AdTheorent introduces a ‘cost per incremental visit’ model for offline foot traffic

Once it became possible to measure when someone walked into a store after a digital ad exposure, it was only a matter of time before marketing platforms built performance ad models around that concept.

GroundTruth (formerly xAd) was the first to introduce a “cost-per-visit” ad model, which was quickly followed by a “guaranteed visits” model from Retale, and then a “cost-per-revenue” model from Empyr.

GroundTruth and Retale both work with Placed (recently acquired by Snap) for third-party validation. So does a company called AdTheorent, which this week introduced a “cost per incremental visit” (CPIV) model. In other words, advertisers don’t pay for store visits per se; they pay only for additional visits that they wouldn’t have received but for the ad campaign.

I spoke with Josh Walsh, AdTheorent’s president, about the company’s methodology and the reaction of customers. He told me that he was seeing a “ton of demand” from advertisers for CPIV (GroundTruth said something similar not long ago). “If 100 people saw the ad and came into the store, but only 20 were incremental, they’re only paying for 20,” explained Walsh.

Placed provides the reporting, and the customer is billed based on its numbers, though AdTheorent is also recording visits for its own modeling. Quick service restaurants, convenience stores, lottery ticket sellers and retailers are current customers, according to Walsh. He said that they pay a flat fee for each customer, which is negotiated on a case-by-case basis.

AdTheorent uses historical visitation data and predictive modeling to price incremental visits. Using data obtained from a range of sources, AdTheorent creates a score for each user that appears to the system and determines when to serve specific ads to maximize engagement and visitation probability.

Walsh said that AdTheorent creates household profiles, with location history data, augmented with CRM data, device IDs, cookie IDs and other information to round out audience profiles.

While the system is highly sophisticated, Walsh admits it’s not perfect. “There’s a lot of gray in determining incremental visits.” But that’s where Placed comes in — to create confidence that the visits are happening as reported.

In addition to the companies already mentioned, Google, Facebook and numerous others are investing in or beefing up offline measurement capabilities, which leads me to believe that we’ll see more of these CPIV-style models in the future.