A retail media network (RMN) is an advertising platform operated by a retailer that lets brands buy ad placements on the retailer's owned properties — search results, product pages, email, apps, and in-store screens. Ads use the retailer's first-party purchase data to reach shoppers at the moment of intent, and sales are measured in the same checkout system. US retail media ad spending reached $45 billion in 2024.
Amazon Ads built the playbook. Walmart Connect, Target Roundel, Kroger Precision Marketing, and Instacart Ads followed. If you sell through major retailers, you now have no choice but to understand retail media — it controls shelf placement on the digital aisle.
What is a retail media network?
A retail media network is the advertising infrastructure a retailer builds on top of its commerce platform. The retailer becomes a media company: it sells ad inventory across search results pages, product detail pages, category browse pages, email newsletters, mobile apps, and in-store digital screens — all to the brands that already sell through it.
Amazon pioneered this model at scale. Its advertising business now generates over $50 billion annually, making it the third-largest digital ad platform in the US behind only Google and Meta. The model works because retailers hold something Google and Meta lack: actual purchase data. Not intent signals, not demographic proxies — verified transactional history tied to real people buying real products.
- Sponsored search ads: appear in keyword search results on the retailer's site (Amazon Sponsored Products, Walmart Sponsored Products)
- Display ads: banner and native placements on category and product pages
- Off-platform programmatic: retailer audience data applied to open-web buys (Amazon DSP, Walmart DSP)
- In-store digital: end-caps, checkout screens, digital shelf-edge labels (Walmart, Kroger)
- Retailer email/push: sponsored placements in retailer-owned newsletters and app notifications
US retail media ad spending reached $45 billion in 2024 according to eMarketer and is forecast to surpass traditional TV advertising by 2028. This is the fastest-growing segment of digital advertising — driven by cookie deprecation forcing advertisers toward first-party data ecosystems.
Why retail media networks matter for brands
For any brand selling through a major retailer, RMNs are no longer optional. They control digital shelf position the same way physical shelf position controls in-store visibility. Three structural reasons they matter:
- Bottom-funnel precision. Retail media reaches shoppers already browsing product categories with purchase intent. A shopper searching "protein bars" on Amazon is moments from a buying decision — far closer than someone reading a fitness blog where display ads reach them.
- First-party data durability. Retailers hold purchase history, loyalty program data, browsing behavior, and search queries that survive cookie deprecation. As third-party cookies collapse, this data becomes more valuable, not less.
- Closed-loop measurement. The retailer operates both the ad platform and the checkout. You can see, precisely, how many units an ad campaign sold — including in-store purchases linked through loyalty cards. That's a measurement standard no traditional media channel matches.
The counter-argument: retail media is expensive and controlled by the retailer. Margins can thin when brands spend heavily to maintain algorithmic placement. The discipline is to treat it like any performance channel — track ROAS, set targets, and pull spend from campaigns that don't convert.
How retail media networks work
Three systems work together in every RMN: inventory, targeting, and measurement.
Inventory
Retailers offer ad placements across owned properties. Amazon provides 30+ placement types. The highest-intent placements are sponsored search results — appearing at the top of results when a shopper searches your category keyword. Lower-funnel than Google search because the platform is already a shopping environment, not an information resource.
Targeting
Ads use the retailer's first-party data: purchase history, browsing behaviour, search queries, loyalty program membership, demographic data, and household size. An Instacart advertiser can target households that purchased hamburger buns and ground beef in the last 30 days with condiment ads. That specificity is impossible on Google or Meta without a retailer data partnership.
Measurement
Because the retailer owns checkout, it can attribute sales to specific ad exposures. Key metrics: attributed sales, ROAS, new-to-brand customers (did this ad reach someone who hadn't bought your category before?), and in-store sales lift. Amazon's reporting separates same-SKU sales, related-SKU sales, and new-to-brand purchases — giving brands visibility into whether ads are defending existing customers or acquiring new ones.
Types of retail media placements
| Placement type | Platform examples | Best for | Typical ROAS |
|---|---|---|---|
| Sponsored search (on-site) | Amazon, Walmart, Instacart | Bottom-funnel, high intent | 5–10x |
| Sponsored display (on-site) | Amazon, Target, Kroger | Category browsing, retargeting | 3–6x |
| Off-platform programmatic | Amazon DSP, Walmart DSP | Upper-funnel awareness | 1.5–4x |
| In-store digital | Walmart, Kroger, CVS | Physical shelf adjacency | Hard to isolate |
| Retailer email/push | Instacart, Target | Loyalty segment offers | 2–5x |
Real retail media network examples
Three concrete cases showing how different brands use RMNs and what results look like.
1. CPG brand on Amazon Sponsored Products
A protein bar brand spent $15,000 per month on Sponsored Products and Sponsored Brands targeting category keywords and competitor product detail pages. Result: 9.2x ROAS with $138,000 in attributed monthly sales. The product had strong reviews (4.5 stars, 2,000+ ratings) which amplified ad performance — organic conversion rate lifted the paid conversion rate.
