On-site, off-site and in-store
On-site retail media covers sponsored product listings in the retailer's search results, brand units with a logo and headline, and display placements across the retailer's site and app. Off-site media uses the retailer's shopper data to target audiences on social platforms, display, video and connected TV. Some networks also sell the digital screens in their stores. Each network publishes its own specs and creative guidelines, so a campaign across several retailers becomes dozens of sizes and variants. Some units are built automatically from the product catalogue, with the brand supplying only a headline and logo; others take fully designed creative.
Approvals run on the retailer's rules
Networks review creative before it runs, and they apply their own policies on top of the law. Common reasons for rejection: a price that differs from the retailer's own, a mention of a competing retailer, an unsupported claim such as lowest price, outdated packaging, or a layout that breaks the retailer's brand guidelines. A rejection late in the process means a missed flight, often a missed promotion window. The studio generates one approved master into each network's formats, with product-lock keeping the pack exact and approved claims enforced, and approvals run on the whole set with an audit trail rather than file by file.
The product and price must be exact
A retail media ad is one click from the product page and a short walk from the shelf, so any gap between them is visible at once. A generated pack shot showing the wrong variant, last year's packaging or a colour that drifted will be compared with the real thing by every shopper who clicks. It is also a consumer law problem: a misleading picture of a product is a misleading action under the UK's DMCC Act 2024 and the EU's Unfair Commercial Practices Directive. Prices and promotions must match what the retailer charges when the ad is seen, with dates that match the offer.
Closed-loop measurement changes creative decisions
Retailers link ad exposure to purchases in their own sales data, so creative versions can be judged on what sold rather than on clicks. That rewards disciplined testing: tag every asset by concept, format and product so results can be read back to the creative decisions behind them. It does not make networks comparable with each other. Each measures inside its own walls, with its own attribution windows and methods, which is why the IAB and the MRC published Retail Media Measurement Guidelines in January 2024. Compare versions within a network, and treat cross-network comparisons as rough until incrementality tests back them up.
Updated 25 September 2026