Wednesday, May 11, 2011

What is fact-based, anyway?

By: Frank Buytendijk, chief marketing officer

After the series of articles on how Marx predicted the end of Google and Facebook (part 1, part 2, and part 3), I have just finished a new series of articles on the "analytical world". Again, this series has three parts.

In part 1, I argue that although "fact-based management" and "competing on analytics" both sound really cool, the concepts have some serious shortcomings. According to most philosophers, there is not much we can really know for sure. And analytics, as a model of the world, are removed even one step more from reality.

Part 2 discusses the practical consequences of this analysis, taking recent examples such as the outburst of the Eyjafjallajokull volcano in Iceland and the credit crunch. Both cases where people relied too much on analytical models. Part 2 also contains a critique on predictive analytics.

I do realize it is easy to criticize things, so if I do I should present an alternative view. This I have done in part 3, where I will introduce the concept of "messy analytics". We should pay more attention to factors that mess up the model, as they represent this nasty thing called reality. In fact, I propose to mess up models on purpose, just to see if they can handle what is often called "black swans". Messing up model so they become more precise may sound counterintuitive, but I am almost convinced you will agree with the logic I will present.