Attribution Is Not Incrementality:
The Measurement Gap in Retail Media.

One of the biggest challenges in retail media is also one of the easiest to gloss over.

A campaign runs.

-The dashboard looks positive.
-Sales are attributed.
-The report says performance was strong.

None of the above automatically proves that customer behavior has changed.That is the measurement gap I see repeatedly in retail media.

Attribution has value. It can help tell part of the story. It can show where activity happened, where engagement occurred, and where sales were associated with a campaign.

Please keep in mind that attribution is not the same thing as incrementality. That distinction matters much more than many teams want to admit.

Because vendors are not really paying for reports that show correlation. They are paying for confidence that the spend influenced an outcome that would not have happened, or would not have happened in the same way, without the campaign.

That is a much tougher standard, and it should be.

That shift is happening as the market matures. IAB described commerce media as the fastest-growing digital advertising channel, but also noted that projected growth would slow from 25.1% in 2024 to 15.6% in 2025. In other words, the category is still expanding, but the easy momentum phase is already giving way to a more disciplined one.

Retailers are now operating in an environment where brand partners are becoming more disciplined about where they place funds and how they evaluate return. That means RMN reporting cannot stop at surface-level performance metrics. Impressions, clicks, conversions, attributed sales, and ROAS may all be directionally useful. But on their own, they often leave too much room for ambiguity.

-Did the media create demand? Or did it simply get credit for demand that was already likely to happen?

-Did it influence a customer to buy sooner, buy more, switch brands, or enter the category? Or did it show up near a purchase that was going to occur anyway?

These are not academic questions. They sit at the core of RMN credibility.

I have seen many retail media discussions where the reported performance appears strong at first glance, but becomes harder to defend once a merchant leader, finance partner, or vendor analyst starts asking more detailed questions. That does not always mean the reporting is wrong. Often, it simply means the measurement model is not mature enough yet.

The danger is not only external skepticism. It is internal mislearning.

When weak measurement is accepted too easily, businesses start rewarding the wrong behaviors. Campaigns that look effective on paper get repeated. Budget gets allocated to tactics that may be overstated. Teams optimize to what is easiest to report rather than what is most commercially meaningful.

That is how a category can grow while quietly weakening its own foundations. The better path is not to abandon performance reporting. It is to strengthen it.

Retailers should work toward measurement approaches that more closely reflect actual influence, actual lift, and actual business impact. That means asking harder questions about customer exposure, timing, repeatability, margin effect, vendor relevance, and what changed versus what appeared in the data.

The strongest RMNs in the next phase of the market will not be the ones with the most polished dashboards.

They will be the ones who can sit across from a vendor and, with discipline and credibility, explain why this activity mattered and whether it is worth funding again.

That is a much more valuable capability, and in the long run, it is what will separate a media network from a media story.

Source note: IAB, Commerce Media: At the End of the Beginning? (Apr. 8, 2025) and the IAB 2025 Outlook Study.