Retail Media Wars: Amazon DSP vs Walmart Connect vs Meta in 2026

The three-front war for retail media dollars

Retail media has become the fastest-growing segment in digital advertising. In 2025, brands spent $142 billion on retail media globally. Projections for 2026 push past $180 billion, with the growth rate accelerating as traditional display and social budgets migrate toward commerce-connected channels.

But the landscape is no longer a single-platform game. Three distinct ecosystems now compete for brand budgets, each with fundamentally different capabilities, audience compositions, and performance characteristics. Allocating effectively across Amazon DSP, Walmart Connect, and Meta's commerce suite requires understanding not just what each platform offers, but how they interact—and where the compounding returns hide.

Amazon DSP: The closed-loop powerhouse

Amazon's Demand-Side Platform remains the dominant force in retail media, commanding roughly 37% of all US retail media spend in Q1 2026. Its core advantage hasn't changed: closed-loop attribution from impression to purchase, powered by Amazon's deterministic purchase data on 310 million active customer accounts.

What's new in 2026

Amazon DSP has made three significant advances that change how brands should approach the platform.

Audience modeling with purchase-intent signals. Amazon's new "Commerce Audiences" feature goes beyond demographic and behavioral targeting. It identifies consumers who have demonstrated purchase intent for your product category within the past 7, 14, or 30 days—based on actual search queries, product page views, and add-to-cart actions. Early adopters report 23-31% improvements in ROAS compared to standard in-market audiences.

Cross-device frequency management. Amazon now offers unified frequency capping across Fire TV, mobile, desktop, and Alexa-enabled devices. This eliminates the waste of showing the same ad eight times to the same household across different screens. Brands with multi-device strategies report 15-18% efficiency gains from frequency optimization alone.

Clean room integrations. Amazon Marketing Cloud (AMC) now supports direct first-party data integration through clean room technology. Brands can match their CRM lists against Amazon's shopper base to build lookalike audiences, measure incrementality, and attribute offline-to-online purchasing behavior. This is particularly valuable for brands with significant DTC or retail channels alongside their Amazon presence.

Amazon DSP benchmarks (Q1 2026)

The current performance benchmarks for Amazon DSP across categories:

Average CPM for Sponsored Display: $4.80-$7.20 depending on category. Beauty and personal care index higher at $6.90; home and kitchen lower at $4.40.

ROAS for mid-funnel campaigns: 4.2x-6.8x for established brands with strong organic rankings. New-to-brand campaigns typically deliver 2.1x-3.5x ROAS, but with higher long-term customer value.

Video completion rates on Streaming TV: 94-97% for 15-second spots, 82-88% for 30-second spots. Amazon's non-skippable inventory drives significantly higher completion than YouTube or social video.

When to prioritize Amazon DSP

Amazon DSP should receive the largest share of retail media budget when the primary objective is driving direct Amazon sales, when the brand has an established product catalog with strong organic rankings, when closed-loop attribution is critical for justifying spend, and when the target audience skews toward high-intent shoppers already within the Amazon ecosystem.

Walmart Connect: The undervalued challenger

Walmart Connect has undergone a transformation that many brand marketers haven't fully registered. In 2025, Walmart's advertising revenue grew 28% year-over-year, reaching $4.1 billion. The platform now serves 159 million weekly customers across stores, Walmart.com, and the Walmart app—an audience that differs meaningfully from Amazon's customer base.

The Walmart advantage: Omnichannel attribution

Walmart's most significant competitive advantage is something Amazon cannot replicate at the same scale: in-store purchase attribution. Through its loyalty program and payment data, Walmart can connect digital ad impressions to physical store purchases. For brands where the majority of volume moves through brick-and-mortar retail, this closes an attribution gap that Amazon DSP cannot address.

Walmart's Store Attribution data shows that for CPG brands, in-store sales driven by Walmart Connect advertising are 3.2x the volume of online sales from the same campaigns. Brands that only measure e-commerce ROAS are undervaluing their Walmart Connect investment by more than two-thirds.

What's new in 2026

Self-service DSP access. Walmart has expanded its self-service DSP to brands with as little as $10,000 monthly spend, down from the previous $50,000 minimum. This opens the platform to mid-market brands that previously couldn't access programmatic inventory through Walmart.

Walmart Creator integration. Walmart Connect now offers native creator content amplification, allowing brands to boost creator-produced content (from Walmart Creator) as sponsored placements across Walmart's owned properties. This blends the authenticity of creator content with the targeting precision of retail media.

Search brand amplification. Walmart's equivalent of Amazon's Sponsored Brands, now enhanced with dynamic creative optimization. The platform automatically tests headline variations, product combinations, and imagery to optimize click-through and conversion rates without manual A/B testing.

Walmart Connect benchmarks (Q1 2026)

Average CPC for Sponsored Products: $0.62-$1.15. This remains 35-45% lower than comparable Amazon Sponsored Products CPCs, creating an efficiency arbitrage that many brands haven't fully exploited.

ROAS for omnichannel campaigns (including in-store attribution): 5.8x-9.2x for grocery and CPG. For general merchandise: 3.4x-5.1x.

New customer acquisition cost: 22-30% lower than Amazon for comparable categories, driven by Walmart's stronger penetration in suburban and rural demographics that are less heavily targeted by competing brands.

