At a glance
A high retail media ROAS does not prove the campaign created additional sales. Compare networks only after aligning sales definitions, attribution windows and product scope, then test incrementality and assess contribution margin before moving more budget.
A retail media campaign reports a return on ad spend of eight to one. That sounds excellent. Before moving more budget, ask what counts as a sale, how long the attribution window lasts and how many of those customers were likely to buy anyway.
Retail media measurement connects advertising exposure with shopping behaviour in a retailer’s environment. That proximity is valuable. It does not make every attributed sale an additional sale.
Agree the definition before comparing the number
IAB Europe’s Commerce Media Measurement Standards V2.1, published in January 2026, clarify gross and net sales definitions, introduce a formal definition of incrementality and update new-to-brand reporting. The guidance retains a default 30-day reporting lookback while calling for flexible alternatives.
Those details matter because two apparently identical ROAS figures may have very different ingredients. One report may include returns differently. Another may count a wider set of products or a longer period after exposure.
Ask each network for its definitions in writing. Keep the original report alongside any normalised comparison so the team can trace how the numbers were calculated.
Separate four questions
| Question | What to inspect |
|---|---|
| Was the advertising delivered? | Impressions, placement, viewability and relevant reach |
| Were purchases attributed? | Click/view rules, windows and included products |
| Did the brand gain customers? | New-to-brand definition and purchase history window |
| Did advertising create additional sales? | Incrementality method and comparison design |
Do not make one metric answer all four. A new-to-brand buyer may have purchased without the ad. A repeat customer may generate genuinely additional value. The labels describe different aspects of the result.
Put margin beside revenue
If you spend £10,000 and attribute £80,000 in sales, reported ROAS is eight. If those sales carry a 25% contribution margin before advertising, they provide £20,000 before the media cost. Subtract the £10,000 media spend and £10,000 remains before any other relevant costs.
That example still assumes all attributed sales belong in the calculation. If only part of the revenue is incremental, the commercial answer changes. Promotions, retailer fees, returns and fulfilment can change it further.
Our view is that retail media should be judged as a commercial investment, not a competition for the largest attributed revenue multiple. Revenue is an input to that decision, not the decision itself.
Ask what the comparison actually compares
A credible incrementality study needs a defensible comparison between people, stores or areas that differ in exposure. Ask how groups were assigned, whether promotions or availability differed and how uncertainty is reported.
A before-and-after sales chart can be useful context, but it may also reflect seasonality or a price change. Our incrementality testing guide explains why a comparison needs to address what would have happened without the campaign.
Product availability deserves particular attention. An apparently weak campaign may be advertising items that are out of stock or difficult to find. A strong one may coincide with a promotion that would have driven sales anyway.
Build a comparison sheet your finance team can read
For each network, record spend, attributed net sales where available, attribution rules, product scope, margin assumptions and the evidence for additional sales. Add the campaign’s job: defending existing customers, launching a product or reaching a different audience.
Use separate measurement methods where the question requires them. Do not hide incompatible figures inside one blended return.
Then decide what you have learned. A network can be valuable even when its reported ROAS is lower, if it reaches customers you would otherwise miss or delivers stronger incremental profit. The best number is the one that helps you make the right next investment.
