In an earlier post, I wrote about the importance of the decision model. The decision model seems complicated, because most companies haven’t been building them. But in fact the decision model can increase transparency and understanding — an important role of analytics.
Businesses are operating in an increasingly complex world, and it follows that analytics must often be complex as well. Nevertheless, analytic complexity must be justified. A senior manager should be able to get a clear answer to “Explain to me why this customer decision needs to be so complicated. What value are we getting from it? What are we learning? How is this going to make tomorrow’s decisions better than today?”
Much unnecessary complexity can be traced back to trying to improve decisions by starting with the decision tree rather than the decision model. Editing a decision tree of any size is an inherently risky venture. This form of representation is visually complicated, and therefore, working within it, one can easily make mistakes.
By comparison, the underlying decision model can be much simpler to understand and change. Edit the decision model, and you will have a better grasp of the structural reasons for the changes you’re making, and thus can have confidence in the decision tree derived from it, no matter how complex.