This website uses cookies
Read our Privacy policy and Terms of use for more information.

File / Download
This is not a report about measurement. Not really. It's a report about who controls the economic decisioning layer of commerce—the infrastructure.
This is not a report about measurement. Not really. It's a report about who controls the economic decisioning layer of commerce—the infrastructure that determines which products get seen, which brands get discovered, and which purchases actually happen. Retail media is where that control is being contested right now. And most of the contestants have already lost. They just don't know it yet. Start with the money: $70 billion in 2026, growing 20%+ annually. That's not an advertising line item. That's a toll booth on commerce itself. The retailers who control that toll booth—who can prove what works, who can make allocation decisions in real time, who can capture demand at the moment of purchase—are building structural advantages that compound over time. Everyone else is subsidizing sales that would have happened anyway and calling it marketing. The measurement crisis is a symptom, not the disease. The disease is incentive misalignment baked into the structure of how retail media was built. RMNs grade their own homework. Brands can't compare across platforms. Organizational silos prevent anyone from seeing the whole picture. And now, AI agents are emerging that may skip the entire apparatus—going straight from consumer intent to purchase without ever showing an ad. It started in 2019, when Amazon disclosed that its "other revenue" line had crossed $14 billion. From ads. On a shopping site. CFOs at every retailer with a functioning website had the same thought: We're sitting on a goldmine. By 2022, everyone had launched a "retail media network." The playbook was simple: hire a GM from Amazon, license some ad tech, issue a press release with "unique first-party data," and wait for the money to roll in. The dashboards said everyone was winning. The dashboards were lying.