IT Should Build BI Systems, Not Justify Them

“IT Struggles to Show BI Value” reads the headline in this week’s Computerworld. The article itself says pretty much the same thing.

This spawns three quick thoughts:

- It’s not IT’s job to show the value of Business Intelligence. It’s the job of the users. If IT is pushing BI on users who don’t want it, then something is wrong.

- Even though BI is an indirect expense, users can and should still justify it based on improving customer value. I wrote about this in detail here back in September. Basically, BI can directly improve customer experiences or indirectly improve business economics.

- No, I don’t really think you can measure the precise impact of each decision on customer value. This points back to yesterday’s comment about the value of Mini Cooper’s personalized billboards. You can’t measure that precisely either. But you can at least describe in general how it would translate into increased customer value and make a back-of-the-envelope calculation to see whether there is any hope of recouping the investment. (In Mini Cooper’s case, the expected value was more purchases and referrals from more enthusiastic owners.) You can also assess whether alternative investments are likely to be more effective: if it cost the same, would it be better to have the personalized billboards, send every Mini Cooper owner a personalized Mini Cooper jacket, or use the RFID to pay their highway tolls for a week? Beats the heck out of me, but if you accept something like Net Promoter Score as a reasonable way to measure customer enthusiasm, this is now something you can test, not just a bunch of bright ideas to pick from randomly.