Retail media is the fastest-growing channel in digital advertising. It is also one of the least standardised. Every retail media network uses different measurement methodologies, different attribution windows and different definitions of what counts as a conversion.
Brands running campaigns across multiple retail media networks are comparing results that are structurally incomparable. A sale attributed to Amazon Sponsored Products is measured differently than the same sale attributed to a Cartology campaign. The reporting looks identical. The methodology underneath is not.
Retail media growth rate in the UK in 2025. Australian growth is tracking similarly. The measurement problem scales with the spend.
The industry bodies are aware of the problem. The IAB and Media Rating Council have published retail media measurement guidelines. Most networks have not adopted them. Networks have a financial incentive to maintain proprietary measurement: it makes their numbers harder to compare and easier to present favourably.
Why it matters
If you are increasing retail media spend, you are likely doing so based on ROAS figures supplied by the networks themselves. That creates a structural conflict of interest. Networks benefit from showing high ROAS. Their measurement choices, including attribution windows and incrementality calculations, reflect that incentive.
This is not necessarily fraud. It is a measurement design problem with commercial consequences. The brands being hurt most are those with spend spread across multiple networks who lack a neutral third-party layer to reconcile the numbers.
The problem also flows upstream into budget allocation. If Network A reports a 4x ROAS and Network B reports a 3x ROAS using different methodologies, reallocating budget from B to A is a decision built on incomparable inputs. You may be optimising in the wrong direction without knowing it.
What to do about it
Request methodology documentation from every retail media network you work with. Ask specifically about attribution windows, what counts as a conversion and how incrementality is calculated. Most networks will provide this on request. Comparing the documentation will show you how comparable the numbers actually are.
Run incrementality tests. Holdout tests, where a matched audience segment sees no retail media advertising, give you a baseline independent of network reporting. They take more work but they are the only way to measure true incremental sales.
Push for third-party measurement. Platforms including DoubleVerify and Integral Ad Science now offer retail media measurement capabilities. A neutral measurement layer does not eliminate the problem but it reduces your exposure to network-favourable reporting.
The spend is real. The attribution may not be. Build the infrastructure to know the difference.
