The 84% Problem: Why Most Retail Media Networks Aren't Winning

Market concentration and what it means for brands navigating the retail media landscape

Retail media is the fastest-growing advertising channel in the US, expected to reach $166 billion in digital spend by 2025. More than 200 retail media networks now exist globally. Airlines, banks, convenience stores, and grocery chains have all launched ad businesses. The opportunity seems clear: retailers have first-party purchase data, advertisers want access to that data, and the economics are attractive for everyone involved.

Except that’s not quite how it’s playing out.

Amazon and Walmart will capture 84% of all retail media ad spending in 2025. The remaining 200-plus networks will compete for the other 16%. More striking: that 16% share has barely grown—increasing by less than one percentage point between 2019 and 2024, even as the total market expanded nearly five times.

The market is growing rapidly, but the benefits are concentrating rather than distributing. This has significant implications for advertisers trying to navigate retail media and for retailers contemplating their own ad networks.

Why Scale Matters More Than You’d Think

Retail media seems like it should favor specialization. Target knows its customers. Kroger knows its shoppers. Instacart knows online grocery buyers. Each network offers access to distinct audiences with unique purchase behaviors.

In theory, advertisers should work with multiple networks to reach different customer segments. In practice, most work with four or fewer partners, even though many have relationships with five or more networks. According to a January 2024 Association of National Advertisers study, 58% of US marketers work with at least five retail media networks.

The disconnect suggests execution challenges. Managing campaigns across multiple platforms, each with different reporting standards, different attribution methodologies, and different optimization interfaces, creates operational burden. For brands with limited teams, consolidation around a few large platforms makes practical sense.

But there’s a deeper issue: measurement standardization. Advertisers can’t easily compare performance across retail media networks because networks measure differently. Does a “view” mean the same thing on Walmart Connect as on Kroger Precision Marketing? Are “conversions” counted consistently? Without standardized metrics, advertisers struggle to allocate budgets efficiently across platforms.

Industry trade associations have published guidelines for standardization. The IAB released frameworks for incremental measurement in commerce media in October 2025, outlining when experiments, model-based counterfactuals, and econometric models are most appropriate. But adoption has been slow. Networks aren’t strongly incentivized to standardize when it might reveal unfavorable performance comparisons.

Amazon and Walmart benefit from this measurement chaos. When cross-platform comparison is difficult, advertisers default to platforms with proven scale and established ROI—which reinforces the concentration of spend.

The Data Story Is More Complicated

First-party retail data is the core value proposition of retail media networks. Advertisers can target based on actual purchase behavior rather than inferred intent. Someone who bought diapers last week is a better target for baby products than someone who merely searched for parenting content.

But access to first-party data isn’t uniform. Amazon and Walmart have hundreds of millions of customers across diverse product categories. They can build detailed profiles showing purchase patterns over time. A smaller specialty retailer has fewer customers, narrower categories, and less comprehensive purchase history.

This creates a data quality gap. Amazon can tell you not just that someone bought batteries, but that they buy batteries every three months, always choose premium brands, also buy outdoor equipment, and typically shop on weekends. A specialty outdoor retailer knows someone bought batteries but lacks the broader context.

For advertisers, richer data enables better targeting, which drives better performance, which justifies more spend. The data gap compounds the scale advantage.

Walmart’s $2.3 billion Vizio acquisition in 2024 illustrates how large players are expanding their data assets. The deal gives Walmart connected TV data—viewing habits, streaming preferences, household composition signals—that can be linked to purchase data. This creates targeting capabilities that smaller networks can’t match without similar acquisitions.

The gap is widening, not closing.

The In-Store Challenge

Digital retail media—sponsored products on websites, display ads on mobile apps—is well-established. In-store retail media represents the next growth frontier, with spending expected to grow 47% in 2025 according to eMarketer forecasts.

The opportunity makes sense. Most retail sales still happen in physical stores. Digital screens at entrances, checkouts, and end caps can deliver contextually relevant ads to shoppers who are about to make purchase decisions. The attribution loop can close tightly: show ad, measure immediate purchase lift.

But in-store retail media requires significant infrastructure investment. Digital screens, content management systems, data platforms to power targeting, attribution technology to link ad exposure to purchase—the costs add up quickly. Hy-Vee, which operates 570 Midwest grocery stores, announced plans to partner with Grocery TV to power 10,000 screens across locations. That’s substantial capital deployment for uncertain returns.

Large retailers can justify this investment. Amazon owns Whole Foods. Walmart has 4,700 US stores. The fixed costs of technology and infrastructure spread across massive store counts and customer bases create favorable unit economics.

Smaller retailers face harder math. Installing screens across 50 locations costs roughly the same per store as installing across 500 locations, but the advertising inventory generated is much less valuable. Fewer stores mean fewer impressions, which means less advertiser demand, which means lower CPMs. The investment may not clear the ROI threshold.

This capital requirement creates another advantage for large players. They can invest in formats that smaller networks can’t afford, offering advertisers more inventory across more touchpoints.

The CTV Play

Connected TV represents a particularly lucrative extension of retail media. Retail media CTV ad spending will grow 45.5% in 2025, and one in five CTV ad dollars will go to retail media by 2027, according to eMarketer projections.

Walmart’s Vizio acquisition positions the company to capture this spend. Amazon already has substantial CTV inventory through Prime Video and Fire TV. Both can link TV ad exposure to purchase behavior, proving incremental sales lift with closed-loop measurement.

Most other retail media networks don’t have owned CTV inventory. They can partner with CTV platforms, but that introduces intermediaries, reduces margins, and limits measurement capabilities. Direct ownership of CTV inventory creates a structural advantage.