When to prioritize Walmart Connect

Walmart Connect deserves increased budget allocation when the brand has significant in-store retail distribution at Walmart, when the target demographic includes suburban families, value-conscious households, or grocery buyers, when CPCs on Amazon have reached diminishing-return thresholds in the category, and when omnichannel attribution (digital-to-store) is important for demonstrating marketing ROI.

Meta: The full-funnel commerce engine

Meta's position in the retail media conversation is often misunderstood. It's not a retail media network in the traditional sense—it doesn't own a marketplace. But Meta's commerce capabilities have evolved to the point where separating "social advertising" from "retail media" is increasingly artificial.

Meta's commerce-specific capabilities

Advantage+ Shopping Campaigns (ASC). Meta's AI-driven campaign type now accounts for 62% of all e-commerce advertising spend on the platform. ASC uses machine learning to dynamically allocate budget across audience segments, creative variations, and placements. Brands using ASC report average ROAS of 4.8x—competitive with retail media networks, and often superior for new customer acquisition.

Shops integration and checkout. Products listed in Facebook and Instagram Shops now support native checkout in 31 markets. When a consumer completes a purchase without leaving Instagram, Meta provides full closed-loop attribution—similar to what Amazon DSP offers within its ecosystem. Native checkout conversion rates are 2.3x higher than redirect-to-site flows.

Collaborative Ads with retailers. Meta's Collaborative Ads program allows brands to target consumers and drive them to retailer product pages (Amazon, Walmart, Target) using the retailer's product catalog data. This creates a bridge between Meta's massive audience reach and retail media attribution. Brands can use Meta's targeting to drive Amazon sales and measure the impact through Amazon Attribution or AMC.

Meta benchmarks (Q1 2026)

Average CPM for ASC campaigns: $11.40-$18.60, varying by vertical. Fashion and beauty index higher; home goods and electronics lower.

ROAS for ASC (e-commerce): 3.8x-5.4x average across all advertisers. Top-quartile performers achieve 8x+ through creative optimization and first-party data audiences.

Cost per new customer acquisition: $12-$28 depending on category and AOV. Meta consistently outperforms retail media networks on new customer acquisition efficiency due to its massive reach outside of active shopping environments.

When to prioritize Meta

Meta should receive increased allocation when brand awareness and new customer acquisition are primary objectives, when the product category benefits from visual storytelling and lifestyle positioning, when the brand has strong creative assets (video, UGC, creator content), and when the DTC channel is a meaningful revenue contributor alongside marketplace sales.

The allocation framework

Rather than debating which single platform deserves the budget, successful brands in 2026 are building allocation frameworks that leverage the distinct strengths of each.

The 50/25/25 baseline

For brands with established presence on both Amazon and Walmart, a common starting allocation is 50% Amazon DSP, 25% Walmart Connect, and 25% Meta. This baseline reflects Amazon's market dominance while ensuring sufficient scale on Walmart and Meta to generate meaningful performance data.

Adjusting for business context

DTC-heavy brands should shift toward 35% Amazon DSP, 15% Walmart Connect, and 50% Meta. The rationale: Meta's ability to drive DTC traffic and build brand awareness compounds the value of marketplace presence without over-indexing on retail media.

CPG and grocery brands should shift toward 35% Amazon DSP, 40% Walmart Connect, and 25% Meta. Walmart's omnichannel attribution captures the full value of campaigns that drive in-store purchases.

New-to-market brands should start with 30% Amazon DSP, 20% Walmart Connect, and 50% Meta. Building awareness first, then converting that awareness into marketplace sales, is more capital-efficient than competing on already-expensive marketplace advertising from day one.

The incrementality test

Whatever allocation you choose, run incrementality tests quarterly. Pause spend on one platform for a controlled period and measure whether sales on that platform and others change. The results often reveal that certain platform spend is largely driving cannibalization rather than true incremental revenue.

Amazon and Walmart advertising frequently show high measured ROAS but low incrementality for established products—the sales would have occurred organically. Meta typically shows lower measured ROAS but higher incrementality, particularly for products with lower brand awareness.

Cross-platform measurement

The biggest challenge in multi-platform retail media is measurement. Each platform reports its own ROAS using its own attribution model, and the numbers don't add up to reality when summed.

Unified measurement approaches

Media mix modeling (MMM). Use MMM to understand the true contribution of each platform at a macro level. Modern MMM tools like Meridian (Google's open-source solution) and Meta's Robyn can process weekly spend and revenue data across platforms to estimate each channel's true incremental impact.

Geo-testing. Run matched market tests where you increase or decrease spend on specific platforms in specific geographic regions while holding other regions constant. This provides causal evidence of platform impact that observational data cannot.

Unified tracking parameters. Implement consistent UTM parameters and tracking IDs across all platforms. While Amazon and Walmart have limited external tracking support, Meta's Collaborative Ads and Amazon Attribution can bridge some of the cross-platform measurement gaps.

Looking ahead: The convergence

The retail media landscape is converging. Amazon is building brand advertising capabilities. Walmart is building programmatic sophistication. Meta is building commerce infrastructure. Within 18 months, the functional differences between these platforms will narrow significantly.

The brands that will thrive are those building platform-agnostic creative and data capabilities today. Invest in first-party data infrastructure that feeds all three platforms. Develop creative workflows that produce platform-optimized variations efficiently. Build measurement frameworks that evaluate platforms on incremental contribution rather than self-reported ROAS.

The retail media wars aren't about picking a winner. They're about building a portfolio strategy that captures the unique value each platform offers—and having the measurement discipline to know when to shift allocation as the competitive dynamics evolve.