Retail Media Networks: Why Every Grocer Is Now an Advertising Business
A supermarket earns two or three cents of net margin on a euro of groceries. It earns far more on a euro of advertising. That single arithmetic explains the fastest-growing category in media.
A supermarket makes a very thin net margin on the food it sells — low single digits is normal, and in competitive markets it is thinner still. On advertising, the margin is a different species of number entirely. Once a retailer realises that its checkout data and its website are a media asset, the decision to build an advertising network stops being a marketing question and becomes a P&L question.
What a retail media network actually is
A retail media network is a retailer selling advertising against its own audience, using its own purchase data to target and measure. It has three surfaces:
- On-site. Sponsored product listings, banners and search placements on the retailer’s own site and app. This is the highest-margin inventory because the shopper is already in a buying state.
- Off-site. The retailer’s purchase data used to target the same shopper on social platforms, connected TV and the open web. Lower margin, much larger reach.
- In-store. Digital screens, shelf displays, audio and sampling — increasingly measured against loyalty data.
The asset underneath all three is the same: verified purchase history tied to an identifiable customer. That is the one dataset the open web lost with third-party cookies and the one retailers had all along.
The margin logic
Retail is a volume business with very little pricing headroom. Advertising is a capacity business where the incremental cost of showing one more sponsored listing is close to zero. When a retailer converts a percentage of its site traffic into advertising inventory, the revenue drops through to profit at a rate the core business can never match. For several large grocers, retail media now contributes a share of group operating profit wildly disproportionate to its share of revenue.
This is why the category grew so fast. Amazon built an advertising business that generates more revenue than almost every European media group on the continent, and did it largely from inventory that already existed. Every other retailer with meaningful traffic drew the obvious conclusion.
Who is building what
The landscape now spans three tiers. Global platforms — Amazon above all — operate at a scale that makes them a mandatory line in most consumer goods media plans. National grocery and general merchandise retailers across Europe have launched networks built on loyalty data, often with a data science partner. And a long tail of category specialists — pharmacy, DIY, electronics, fashion marketplaces — is monetising narrower but highly qualified audiences.
Fragmentation is the problem this creates. A brand selling across ten European markets may face fifteen separate retail media contracts, each with its own ad server, taxonomy, reporting cadence and definition of a "sale". Standardisation efforts are underway through industry bodies, but adoption remains uneven, and the practical consequence for advertisers is a real increase in operational overhead.
The measurement problem nobody has solved
Retail media's pitch is closed-loop measurement: the network shows the ad and observes the purchase. That is genuinely powerful, and it is also a conflict of interest. The seller of the inventory is also the arbiter of whether the inventory worked.
The specific risk is attribution of sales that would have happened anyway. A sponsored listing on a search for your own brand name captures shoppers already looking for you. Reported return on ad spend looks superb; incremental sales may be close to zero. This is the same trap that made branded paid search look miraculous a decade ago, and the remedy is the same: geo-based holdout tests, matched-market experiments and periodic incrementality studies run by someone other than the seller.
A practical rule: any retail media reporting that only ever produces flattering numbers is not measurement, it is invoicing with charts.
Where the budget comes from — and why it matters
The most consequential organisational question in retail media is which budget pays for it. In many consumer goods companies, retail media is funded from trade or promotional budgets negotiated by the sales team, not from the media budget managed by marketing. That has three consequences: the spend is often negotiated as part of a listing or promotional agreement rather than assessed on media merit; it rarely appears in marketing mix models, so its true contribution is invisible; and nobody owns the trade-off between a retail media placement and the broadcast, video or out-of-home investment it may be displacing.
Advertisers that handle retail media well tend to do one structural thing: they measure it inside the same framework as everything else, even when a different team negotiates it.
What to negotiate
- Log-level or granular data access, not just dashboard summaries. Without it, independent verification is impossible.
- A defined incrementality testing programme, with holdout groups agreed in advance and results shared regardless of outcome.
- Attribution windows and rules stated in the contract, not configurable by the seller mid-flight.
- Separation of trade terms from media terms, so that advertising investment is not implicitly a condition of shelf space.
- Off-site data usage limits — what the retailer may do with your campaign data, and for how long.
Where this goes next
Two developments will define the next phase. The first is in-store retail media, where digital screens at the point of decision are becoming measurable against loyalty data — the last unmeasured surface in the shopper journey. The second is the extension of the model beyond retail: airlines, banks, delivery platforms, telecom operators and hotel groups all hold transaction data and audience attention, and several have already launched commerce media offerings. The category will probably stop being called retail media, because retailers will no longer be the only ones selling it.
Note on data. Retail media market sizing varies substantially between research firms because definitions differ — some include marketplace search advertising, some include in-store, some count off-site extension in the digital display total instead. Treat any single market-size figure with caution and check the underlying definition before comparing two forecasts.
Sources
The claims in this article rest on the documents below. Each is linked to what it establishes, so you can check any statement against its origin rather than taking ours for it.
- IAB Europe, Retail Media Measurement Standards V1 (April 2024) — the first European framework for comparing retail media investment on consistent metrics
- IAB Europe, Commerce Media Measurement Standards V2.1 — the revised standard following public comment in late 2025, broadened beyond retail
- IAB Europe, 101 Guide to Retail Media (June 2024) — definitions of on-site, off-site and in-store inventory
- IAB and IAB Europe, in-store retail media definitions — the measurement guidelines for physical store inventory
Frequently asked questions
What is a retail media network?
It is a retailer selling advertising against its own shoppers, using its own purchase and loyalty data to target and measure. Inventory spans the retailer’s site and app, off-site placements on social and connected TV powered by its data, and increasingly in-store digital screens.
Why are retailers so keen on advertising?
Because the margin is incomparably better. Grocery retail runs on low single-digit net margins; advertising inventory has near-zero incremental cost, so the revenue converts to profit at a far higher rate. For several large retailers, retail media now contributes a share of operating profit far exceeding its share of revenue.
Is retail media worth it for smaller brands?
It can be, but the economics turn on category and competition. Sponsored listings work best where shoppers browse rather than arrive with a fixed brand in mind. In categories where a shopper searches for a specific brand, much of the measured return is likely to be non-incremental — you are paying to capture demand you already had.
How do I know if retail media is actually driving sales?
Run holdout tests. Withhold retail media in matched regions or store clusters and compare sales against exposed areas. Reported return on ad spend from the network itself is not sufficient evidence, because the seller of the inventory is also the party attributing the outcome.
Should retail media come out of the trade budget or the media budget?
Functionally it is media, so it should be evaluated against media alternatives even when the sales team negotiates it commercially. The most common failure is retail media funded from trade budgets, excluded from marketing mix models, and therefore never compared with the broadcast or video investment it displaces.