Target, the third-largest retail media network by share, doesn’t have a clear CTV strategy comparable to Amazon or Walmart. This suggests that even established networks struggle to compete across all formats without significant M&A activity or partnerships.

What About Non-Endemic?

Endemic advertising—brands selling products on a retail platform advertising on that same platform—was the original retail media model. Unilever advertising Dove soap on Walmart.com makes intuitive sense.

Non-endemic advertising expands the opportunity. Brands that don’t sell on the platform can still advertise to reach the retailer’s audience. Insurance companies, auto manufacturers, financial services firms—categories that aren’t sold in retail—can buy ads targeted using retail purchase data.

Amazon, Walmart, and Best Buy are embracing non-endemic opportunities to expand reach. This requires different infrastructure—ad formats suitable for awareness and consideration rather than just purchase, attribution models that don’t rely on immediate transaction data, creative capabilities beyond sponsored product placements.

It also requires scale. An advertiser considering non-endemic spending wants reach. They’re not targeting the 3% of households that shop at a specific specialty retailer; they want meaningful national reach. Only the largest networks can credibly offer this.

For smaller networks, non-endemic advertising remains largely theoretical. They lack the audience scale, the technology infrastructure, and the sales relationships with non-endemic advertisers to make it viable.

The Margin Story

From retailers’ perspective, retail media is extraordinarily profitable. Advertising now accounts for almost a third of Walmart’s $6.7 billion operating income. For a business operating on thin retail margins, this represents a significant earnings contribution.

But building and operating a retail media network requires investment: technology platforms, data infrastructure, sales teams, advertiser support, creative services. These are fixed costs that need to be covered by ad revenue.

Large networks achieve strong unit economics. Amazon’s retail media business likely operates at 60%+ margins—mostly software and sales, minimal variable costs. As revenue scales, margins improve.

Smaller networks face different math. A network generating $10 million annually in ad revenue might spend $4-5 million on platform costs, sales, and support. A network generating $100 million might spend $15-20 million on the same functions. Margins improve substantially with scale.

Coresight Research projects that retailers can expect retail media networks to generate a 70% increase in gross margin compared to core retail operations. But that projection likely applies more to large networks than small ones. A specialty retailer with limited ad inventory and high operational costs relative to revenue may see much lower margins—perhaps not enough to justify continued investment.

The Advertiser’s Dilemma

For brands, the concentration creates practical challenges. Working with Amazon and Walmart makes sense—they deliver scale, proven ROI, and sophisticated targeting. But relying exclusively on two platforms limits reach and creates dependency.

The promise of retail media was that brands could reach customers across multiple purchase environments, tailoring messages to different contexts. That vision requires a healthy ecosystem of diverse networks.

Instead, advertisers face a binary choice: work primarily with the two dominant platforms and accept limited reach, or spread budgets across many smaller networks and accept operational complexity, inconsistent measurement, and uncertain ROI.

Some categories have no choice. If you sell grocery products, you need to advertise on Amazon and Walmart. But you also need to reach shoppers at regional chains, specialty grocers, and meal delivery services. Concentration at the top means underinvestment in the long tail.

Where This Leads

Three scenarios seem possible:

First, continued concentration. The measurement and infrastructure challenges persist. More ad dollars flow to Amazon and Walmart. Smaller networks struggle to prove ROI, lose advertiser support, and eventually shut down or exist as minimal operations.

Second, consolidation. Mid-sized networks merge to achieve greater scale. We might see regional grocery chains pooling their retail media operations, specialty retailers forming alliances, or acquisitions by media companies looking to enter retail media. This could create a few networks with enough scale to compete, even if they don’t match Amazon and Walmart.

Third, standardization. Industry bodies successfully push measurement standardization. Advertisers can compare performance across platforms. Technology infrastructure becomes available as white-label solutions, reducing fixed costs. Smaller networks prove ROI in specific niches and capture profitable segments. The market becomes more fragmented but healthier.

The second scenario seems most likely. We’re already seeing moves in this direction—Criteo offers retail media platforms to multiple retailers including Target, CVS, and Best Buy. These infrastructure partnerships lower barriers and enable smaller players to offer sophisticated capabilities without building from scratch.

But even with consolidation and standardization, the fundamental advantages of scale, data depth, and CTV ownership favor the largest players. Retail media may always be a concentrated market.

What Advertisers Should Do

Given this landscape, what’s the practical path forward for brands?

First, establish baseline performance on the big platforms. Amazon and Walmart are where most advertisers need to start. Build competency there, understand what good looks like, and establish performance benchmarks.

Second, test selectively on smaller networks. Don’t write off all non-Amazon/Walmart networks. Test a few that reach your specific audience or offer unique inventory. Measure carefully and expand what works.

Third, demand better measurement. Push networks to provide incrementality testing, not just attributed conversions. Insist on standardized metrics where possible. Vote with your budgets for platforms that offer transparent performance data.

Fourth, watch for consolidation opportunities. As networks merge or offer cross-platform buying, efficiency may improve. Stay informed about partnerships and infrastructure developments that could simplify multi-network campaigns.

Fifth, plan for a concentrated future. Amazon and Walmart will likely remain dominant. Build your retail media strategy accepting this reality rather than hoping for a more distributed market.

The 84% problem isn’t going away soon. The forces driving concentration—measurement challenges, scale advantages, data depth, infrastructure costs—are structural rather than temporary. Brands that understand and adapt to this reality will navigate retail media more effectively than those hoping for a different market structure.