Any Information Technology manager lives in fear that some leading, widely-read business periodical will publish an article touting the amazing benefits of a technologically-based business process, such as Data Warehousing. Some senior executive inevitably reads it, walks down to the IT Department and announces, "Let's buy one of these!" It is the latest (though probably not the final) chapter in the search for the "silver bullet" - that long-awaited, eagerly-anticipated "killer application" which will finally deliver the promised return on the IT investment.

Okay, so now you have "bought" a Data Warehouse. The votes are coming in, and the results are not encouraging:

"Response time is too slow!"
"I can produce nice reports, but I still don't have any real insight into my business."
"If we had bought the right technology, we would be getting better results!"
"We spent $2 million on this project! Where's my ROI?"
What happened?

You see, the problem is that you really can't just "buy one of these" Data Warehouses.
While you may indeed be able to buy the technology components, you can't just buy their integration with the business strategy, the culture, or the organization. Successful implementation of a Data Warehouse requires, among other things, a significant investment of time and energy on the part of many, maybe most, of those who will be its users, to ensure that the end result meets their needs. It is a tedious, difficult, lengthy, and frequently contentious process. What's more, as this article explains, much of this time and energy come from the most senior executive levels of the organization - the very people who are inclined to decree, "Let's buy one of these!"

Well, even though you can't just "buy one of these" Data Warehouses, maybe we have reached the point where technology, business demand, frustration, and process insight have converged on something that really can deliver the long-promised return.

There are a growing number of successful examples which prove not only that it can be done, but also that the sometimes dramatic results are more than worth the investment.

The purpose of this article is to further define these reasons for failure, and to provide some ways by which an organization can avoid these pitfalls as it implements its own Data Warehouse. Some will argue that this list is incomplete - that there are other essential points of equal importance.

The fact is that many books have been written, explaining all the problems and pitfalls of Data Warehousing. It would be sheer folly to represent the thoughts that can reasonably be captured in a brief magazine article as comprehensive treatment of this subject. Still, if an organization can avoid the three critical failures mentioned above, it will most likely be delighted with the outcome. Furthermore, it will have put a process in place that allows the "journey" to continue beyond the design and implementation phase.

Driven by technology instead of business need...

When the edict comes from on high to go "buy one of these," most often the technologists assume control of the project, and the focus becomes putting the technology in place. The Data Warehouse project becomes technology-driven. The problem is that an enterprise Data Warehouse is not merely a technology problem. There are several things that must be done to avoid this critical failure.