Defining the Next Phase of Retail Media in 2026

Retail media moves beyond endemic brands

Home Depot’s Orange Apron Media opened a beta integration with Pinterest Media Network Connect in 2026, letting insurance companies and telecom providers target homeownership audiences that have never bought a single SKU from Home Depot’s shelves. [13] That move, quiet as it was, signals something bigger than one retailer’s partnership strategy: retail media networks are no longer just a place where CPG brands bid on sponsored product listings. They are becoming audience platforms, and the advertisers showing up have nothing to sell in the store.

Global retail media ad spend is projected to hit $196.7 billion in 2026, representing roughly 16% of all global ad spend. [11] A meaningful share of that growth is coming from non-endemic advertisers, brands that don’t sell through the retailer but want access to its first-party purchase data. Think of a credit-card company buying placement on a grocery site’s checkout page, or a travel firm targeting high-basket electronics buyers on Walmart.com. [7] These advertisers are willing to pay a premium for deterministic purchase signals that no third-party cookie replacement can match, and retailers have figured out how to package those signals into audience segments like “new parents” or “loyalty members” with reach and outcomes data attached.

Home Depot’s case is instructive because its audience is so specific. Homeowners researching kitchen renovations or HVAC systems carry intent signals that are gold for mortgage lenders, home insurance providers, and smart-home companies, none of which sell through Home Depot’s aisles. [12] By activating those audiences off-site through Pinterest and Reddit partnerships, Orange Apron Media can monetize its data without cluttering its own product pages with irrelevant ads. [17] For the advertiser, the value proposition is straightforward: reach people who are demonstrably spending money on their homes, verified by transaction data rather than inferred from browsing behavior.

This non-endemic expansion is where the real revenue diversification lives. Endemic advertisers have finite budgets constrained by their trade marketing allocations, but non-endemic demand is theoretically unbounded because it draws from brand budgets across every category. The question for 2026 is whether retailers can build the measurement infrastructure to prove incremental value to advertisers who can’t close the loop with a same-platform sale. [4]

Programmatic buying becomes the default standard

U.S. programmatic ad spend is forecast at $203 billion in 2026, and commerce media (the category that includes RMNs) grew twice as fast as programmatic display in 2025. [6] That growth rate tells you where the money is moving: advertisers want to buy retail media inventory through the same DSPs they use for everything else, not through bespoke self-serve portals that each retailer built independently.

For years, the friction in retail media buying was operational. Every RMN had its own interface, its own targeting taxonomy, its own reporting cadence. Running campaigns across Amazon, Walmart Connect, and Instacart meant managing three separate workflows with incompatible metrics. Programmatic integration collapses that complexity. When retail media inventory is available through Trade Desk, DV360, or Amazon DSP, buyers can allocate budgets across networks in a single campaign, optimize bids with unified frequency caps, and compare performance in one reporting layer.

Walmart has been aggressive here. Walmart Connect generated $6.4 billion in global ad revenue in FY2026, with U.S. growth of 41% year-over-year. [16] [15] Part of that acceleration comes from making inventory programmatically accessible, which lowers the barrier for mid-market brands that don’t have dedicated retail media teams. Home Depot followed a similar path, integrating with Trade Desk and Yahoo before layering on the Pinterest beta for non-endemic activation. [13]

I’ve seen some claims that programmatic middleman cost savings range from 10% to 30%, but platform-specific benchmarks remain murky, and the actual take rates vary so widely by network that quoting a single number feels misleading. [6] What is clear is that the operational savings in workflow consolidation are real, even if the media cost savings are harder to pin down. For agencies managing ten or more RMN relationships, the ability to run cross-network campaigns from a single DSP is worth the programmatic fee regardless of CPM arbitrage.

Amazon still dominates this space with $54 billion in ad revenue, capturing roughly 87% of the RMN market by spend. [14] That concentration is a problem for the rest of the ecosystem. Walmart, Kroger, Target, and Home Depot are all growing fast in percentage terms, but the absolute dollar gap with Amazon remains enormous. Programmatic access helps smaller networks compete by reducing the switching cost for advertisers, but it also commoditizes inventory in ways that could compress margins for retailers without differentiated audiences.

