EMarketer’s new forecast shows 17.9% annual growth
eMarketer’s March 2026 forecast projects US omnichannel retail media ad spend will hit $69.33 billion this year, a 17.9% jump from 2025 levels. [7] That growth rate outpaces nearly every other digital ad category, and it raises an uncomfortable question for brands still treating retail media as an experimental line item: how long can you afford to underfund the channel where your customers are closest to buying?
The $69.33 billion figure covers onsite formats (sponsored listings on retailer websites and apps), offsite placements bought through retailer DSPs, in-store digital screens, social ads run through retail media networks, and connected TV inventory sold by retailers. [2] That breadth matters because it signals retail media is no longer synonymous with “sponsored product ads on Amazon.” It is a full-funnel system, and the forecast reflects that expansion.
Some secondary reports have floated global retail media figures in the $165 to $200 billion range for 2026, but I could not verify those numbers against a primary eMarketer source. [11] The US-specific $69.33 billion projection, however, is well-documented and draws on eMarketer’s standard methodology: historical revenue data, economic conditions, executive interviews, and reported earnings from the major retail media networks. [2] There is also a mild disagreement in the data ecosystem about whether growth will exceed 20% this year; some analyst posts on LinkedIn suggest higher rates, but the 17.9% figure is the one eMarketer’s own forecast methodology supports when you back into it from the $69.33 billion total. [10]
What makes 17.9% particularly striking is the base it is growing from. Retail media was already a $58-plus billion channel in 2025. Maintaining near-20% growth at that scale is unusual for any ad format that has been commercially available for over a decade. For comparison, Meta is projected to surpass Google in total digital ad revenues this year, [4] but the growth rates powering that shift are considerably more modest than what retail media is posting.
Why first-party data drives this budget shift
Retail media’s growth story is inseparable from the collapse of third-party cookie targeting and the broader signal loss that has plagued programmatic advertising since Apple’s ATT rollout. Brands are not flocking to retail media networks because the ad units are beautiful or the reporting dashboards are elegant (they often aren’t). They are moving budgets because RMNs sit on deterministic purchase data that no other channel can match.
When a brand runs a sponsored product campaign on Amazon or Walmart Connect, the targeting relies on what shoppers have actually searched for, browsed, and purchased on that retailer’s platform. That first-party data loop closes the gap between ad exposure and transaction in a way that walled-garden social platforms and open-web programmatic simply cannot replicate. A Meta campaign can optimize toward modeled conversions; a retail media campaign can show you whether the person who saw your ad put the product in their cart and checked out, sometimes down to the SKU level.
This data advantage compounds over time. As more brands invest in retail media, the networks accumulate richer behavioral signals, which in turn improve targeting precision, which attracts more spend. It is a flywheel, and the 17.9% growth rate reflects a flywheel that has been spinning for several years and is still accelerating. In my analysis, the single biggest driver of that acceleration is not any one platform’s innovation but rather the industry-wide realization that first-party purchase data is the most defensible targeting asset left in digital advertising.
There is a caveat worth acknowledging. First-party data from RMNs is powerful for lower-funnel conversion campaigns, but its value for upper-funnel brand building is less proven. Brands that shift too much budget into retail media risk starving the demand-generation activities that feed the purchase funnel in the first place. The 17.9% growth figure does not distinguish between brands making strategically sound allocations and brands simply chasing the path of least measurement resistance.
Amazon’s ad platform shows market dominance
Amazon remains the gravitational center of US retail media. While eMarketer’s detailed company-level revenue share breakdowns for 2026 are locked behind a PRO+ paywall, [2] every publicly available indicator points to Amazon commanding a majority share of that $69.33 billion. Amazon’s advertising services segment has consistently reported double-digit quarterly revenue growth in its earnings, and its DSP infrastructure gives it reach well beyond its own properties.
Amazon’s dominance rests on three structural advantages that competitors have struggled to erode. First, its search volume is enormous: for many product categories, more purchase-intent searches start on Amazon than on Google. Second, its self-service ad platform is mature, with years of iteration on bidding algorithms, keyword tools, and reporting. Third, Amazon Prime’s membership base creates a logged-in, identity-rich environment where ad targeting and attribution operate with a precision that open-web advertising lost years ago.
Walmart Connect is the most credible challenger, and its recent moves suggest it is trying to compete not by matching Amazon’s onsite scale but by outflanking it off-site. Walmart Connect now offers self-service full-funnel campaigns on Meta and TikTok, complete with an “Add to Cart” beta that lets shoppers add up to 10 Walmart items without leaving the social platform. [6] A beta test with Burt’s Bees produced 42% ad-attributed sales and 95% cart completion within one week. [6] Those are eye-catching numbers, though a single-brand beta is far from proof of scalable performance.
