ROAS has been the comfort metric of performance marketing because it turns complicated promotional activity into one clean number. But a clean number is not always a useful one.
Ibotta founder and CEO Bryan Leach and Liquid Death chief media and digital commerce officer Benoit Vatere argue that brands need to look harder at what their promotions actually changed. The shift is from reported return to incremental business impact.
Incrementality is harder, and more useful
ROAS can show the revenue associated with a promotion. It does not necessarily show how many of those sales would have happened anyway. That distinction matters in retail, where shoppers may already be close to buying and a discount can receive credit for a decision it did not create.
The discussion points to matched-audience testing as one way to separate genuinely incremental sales from activity that would have occurred without the promotion. It also highlights cost per incremental dollar as a more useful measure for protecting margins than a top-line return figure alone.
That is a more demanding standard. It asks marketers to compare exposed and unexposed audiences, watch sales signals in real time, and accept that some apparently efficient activity may not be creating much additional demand.
The measurement shift changes the budget conversation
The implication is not that ROAS disappears. It is that ROAS becomes one input rather than the final answer. If brands can identify incremental sales, they can move money between retail channels before a quarterly review tells them the original allocation was wrong.
The same logic applies beyond immediate conversion. Adweek’s report also connects measurement with reducing customer churn and growing household penetration. A promotion that brings back an existing buyer, reaches a new household, or protects a relationship may create value that a short-term transaction report misses.
That makes independent verification more important, too. Retail media is asking brands to trust increasingly complex data environments while also taking more of their promotional budgets. Third-party checks are not a decorative layer of credibility. They are part of the argument that the result is real.
For brands, the consequence is straightforward: the best-performing channel will increasingly be the one that can prove what it changed, not simply the one that can claim the most attributed revenue. The metric is becoming less about making spend look efficient and more about showing that it produced something that would not otherwise have happened.