Off-site and in-store extensions expand reach

Retail media’s original value proposition was proximity to the point of purchase: a sponsored listing on a search results page, a banner on a category page. That proximity still matters, but the growth story in 2026 is happening off-site and in-store, where retailers are extending their first-party data into channels they don’t own.

Off-site activation works through clean rooms and DSP integrations. A retailer like Walmart or Home Depot packages an audience segment (say, “customers who bought power tools in the last 90 days”), makes it available through a demand-side platform, and lets advertisers target that segment across the open web, social platforms, or CTV. Walmart’s Vizio acquisition gave it a direct CTV pipeline, and the results have been notable: a Café Bustelo campaign on Walmart Connect’s CTV inventory saw a 44% lift in view rate compared to standard display. [16] That kind of performance data is what makes off-site retail media compelling for brand budgets that historically went to linear TV or YouTube.

In-store is the other frontier, and it is messier. Andrew Lipsman has described in-store retail media as a “royal rumble” by end-2026, with digital screens, AR overlays, and audio ads all competing for attention in physical aisles. [6] Programmatic out-of-home is powering some of this evolution, connecting digital street-level inventory to the same buying platforms that serve online display. [10] But in-store measurement is still primitive compared to digital. Foot traffic attribution, dwell time on screens, and basket-level incrementality are all being tested, yet none of them have the precision of a click-through conversion on a product detail page.

In my view, the off-site story is overhyped relative to in-store. Off-site retail media is essentially a data licensing play dressed up as media, and it competes directly with Meta and Google’s own audience products. Retailers have better purchase data, yes, but Meta has better optimization algorithms and vastly more inventory. In-store, by contrast, is a genuinely differentiated environment where retailers have a monopoly on attention at the physical shelf. If the measurement catches up, in-store could be the higher-margin opportunity.

Why measurement standardization remains a challenge

Every conversation about retail media’s maturity eventually hits the same wall: measurement. And in 2026, that wall has not moved as much as the spend numbers would suggest it should have.

For endemic advertisers, measurement is relatively straightforward because the purchase happens on the same platform where the ad ran. ROAS calculations, while imperfect, at least have a closed loop. Non-endemic measurement is a different problem entirely. When a home insurance company runs ads through Orange Apron Media’s Pinterest integration, how does anyone prove that the campaign drove incremental policy sign-ups? The retailer has purchase data but not the advertiser’s conversion data, and the advertiser has conversion data but can’t match it back to the retailer’s audience without a clean room or some form of identity resolution. [4]

Agency sentiment reflects this gap. Roughly 40% of agencies view retail media as a comprehensive full-funnel channel, which means 60% still see it as something less than that. [11] The skepticism is rational. Without standardized incrementality frameworks, every RMN is grading its own homework. Amazon’s measurement methodology differs from Walmart’s, which differs from Kroger’s, which makes cross-network comparison nearly impossible for multi-retailer advertisers.

IAB and MRC have been working on retail media measurement standards for over a year, but adoption is uneven. Retailers with dominant market positions (Amazon, primarily) have little incentive to adopt common standards that would make their inventory directly comparable to smaller competitors. Smaller networks want standardization because it reduces the friction of getting on a media plan, but they lack the use to force it. This dynamic is familiar from the early days of digital video measurement, where YouTube resisted MRC accreditation for years because it didn’t need external validation to attract spend.

The practical consequence for advertisers is that retail media still requires more manual analysis than most digital channels. You can’t just pull a cross-network report from your DSP and compare apples to apples. Each network’s attribution window, incrementality methodology, and reporting granularity are different enough that media mix modeling or matched-market tests remain the most reliable way to evaluate true contribution, and those approaches are slow and expensive.

How to adapt your media mix now

Given the speed at which retail media is absorbing budget, the question for most marketing teams isn’t whether to invest but how to restructure existing allocations without double-counting audiences or cannibalizing other channels.

