Most retail media reporting counts sales the retailer would have made anyway. The question nobody asks is what the spend added.
The pitch for retail media is that it finally closes the loop. You buy the placement, the shopper buys the product, and the retailer can show you the receipt. No panel, no modelled attribution, no arguing about the last click. After twenty years of marketing measurement that never quite settled, a number that ties spend to a basket looks like the end of the argument.
It is not, because the receipt does not say why the basket happened. A sponsored placement on a retailer's search results sits in front of a shopper who has already opened the app, already typed the category, and already decided to buy something in it. Attributing that sale to the placement is not measurement. It is bookkeeping. The brands with the highest reported return on ad spend are usually the ones buying their own brand terms, which is to say paying a toll on demand they created somewhere else and already owned.
The number that matters is the incremental one, and it is always smaller. Ask what would have sold without the placement, and the honest answer needs a test rather than a report: hold the spend out of a matched set of stores or regions, run it for long enough to clear the noise, and read the difference. Every retailer with a media business can run this. Very few offer it, because the answer is not flattering to the rate card, and very few brands demand it, because the reported figure is the one that goes in the deck.
This matters more each year, because retail media is no longer a rounding error in the plan. It is now a serious share of trade spend in most consumer businesses, and in many it has quietly absorbed money that used to build the brand. That is a live trade. Short-term conversion is easier to prove than long-term demand, so it wins every budget round on evidence, and the loss shows up three years later as a brand that has to discount to be chosen.
The way to run this is not to spend less. It is to know which part of the spend is buying growth and which part is buying credit for growth you already had. Test one large category, hold the spend out, and read the incrementality against the reported figure. Then negotiate the rate against the real number rather than the reported one. The retailers running the best media businesses will meet you on this, because they would rather have a brand that keeps investing than a brand that discovers the gap on its own.
