the intersection of IT and business processes. If implementation doesn't match specific business needs, the project won't be successful.