Data Warehousing Mistakes – 6 Things Not To Do

Data warehousing pros avoid these pitfalls.

David Haertzen.

You can improve your odds of bringing in a successful data warehousing project by avoiding critical mistakes that others have made.  In this article, you will learn what these six critical mistakes are and how to avoid them.

First, create a data warehouse without business input and expect that business users will embrace and use that system.  This “build it and they will come” approach is usually doomed to failure.  There are two reasons why this is a problem.  For one thing, a system developed without business input is unlikely to meet business requirements.  For another thing, business people who are left out will be unlikely to support a system that is pushed at them without their involvement.

Second, save the most difficult part of the project until the end.  Obtaining the right data is often the most difficult part of the data warehousing and business intelligence project.  If you wait until the end to see what the outputs look like, you may get the nasty surprise that data is not what the business is seeking.  Instead, work to explore data and show parts of the data to business people as early in the project as possible.  Then any problems that are discovered can be more readily corrected.

Third, start without basic pieces in place.  If you start without being ready you run the risk of failure and loss of credibility.  Basic pieces that should be in place include: business sponsor, analytics champions, a specific business goals, and basic technical architecture.

Fourth, wait until conditions are perfect.  It is difficult to find perfect conditions.  It is OK to start in a focused area with basic technology.  Success in one area can be a stepping stone to further success.  If you wait for all conditions to be perfect, you run the risk that your project may never start.

Fifth, set unrealistic expectations with executives.  If you tell executives that the new system will solve all problems you are running the risk that executives will be disappointed when they learn otherwise.

Sixth and finally, antagonize executives by implying that they are making poor decisions.  If you sell the project by saying that “executives will now be able to make informed decisions”, the executive supporters that you need may think that you are saying that they are uniformed decision makers and then feel antagonistic toward your project.

In conclusion, we have reviewed some basic mistakes that can harm your data warehousing project.  Armed with this information, you can improve your odds of a successful project.