How to Increase the Reliability of Your IT Infrastructure Using Predictive Analytics REGISTER >
Even where we have been able to address these issues, more challenges appear. While our summarized and filtered data is often just what we need to assist in making a decision, it isn't always enough to drive immediate understanding. We may need to "drill down" to the original information to gain a full understanding of the situation, and hence make the best decision.
This is easy in theory with today's BI tools--the dashboards that present summarized information connect directly to the underlying data--but the capability in practice raises some challenges. What's the computational and storage cost of unrestricted "drill down?" Can I manage security and privacy if I provide complete visibility to everything in the warehouse? If I find errors, can I fix them (and are they really errors?), and what does that do to everyone else's view of the "correctness" of the data? Does a change I make invalidate a decision made seconds (or hours or days) before on the previous, then-thought-to-be-correct, view of the data? How much history do I need to keep in case I have to undo changes? How do I avoid the kind of "editorial" clashes over the truth that we have seen flare up in wikis? How do I arbitrate disputes? How do I deal with the inevitable "hot spots" in the database that always seem to occur periodically when everyone wants the same information at the same time and the sudden demand triggers significant peaks in usage?
So, pervasive BI isn't easy. It requires an architectural approach. It needs (much) better tools for most users. It needs more than just a pile of data and some analytic models. And it requires us to ask ourselves: If information runs the business, what role do managers and human decision makers play?When can a human override the decision indicated by the information? Should we monitor our decision making to better understand what information actually improves our decisions and why? Should we measure our managers' and associates' ability to use information to do a "better" job? Is "BI blindness" a new impediment to career advancement?
The promise of an information-based (intelligent) business is certainly there. I know we want to realize it. I'm just not sure we're quite ready for the consequences--intended and otherwise.
John Parkinson, the former CTO of TransUnion LLC, has been a technology executive and consultant for over 30 years, advising many of the world's leading companies on the issues associated with the effective use of IT.