Reading this made me smile and reminded me of a conversation with a data science guy a few months ago. L was solving for a problem that all sales organizations can relate to. Are they catching the buying (or churn) signals early and often and are they equipped to do what is needed once such signal is detected. In L's case the sales people were selling into complex organizations where buying decisions were made by an array of factors including geo-political issues.
He had identified the full laundry list of features and set up some Markov Chain Monte Carlo methods to build his prediction models. The efficacy of this thing was apparently very hard to discern but the key variables that signaled a buying interest L could have deduced through pure commonsense.
I asked him what he thought of his grand adventure to arrive at confirming answers he already knew. L was not sure he was ready to call his bag of tools not dazzling enough. He was more inclined to believe that if he thought about the problem harder a smarter solution would emerge - and one that he would not have already known. We promised to check on his progress a few months out. I am all but certain there is no golden needle in his haystack but I look forward to being proven wrong.
Comments