Just the other day, I was in a discussion talking about why BI projects fail. We have all heard of stories about why it fails, be it scope, requirements, design, development, technology...etc. Then, there were comments that were made that really made me think. Comments such as "it's information delivery...it's performance measurement...it's not "rocket science". Maybe BI is "rocket science" because there are so many of these initiatives that fail.

Over the past few months, I've been involved in turning around failed attempts at BI at various companies. It is pretty apparent that poor scope and requirements documentation are a major cause for a failed BI project. This then leads to faulty designs that not only have poor performance, but they do not answer any of the questions the business is asking. More distressing, is the fact that the users can still get different measures even though everyone is using the same data.

A lot of people talk about being able to build BI solutions (i.s. data warehouse, cubes, reports and analytics). I hear and read "simply connect to your data source and build your cubes and then build your reports". Sure, sounds pretty simple. Anyone can do it, but can they?

There is no question that BI can drive business performance, be it reducing costs, process efficiencies or growing revenue. The problem is building a BI solution that helps drive business performance. A lot of people can claim to build BI solutions but very few can do it properly and successfully.

Is BI "rocket science"? Maybe, maybe not. But when i hear of all the horror stories and assess some of the BI implementations, I wonder...maybe I'll be in my garage working on my rocket.