A retail media network is an advertising business a retailer builds on top of its own checkout and browsing data, selling shelf-adjacent placements, on-site search ads and off-site targeting back to the brands that sell through it. CPG brands are shifting budget there because the retailer can show a sale, not just a click, closing the loop that television and social ads cannot close.
Why did retailers build ad networks?
A large retailer sits on something no ad platform outside its own walls has: a running record of what a specific household actually bought, not what it clicked on or claimed to want. That record is worth more to a brand than almost any other targeting signal, because it is behaviour rather than intent.
Retail margins on the goods themselves are thin and shrinking under private-label pressure and price competition, so a high-margin data and advertising layer sitting on top of the same infrastructure, the same site, the same loyalty programme, is close to free to build and very expensive for anyone else to replicate. Grocers, marketplaces and pharmacy chains reached the same conclusion roughly at the same time, which is why the format spread fast rather than staying a single company's experiment.
- 01
Separate the claim from the channel
Treat a retailer's reported lift as one input, not a verified fact.
- 02
Check what closed-loop actually closes
Confirm whether the attribution covers only on-platform sales or the full basket.
- 03
Weigh the margin cost
Price the placement against what it would have cost to earn the same shelf position.
- 04
Hold back a brand-building budget
Keep spend that builds demand separate from spend that only captures it.
Why are CPG brands shifting budget here?
A television or social impression tells a brand it reached someone. A retail media placement, in the best case, tells a brand that someone reached for the product and paid for it. For a category where the purchase decision is made in seconds at the shelf or in a search bar, that difference in what gets measured is the whole argument.
Trade spend used to buy end-caps and slotting through a broker relationship with limited visibility into what it produced. Retail media replaces some of that with a placement that reports back in the same units a performance marketer already thinks in, cost per acquisition, return on ad spend, incremental units. Budget follows the channel that can defend itself in a spend review, and increasingly that is retail media rather than a linear TV buy or an open-web display campaign.
How is retail media measured?
The pitch is closed-loop attribution: the same retailer that shows the ad also processes the transaction, so it can report sales lift directly from its own checkout data without stitching together a third-party cookie or a survey. That closes a loop open-web advertising has never been able to close cleanly.
The catch is that the retailer grading its own ad product is also the retailer selling it, and the methodology behind a reported lift number is rarely fully open to the brand buying the placement. Brands that treat every network's dashboard as directly comparable to another's, or to their own in-house measurement, tend to overstate what closed-loop actually proves. Comparing a lift number against a brand's own independent shopper research is usually what turns a vendor claim into a defensible planning input.
| Traditional trade spend | Retail media network | |
|---|---|---|
| Bought through | Broker or category manager relationship | Self-serve or managed ad platform |
| Reported back as | Shelf placement, limited visibility | Impressions, clicks, attributed sales |
| Who grades the result | Largely unverified by either side | The retailer selling the placement |
What is the margin tension?
A retailer monetizing its own shelf and search results is, structurally, charging a brand for prominence the brand used to earn through price, packaging or distribution muscle. That cost gets absorbed somewhere, and it usually shows up as pressure on the brand's own margin or on the price the shopper eventually pays.
The tension sharpens where a retailer also sells its own private label next to the brand paying for placement. The retailer is simultaneously the landlord, the auctioneer and a competing tenant, a position no other media channel puts a brand in. That structural conflict is part of why the debate over private label's growth and the debate over retail media spend are really the same conversation about where shelf power actually sits.
Where should retail media sit in a plan?
As one channel with a specific, provable job, converting demand that already exists near the point of purchase, rather than as a substitute for the brand-building work that creates demand in the first place. Retail media is good at capturing intent; it is not designed to generate it.
The brands managing this well tend to hold retailer-reported numbers to the same scrutiny they would apply to any vendor's self-graded results, and they keep an independent read on what is actually driving the sale before reallocating budget at scale.