Case study: Using data modeling to predict marketing’s effect on sales

It’s hard for marketers to make a direct connection between marketing activity and sales. Harder still is trying to predict how advertising and media spend will affect sales before an activity begins, not just measure it after the campaign runs. It is, however, exactly what we do with the Callahan Intelligence Platform.

As an example of the Intelligence Platform’s capabilities, we developed a predictive data model of expected sales for one of our clients. We looked at historic store-by-store sales, marketing and media spend for this client’s 400+ stores across the U.S. After assimilating, processing and analyzing all their historical data, we were able to project the future sales trajectory for each store.

This gave us the ability to compare projected sales against actual performance as we introduced new marketing or media tactics. The Platform is particularly useful for running tests of new media ideas or approaches and measuring the results on a store- or market-level basis before rolling them out systemwide.

Armed with valuable information from our data model, we conducted a test in a single, large DMA in Texas using very different social media tactics than the client had ever implemented before. When we measured the results, we saw an increase of 2% in comp sales versus the same period the year prior.

Now, you might think plus 2% isn’t so great, but to put that 2% in context, our model actually predicted a decrease of 5% in comp sales for the stores in that DMA. This means our social media test produced a seven-point positive swing—a metric that wouldn’t have been observable, and is considerably more significant than if the 2% lift was assessed without the predictive model.

With this type of information, we can use test results to predict sales in other markets or across the entire chain. And, we can phase the rollout, which creates a perpetual testing environment to continue to prove results and add to the model data over time. This further improves the model’s accuracy, and enhances its predictions.

A data model we create with our Callahan Intelligence Platform allows us to predict the effects of marketing activities before they happen—making that direct connection that marketers need and want. So, if we know that a certain promotion will reliably provide a 7% lift in comp sales, we can apply it to every store in the chain and predict the results of a national promotion.

Photo by Evan Kirby on Unsplash