Follow Datanami:

Seeing Around the Corner: Proxies During a Pandemic

How to Predict Store Sales to Within 2% Accuracy During the Pandemic, and Beyond

In fast-changing retail environments, predicting consumer behavior and purchasing patterns is critical to fulfilling demand and maintaining profitability. In this white paper you’ll learn how WorldQuant Predictive (WQP) and CVS Health partnered during the height of the pandemic in 2020 to predict–within 2% accuracy–individual store sales.

In an unprecedented time of pantry loading and toilet paper hoarding, CVS Health knew it couldn’t rely on historical data to plan for future demand. They tapped WorldQuant Predictive and their Global Research Network of data scientists to get some answers about future consumer behavior. Using WQP’s proprietary AI platform, Quanto™, the WQP/CVS team looked at +50 disparate, open-source, alternative datasets. Through an iterative process they quickly generated a customized predictive model for CVS stores.

As it turns out, one of the surprising sources for predictive datasets was Zillow—-it uncovered a hidden alpha signal that helped CVS Health accurately predict future consumer behavior. The takeaways from this white paper can be extended to solving prediction problems common to retailers. You’ll learn how your organization can apply this approach for fast, accurate, transformative insights in dynamic environments that are continually evolving.