Editor's note: Michael Lieberman is founder and president of Multivariate Solutions, a New York data science and strategy firm.

Data. Companies have lots of it. So much that we have begun urging them to commoditize it. Or at the very least put it to practical use. With today’s technology not only can nearly everything be gathered, counted and measured but the information can be stored and then processed at record speeds. The result is analysis that goes beyond sums, averages and basic statistics to aggregates, benchmarks, recommendations and predictions. So what does one do with all of this game-changing data and analysis? Create a data product. They’re all around us and they’re changing the way we as consumers interact with companies and the way businesses interact with each other.

There are stories to be told hidden within the databases of major corporations. Traditional marketing research projects often take data that is fielded by way of panels or other field measures. While these will continue to function, an opportunity to use existing customer databases and purchased data – for example, Nielsen PRIZM segments – to apply to marketing research techniques and produce an actionable product are now becoming more prevalent.

Many companies wish to link their customer experience (CX) variables to specific financial outcomes. They wish to estimate customer retention based on data they already possess and identify variables that will predict financial performance. This article will illustrate how we would go about this. 

Anaheim is a fictional insurance company with a lot of data. Anaheim finds itself at a crossroads; it has accumulated massive amounts of data and has conducted a significant number of analyses. Still, it is removed from having a simple-to-implement input/output model that management can use for making investment decisions in new customer experience initiatives.

At the same ...