Predictive Modeling | WildFig Data

How to more accurately predict lead volume and sales;
and optimize marketing communications strategy
over both near and long-term horizons?


The challenge

A global medical device manufacturer in the diabetes segment had no quantitative framework in place to measure effectiveness or produce accurate forecasts for its entire sales funnel. The data required to answer these questions and build the required tools were siloed in legacy database systems and third-party tools like Salesforce.

The WildFig solution

Our team began with a thorough analysis of all first-party data and previously used reporting tools to create a more modern framework for analysis and forecasts. Upon completion of the munging of their data, our team developed a series of dynamic dashboards for ongoing use and reference by senior leadership.

With the dashboards in place, we then developed a series of predictive models around sales and lead volume; and began quantifying the impact of elements of their sales program. Our models deconstructed the overall trends, daily and weekly trends with and without spend. The most powerful component of our work was the ability for forecasts to be adjusted dynamically based on spend, tactics and temporal components. This was supported by ongoing A-B testing of messaging and platform delivery.

Fifty-two percent of all new patients were directly attributed to campaign and supporting optimizations. Leads turned to customers and advocates as Facebook followers jumped 29 percent YOY, which is equivalent to 38,084 new followers.