Other RMNs, including Etsy’s Offsite Ads and networks from grocery chains like Kroger and Instacart, occupy meaningful niches. [2] But the gap between Amazon and everyone else remains wide enough that most brands treat Amazon as a mandatory line item and everything else as incremental. Until a second player can offer comparable search volume, self-service tooling, and closed-loop attribution, Amazon’s share is unlikely to shrink in absolute terms, even if its percentage of the total dips as the market expands.
Omnichannel is the next growth frontier
Onsite sponsored product ads built the retail media category, but the next wave of growth is coming from off-site, in-store, social, and CTV formats. eMarketer’s forecast explicitly uses the term “omnichannel retail media” to describe the $69.33 billion figure, and that framing is deliberate. [2]
“Even as Walmart continues to build on its owned media business, faster growth is shifting off-site, where revenue is set to grow at more than twice the pace of on-site.”
Sarah Marzano, eMarketer analyst [6]
That quote captures the structural shift well. US retail media social ad spend alone is expected to grow 23.6% to $1.26 billion in 2026, [6] which is a small slice of the $69.33 billion total but growing faster than any other subsegment. According to eMarketer and Bain data from August 2025, 32% of sell-side RMN leaders cited emerging off-site capabilities, including social, as their primary revenue driver over the next 12 months. [6]
Commerce media networks and CTV platforms are now competing for overlapping advertiser budgets, which creates both opportunity and fragmentation. [5] A brand running Walmart Connect campaigns on TikTok while also buying CTV inventory through Amazon’s DSP and running in-store digital screens through Kroger’s network faces a measurement nightmare. Each channel uses different attribution windows, different identity graphs, and different definitions of “conversion.” The omnichannel promise is real, but the omnichannel measurement infrastructure is still catching up.
In-store retail media is a particularly interesting frontier. Digital screens at the point of purchase offer something no other digital channel can: physical proximity to the product at the moment of decision. Local marketing strategies are increasingly incorporating these formats alongside CTV and digital out-of-home. [9] But in-store inventory is inherently limited by the number of screens a retailer is willing to install, and standardization across retailers is nonexistent. I expect in-store to remain a small but high-value piece of the omnichannel mix for the foreseeable future, not a volume play.
How brands can capture this market opportunity
A $69.33 billion market growing at 17.9% sounds like an obvious place to invest, but the operational reality is messier than the headline suggests. Only 40% of retail media networks offer self-service access to sales data, and 41% of industry participants cite the lack of self-serve technology as a barrier to scaling their retail media programs. [6] That means the majority of RMNs still require managed-service relationships, manual reporting, and a level of operational overhead that makes them impractical for mid-market brands without dedicated retail media teams.
For brands already active in retail media, the highest-use move right now is testing off-site formats before they get expensive. Walmart Connect’s social integration with Meta and TikTok is still in its early stages, and early adopters tend to benefit from lower CPMs and less auction competition. The Burt’s Bees beta, with its 42% ad-attributed sales rate, is exactly the kind of result that attracts a wave of new advertisers, which then drives up costs for everyone. [6] If you are going to test, test now.
Brands entering the space for the first time should resist the temptation to spread budget across every available RMN. Amazon is the default starting point for a reason: its self-service tools are the most mature, its audience is the largest, and its attribution is the most granular. Once a brand has a profitable Amazon program with repeatable processes, expanding to Walmart Connect, Instacart, or category-specific networks makes strategic sense. Trying to run five RMN programs simultaneously with a two-person team is a recipe for wasted spend and incomplete data.
Measurement standardization, or the lack of it, is the single biggest obstacle to rational budget allocation across retail media networks. Each RMN defines attribution differently, reports on different timescales, and provides different levels of granularity. Until the industry converges on something resembling a common measurement framework, brands need to build their own internal models for comparing performance across networks. Relying solely on each RMN’s self-reported ROAS is like letting every student grade their own exam.
One thing I would watch closely over the next 12 months is whether CTV inventory sold through retail media networks starts to cannibalize traditional CTV buys. Commerce media networks and CTV platforms are already competing for the same advertiser dollars, [5] and the retailer’s first-party data advantage could tip the balance. If Amazon or Walmart can offer CTV targeting powered by actual purchase history, the value proposition for buying CTV through a traditional DSP weakens considerably. That convergence, more than any single format or platform, is what could push retail media ad spend well past $69 billion in 2027 and beyond.
Sources
- EMARKETER Forecasts, Estimates, & Historical Data
- US Retail Media Ad Revenue Share, by Company Forecasts
- Search and Retail Media Search KPIs Q2 2026
- Meta to Surpass Google in Digital Ad Revenues for First Time Ever
- Commerce Media Networks and CTV platforms are competing for an increasingly concentrated ad market
- Walmart Connect expands social capabilities as retail media chases off-site growth
- Retail Media is a $69B Opportunity. So Why Is It Still So Hard to Get Right?
- James Hay’s Post on Digital OOH
- FAQ on local marketing: How AI, CTV, and digital OOH are reshaping local advertising
- U.S. Retail Media Report Card: 2026 Ad Spending Growth Projections
- Facebook/EMARKETER Post on Commerce Media
