Bill is a sales professional with a mathematical background. He has been working on optimization of sales processes for most of his career. Bill and Bryan worked together on commercial applications of the open source SciDB database project. Separately, Bryan works on a few long-term projects with colleagues at Montefiore and Albert Einstein College of Medicine using survival analysis and other machine learning tools to model health care

outcomes.

 

One day while discussing one of Bill's forecasting problems, we realized how well statistical techniques from survival analysis could apply to sales funnel forecasting. For instance, consider the simple analogy of stages of a disease progression to the stages of an opportunity in a sales pipeline. Instead of the chance of, say a hospitalization (or worse), we're instead modeling the chance of making a sale (a good thing). Similar models are of course also widely applied in actuarial science and econometrics. But, despite the wide use of machine learning models in B2C sales and marketing, we found very little application of sophisticated modeling techniques in the B2B sales process.

 

Funnelcast is our attempt to formalize these models for the sales process, and to automate and simplify their use as much as we can.