Retail media in 2026: closed gardens, paid premiums.
Every brand has a retail-media budget. Most are paying premium for inventory they can’t audit and audiences they can’t take with them. The shape of a defensible plan is finally becoming clear.
Retail media in 2026 is the largest single growth line in advanced media. Global retail-media spend has crossed $100B. Amazon Ads remains the gravitational center. Walmart Connect has crossed an inflection. Kroger Precision Marketing, Albertsons Media Collective, Best Buy Ads, Target Roundel, and Instacart Ads each hold meaningful share in their respective categories. The category is structural now.
It is also the category where senior brand teams have the least clarity on what they’re buying.
The closed-garden problem
Each retail-media network runs on its own ID graph. Amazon’s is the deepest. Walmart’s is the broadest grocery footprint. Kroger’s is the highest-fidelity loyalty data in the US. None of them are portable. The audience built in Amazon Marketing Cloud cannot be activated through Walmart Connect. Cross-network frequency capping doesn’t exist. Cross-network measurement runs on different methodologies and isn’t directly comparable.
The retail-media-network view is that the closed garden is the asset — and they are not wrong from their position. The brand’s view should be that the closed garden is the cost — and most brand plans haven’t priced that cost.
What is actually working
- Amazon DSP for off-Amazon reach against Amazon shoppers. The single most defensible retail-media line item for non-grocery brands. Amazon’s ID resolution against off-Amazon CTV and display is the strongest in the category.
- Walmart Connect for grocery-driven conversion. Best when paired with first-party CRM matching. The closed-loop reporting is reliable in categories where Walmart has scale.
- Direct-API integrations with the top two RMNs. Pull exposure and conversion data into the brand’s own clean room for unified measurement. Operationally heavy. Pays back, especially across multi-RMN plans.
What is not working
- Treating retail media as the CPG channel. Non-CPG brands using RMN inventory as an audience-extension proxy. The audience is what was sold; the inventory was the channel. The performance reads accordingly — strong on intent capture inside the retailer’s walled garden, weak on the broader reach claim.
- Cross-RMN comparison reporting. Each network’s ROAS methodology is different. Comparing Amazon ROAS to Walmart ROAS line-by-line creates false signals. Comparable measurement requires either a standardized third-party (Circana, Nielsen) or roll-up into a unified MMM.
The currency question
Which retail-media network is the brand’s measurement currency? Most senior teams will say Amazon by default because Amazon scaled first. By 2026 that default is contestable. Walmart’s measurement maturity is closing. A brand whose buyer base skews more Walmart than Amazon should be re-examining whether the measurement default has been reviewed recently — not assumed.
The read
“A retail-media plan that can’t be audited isn’t a plan. It’s a cost center.”
Two procurement questions to ask every retail-media line item: (1) Can the audience be exported, or is this a single-network commitment? (2) Is the conversion attribution running through the retailer’s clean room (auditable) or through their managed dashboard (not auditable)? Most line items will fail at least one of these. That doesn’t mean the spend is wrong — it means the spend has a cost the brand hasn’t priced.
A working note. If you want to walk through your retail-media plan at the line-item level — direct line at hello@odellnco.com.