Retail media is the slotting fee, recoded as advertising.
US advertisers will spend $69.33 billion on retail media networks in 2026, up from $58.79 billion in 2025 1. Of the $10.53 billion in incremental spend, $9.42 billion is going to Amazon Ads and Walmart Connect 2. Industry trade press is calling this the next great channel. Agencies are building retail media practices. CPG brands are reallocating from Meta and Google into Sponsored Products.
The firm’s position is the opposite. For most operators outside the top 100 national CPG brands, retail media is a structural cost of doing business on a marketplace, recoded as advertising. The spend earns its keep in narrow circumstances. Most of the rest is a slotting fee.
What retail media is
A retail media network sells ad inventory on a retailer’s owned property: search results on Amazon.com, category pages on Walmart.com, the home feed on Target Circle, the checkout flow on Kroger. The advertiser is usually selling a product the retailer already stocks. The retailer takes a cut of the spend, plus the wholesale margin on the unit, plus, in many cases, slotting fees or trade marketing payments outside the ad budget.
This is different from Google Search or Meta in one important way. On Google, a search for “running shoes” is a query the buyer brought to the platform. The advertiser is paying for proximity to a query the platform did not originate. On Amazon, a search for “running shoes” is a query the buyer brought to Amazon, but the advertiser is also paying Amazon a wholesale-margin tax on the eventual sale, plus a referral fee, plus, in many categories, FBA fees. The ad spend is on top of an existing channel cost.
Said plainly: the retailer is already getting paid by you on every unit. Retail media charges you again for the visibility you used to earn through stocking, packaging, and product performance. It is the slotting fee that grocery brands have always paid, scaled to the marketplace era, and rebadged as a digital advertising opportunity.
The math at the brand level
Take a mid-market consumer electronics brand selling on Amazon with $5 million in annual Amazon GMV. Amazon’s referral fee at 15% takes $750,000. FBA fees take another $400,000 to $600,000 depending on category. Wholesale-margin compression from Amazon’s pricing pressure costs maybe another 5 to 10% versus the brand’s owned site. Before any ad spend, Amazon is collecting roughly 25 to 30% of the GMV in retailer compensation.
Then the brand adds retail media. A typical ACoS (advertising cost of sale) target on Amazon is 15 to 25% 3. Add that to the existing platform cost and Amazon now collects 40 to 55% of GMV across all fees. The brand is paying Amazon to advertise products that Amazon already takes a cut on. Each dollar of incremental ad-driven revenue carries a margin profile that is materially worse than the brand’s owned-channel revenue.
This is fine if the marginal customer would not have purchased without the ad and would not have come to the brand’s owned site. The math breaks when retail media spend is partly cannibalizing organic Amazon ranking, partly cannibalizing the brand’s own site, and partly defending shelf position the brand should already own through product quality. In the firm’s audits across consumer brands, the cannibalization share is usually 30 to 50% of the reported lift.
When retail media earns the spend
Three situations make retail media worth the cost.
The first is when a brand is launching a new product and needs visibility on a platform where organic ranking takes 60 to 120 days to build. Sponsored Products is the only way to get a new SKU in front of buyers during the launch window. Treat the spend as a launch cost, time-boxed, and exit when organic ranking takes hold.
The second is when a brand is defending against a competitor who is bidding on its branded terms. If a competitor is buying ad space on searches for your brand name, you need to be there or you lose that traffic. This is reactive spend, not growth spend, and the budget should be sized to the threat.
The third is when the platform represents a clear competitive moat: the buyer cannot easily comparison-shop, the platform’s audience is unique, and the conversion path is short. Amazon for everyday-consumables in categories where the buyer is brand-loyal fits this. Instacart for grocery does. Most other platforms do not.
When it does not
The list of situations where retail media is a tax, not an investment, is longer.
When a brand has under-invested in product photography, copy, and reviews on its retailer page, retail media buys traffic that converts at half the rate it could. The spend masks an unfixed problem. The fix is to invest in the product detail page; then evaluate ads.
When a brand is using retail media to maintain ranking on terms it ranks for organically, the spend is replacing free traffic with paid traffic at a wholesale-plus-ad-fee cost structure. The right move is to measure organic share of voice and reduce ad spend on terms where organic ranking is durable.
When a brand is buying retail media on marketplaces where its conversion rate is meaningfully below its owned site, retail media is teaching the brand to depend on a channel that earns less margin. The right move is to diagnose why the marketplace conversion rate trails, fix what is fixable (often photography, reviews, or a missing variant), and then decide what fraction of the brand’s volume the marketplace should represent.
The cannibalization that does not show up in the dashboard
Retail media’s reported ROAS looks high because marketplaces use last-click attribution within their own walls. A buyer who searches your brand, clicks a Sponsored Product ad, and converts is counted as ad-driven, even though the buyer would have scrolled five inches further and bought organically. The platform’s ROAS overstates incrementality by design.
This is the same critique the firm and others have made about branded search on Google for years, and the answer is the same: run a holdout test. Drop retail media spend in a single category for 30 days. Measure total category revenue, not just Sponsored Products revenue, against the prior 30 days controlling for seasonality. The difference is the incremental contribution of retail media. In the firm’s experience, that contribution is typically 40 to 60% of the reported number. The rest is captured demand that would have converted without the ad.
The fact that 89% of incremental retail media spend in 2026 is going to Amazon and Walmart 2 is not evidence that those platforms are uniquely valuable. It is evidence that those platforms have the most leverage to extract spend from brands that already sell on them. Concentration of spend follows concentration of platform power, not concentration of marketing efficiency.
The competitor position the firm disagrees with
Several agencies have framed retail media as the next major growth channel for consumer brands, with messaging that frames not investing in retail media as a missed opportunity 4. The framing is convenient for an agency, because retail media is hours-billable work that scales with spend.
The framing is wrong for most operators. Retail media is not a growth channel in the way that adding paid search to a brand-only account was a growth channel. It is a cost of distribution. The brands that win on it are the brands that would have been winning on the same marketplaces without it, plus a few exceptional cases where the channel structure is genuinely advantageous. For the broad middle of consumer brands, retail media is a transfer from brand margin to retailer margin, processed through an ad account.
The firm’s position on diversification, laid out in the more-channels piece, applies here too. Adding retail media to an account that has not yet exhausted owned-channel growth dilutes attention, dilutes data, and shifts spending toward a channel where the marketplace owns the customer relationship. Concentration on channels where the brand owns the customer beats diversification into channels where the marketplace does.
What to do before signing up
Before reallocating budget into Amazon Ads, Walmart Connect, or any retailer network this quarter, run three checks.
Check what fraction of your current marketplace revenue would persist if you spent zero on retail media for 30 days. If the answer is 80% or more, the channel is largely defending demand that would convert organically.
Check what your blended margin looks like on marketplace revenue versus owned-site revenue. If marketplace margin is more than five points lower, growth on the marketplace channel is actively destroying enterprise value, even if the top-line revenue grows.
Check what your owned-site retention and repeat-purchase rate look like compared to marketplace. If owned is materially higher, every retail media dollar that captures a first-time buyer on the marketplace is sending the buyer into a relationship the brand does not own.
The $69 billion will get spent regardless. The question is whether your brand’s share of it is paying for growth or paying the bill.