≡ Menu

Decision modeling, data science and Toyota Financial Services


One of my students (from the UCI Extension Predictive Analytics Certificate in which I teach Business Goals for Predictive Analytics) sent me this article on Toyota Financial Services and its use of data science, predictive analytics, in collections. It’s a great example of how to use analytics to improve your business outcomes and well worth a read. Three key points leap out at me:

  1. What Toyota Financial Services did is a classic example of what I call Micro Decisions (a phrase Neil Raden and I came up with for Smart (Enough) Systems). Instead of treating everyone the same – using a “broad brush” as the article puts it, analytics are used to drive a decision for each specific customer. What will help keep this customer in their car while lowering overall delinquencies.
  2. Solving this kind of problem – a Decision Management problem – often involves a mix of technologies and you need to be solution-focused not technology-focused as a result. As the article says “the whole is greater than the sum its parts”).
  3. It’s essential to keep the analytics team focused on the business problem, not just on the data or the analytic itself. The team co-located and kept its eyes on the decision-making they were trying to improve – “This is a team effort, not just the department, and you have many players that all have to cooperate”.

There’s a great quote in the article:

“Analytics is all about making decisions. Focus on what decisions you have to make and what actions you have to take, rather than starting with data or systems. Understand the business process. Involve the statisticians, and fit the analytics to the corporate culture.”

It’s well worth the read and if you like the article, check out this white paper on framing predictive analytics projects – something that will help you do what Toyota Financial Services did.