This article originally appeared on the BeyeNETWORK.
A few years ago, a popular TV program featured a detective who always told witnesses, "Just the facts, ma'am." One would get the impression that the facts are important. In fact, it can be stated that basing conclusions on nonfactual information is a fool’s game.
It was with this mindset that I was chatting with some government information professionals recently. They told me that "data warehouse" is a dirty term in the government world. I was told, "Whatever you do, don’t mention data warehouses because they have been such failures in government circles."
With this warning, I did some investigation. I was told that a certain justice department had had a spectacular failure with their data warehouse. Their contractor – a well known systems integrator – had used the classical systems development life cycle to build the data warehouse. And the result was a gross failure.
Now here’s the funny part. Every data treatise that I have ever written or have read by others who have built data warehouses explicitly and in no uncertain terms states: When you build a data warehouse, DO NOT USE the classical waterfall development methodology. There are warnings everywhere. DO NOT USE THE CLASSICAL SDLC.
However, the contractor used the SDLC and the development results were a failure. And who got blamed? Why, data warehousing, of course.
Then I reviewed a memo from a top government executive. He stated that the data warehouse was a failure because they had tried an information bus and the information bus did not work.
Wait a minute! An information bus IS NOT A DATA WAREHOUSE. In fact, people build information busses because they can’t or won’t build data warehouses. Data warehouse advocates warn against building information busses because they are great in theory but do not work in practice.
Does anyone else see the lunacy of all of this? Am I the only one that even cares to look at the facts?
These are two cases in which people have NOT built a data warehouse, have failed, and then have blamed the failure on data warehousing. What kind of Alice in Wonderland is all of this? Is it me smoking the funny stuff or someone else?
Blaming the failure on the data warehouse in these two cases is like blaming the Boers for 9/11. It is convenient because the Boers are not around to defend themselves. The problem is that the Boers had absolutely nothing to do with 9/11. But if facts and the truth are no problem, then let’s all start picking on the Boers.
Blaming the Boers makes about as much sense as blaming data warehouse for failure when NOT building a data warehouse is the real culprit. If we really care about facts or the truth, then we can start to form our own Alice in Wonderland.