2. Grocery brand on Instacart Ads
A condiment brand used Instacart's contextual targeting to reach shoppers who had added hamburger buns and ground beef to their carts. Sponsored product placements appeared in the cart and at checkout. Result: 4.5x purchase rate lift versus untargeted display, with a strong new-to-brand metric showing 60% of purchases came from shoppers who had not previously bought the brand on Instacart.
3. Electronics brand on Walmart Connect
A TV manufacturer ran Walmart Connect sponsored search and display ads during the Black Friday period, coordinating with in-store digital end-cap placements in electronics departments. The integrated measurement showed online ad exposure drove 22% of in-store sales uplift — a cross-channel attribution capability unique to retailer-owned networks.
Retail media network vs programmatic advertising — which to use
Both serve digital ads. The difference is data source, intent level, and measurability.
Use retail media when
- You sell through the retailer and want bottom-funnel sales
- Closed-loop ROAS measurement is a priority
- You need to defend shelf position against competitors
- First-party purchase data targeting is critical
- You are launching a new SKU and need trial
Use programmatic when
- You need upper-funnel brand awareness at scale
- Your product doesn't sell through major retailers
- You want reach across the open web, not just one ecosystem
- CPM efficiency matters more than purchase intent
- You are building a new audience, not converting an existing one
6 best practices for retail media networks
- Fix your product listing before spending on ads. Ad spend amplifies what's already on your product page. A listing with weak images, thin descriptions, or below-4-star ratings will generate clicks that don't convert. ROAS suffers. Optimise the listing first.
- Start with exact-match keywords. Broad match on retail media burns budget fast. Start with tight exact-match targeting on your brand keywords and top category terms. Expand once you have ROAS data to justify it.
- Set a target ROAS before spending, not after. Know your breakeven ROAS from your margins. A 70% gross-margin product needs at least 1.5x ROAS to break even on ad spend. Set campaign targets accordingly and pause anything that misses for two consecutive weeks.
- Monitor the new-to-brand metric. ROAS alone doesn't tell you if you're growing market share or just defending existing customers. Amazon's new-to-brand reporting shows what percentage of attributed sales came from shoppers new to your brand. Prioritise campaigns with strong new-to-brand performance.
- Coordinate retail media with trade promotions. Ads perform significantly better when paired with a price promotion or deal badge. The conversion rate jump from a 15%-off badge more than offsets the margin reduction in most categories.
- Use dayparting if your category is time-sensitive. Grocery and meal-kit categories see search spikes in the early morning (breakfast planning) and early evening (dinner planning). Schedule higher bids during those windows and pull back overnight.
Retail media ads disappear the moment you stop paying. Organic search ranking on Amazon and Walmart is driven by sales velocity, reviews, and listing completeness — all of which your ad campaigns influence. The right approach uses paid to generate sales velocity that feeds organic ranking growth, so you are building a sustainable position, not just renting one.
Common retail media mistakes to avoid
- Running campaigns without conversion tracking: If you can't see attributed sales at campaign level, you are flying blind. Set up conversion attribution before launching.
- Treating RMN budget as trade spend: Retail media earns ROAS. Treat it like a performance channel with targets, weekly reviews, and hard pause rules for underperformers.
- Ignoring negative keywords: On Amazon especially, broad campaigns accumulate irrelevant search terms. Review your search term reports weekly and add negatives.
- Only buying on your own brand keywords: Competitor keyword targeting and category keywords drive new-to-brand growth. A pure branded keyword strategy only defends existing customers.
- Over-rotating to one network: Amazon dominates, but Walmart Connect, Instacart, and emerging networks serve different shopper audiences. Diversify if your distribution does.
Frequently asked questions
Generally yes. Most retail media networks require an active seller relationship on that platform. Amazon DSP allows some off-platform targeting using retail audience data, but primary sponsored ad placements require you to have products listed and available to buy.
Retail media reaches shoppers already on a shopping platform with purchase-intent behavior. Google Ads captures broader web search intent but lacks closed-loop purchase measurement. Retail media offers superior ROAS measurement and first-party data quality, but narrower reach. Most brands run both.
No. Self-serve platforms from Amazon and Walmart accommodate sellers of any size. Amazon Sponsored Products has a minimum daily budget of $1, making it accessible for small brands and independent sellers.
Amazon Sponsored Products averages 7–10x ROAS, which exceeds most display and social benchmarks. Results vary significantly by category, competition level, and how well your product listing is optimised. Set your own ROAS target from your margin structure before comparing to averages.
Programmatic advertising buys inventory across open-web exchanges using third-party or second-party data. Retail media is sold by the retailer directly and uses that retailer's first-party purchase data. Retail media is typically lower-funnel and more measurable; programmatic is broader reach.
Related glossary terms
Sources