Start by auditing where your retail media spend overlaps with your programmatic display and social budgets. If you’re running off-site retail media through a DSP and also buying the same audiences directly on Meta, you’re likely bidding against yourself. Frequency capping across those environments is still imperfect, so the risk of oversaturation is real. One approach that I’ve seen work is to treat off-site retail media as a replacement for a portion of prospecting spend on social, not as an incremental line item. You’re buying the same impressions with better data signals, so the budget should come from the same pool.

For non-endemic advertisers exploring retail media for the first time, the entry point matters. Home Depot’s Pinterest beta and Walmart Connect’s CTV inventory are relatively low-risk tests because they operate in familiar formats (social, video) with the added layer of purchase-based targeting. [13] [16] On-site placements on a retailer where you don’t sell products are a harder sell internally because the KPIs are less intuitive, so off-site is a better proving ground for building organizational buy-in.

Measurement infrastructure should be built before scale, not after. If you can’t run incrementality tests at your current spend level, you won’t be able to at three times the budget either. Negotiate clean room access or lift study commitments as part of your insertion orders. Retailers are increasingly willing to offer these because they know measurement is the bottleneck to unlocking larger budgets, particularly from non-endemic categories that are accustomed to rigorous attribution from their digital spend elsewhere.

What new retail media formats will emerge

AR ads are the format getting the most attention in 2026, though “attention” and “adoption” are different things. Programmatic AR in retail media works by serving 3D product overlays (built in USDZ or GLB formats) through RMN inventory, targeted with first-party data and bought via RTB. Minimum budgets sit around $10,000 to $25,000, which is accessible for mid-market brands but still high enough to filter out casual experimentation. [6] Measurement for AR campaigns focuses on dwell time and return reduction rather than traditional CTR, which makes sense for categories like furniture or apparel where visualization reduces purchase anxiety.

CTV is a more immediately scalable format. Walmart’s Vizio integration gives it owned-and-operated CTV inventory that can be targeted with Walmart purchase data, a combination no pure-play streaming service can replicate. [16] Walmart’s AI shopping assistant, Sparky, is also being tested as an ad surface, with 81% of customers reportedly using it for product research. [6] If conversational commerce becomes a real shopping behavior (and that “if” is doing a lot of work), sponsored recommendations within AI assistants could become a format category unto itself.

Programmatic digital out-of-home tied to retail locations is another format gaining traction. Screens in parking lots, at gas stations, and on in-store endcaps can now be bought through the same DSPs that serve online display, with targeting informed by the retailer’s loyalty data. [10] The creative constraints are significant (short dwell times, no click-through, limited interactivity), but the proximity to purchase is unmatched.

What I’d watch most closely isn’t any single format but the convergence pattern: retailers stitching together on-site search ads, off-site social and CTV, and in-store digital screens into a single campaign with unified frequency management and cross-channel attribution. Nobody has fully delivered that yet. Walmart is closest, and Amazon has the data to do it but hasn’t opened its walled garden enough. Whichever network cracks unified cross-surface buying with credible measurement will likely capture a disproportionate share of the non-endemic budgets that are just now entering the channel. That’s the race worth watching through the rest of 2026.

Sources

1. Four Trends That Will Define Retail Media’s Next Phase – RETHINK Retail
2. 4 Retail Media Themes Shaping the Industry in 2026 – Harvest Group
3. Programmatic AR Ads in Retail Media Networks: 2026 Guide
4. The Future of Retail Media: How Retailers Become Media Networks
5. Retail Media Networks Market Evolution – LinkedIn
6. Retail Media Spend to Reach $196.7B by 2026
7. Home Depot Retail Media Strategy: Orange Apron Media Analysis
8. Orange Apron Media Unveils New Partnerships
9. Digital Advertising Statistics 2026
10. From Retailer to Media Platform: Walmart Connect
11. Walmart, Vizio Link Content To Commerce
12. Home Depot deepens retail media push
13. How programmatic Out of Home is powering retail media’s evolution