The Challenge

Every advertiser asks the same question: is my advertising actually driving results, or are these people who would have converted anyway?

Traditional attribution tells you a conversion happened after someone saw an ad — but it can't tell you whether the ad caused it. That distinction matters, because without it, you're measuring correlation, not impact.

We set out to answer this definitively using a methodology most agencies don't have access to: first-party bid-level incrementality analysis.

The Approach

As a programmatic DSP operator, we have access to something unique — not just records of the ads we served, but records of every auction we entered, including the ones we lost. This creates a natural experiment with two groups drawn from the same targeting pool:

This isn't a modeled estimate. Both groups are real people, identified through first-party data, who participated in the same real-time auctions during the same period. The difference in conversion rates between these two groups is the true, measured lift created by the advertising.

Why This Matters

Most digital advertising measurement relies on attribution models that credit an ad whenever a conversion happens after exposure. But this can't distinguish between a customer who converted because of your ad and one who was already going to convert and simply happened to see it along the way.

Our incrementality approach measures the difference in outcomes between exposed and unexposed groups — the same methodology used by academic researchers and enterprise advertisers, now applied at the local and regional level using first-party programmatic data.

The result: a clear, defensible answer to the question "did my advertising work?" — grounded in observed behavior from a real control group, not assumptions or statistical projections.

What Makes This Possible