Amazon's advertising business generated $56.2 billion in 2024. That figure exceeds the total revenue of most retailers. Twelve years earlier, Amazon launched its first advertising services as an experiment to help marketplace sellers gain visibility.

The trajectory from experiment to industry explains why retail media networks now attract more than $58 billion in annual US ad spending, growing at 20% annually while overall advertising grows at 4%. Understanding how retail media evolved from Amazon's 2012 launch through today's fragmented ecosystem reveals both the opportunity and the infrastructure challenges that define the category.

What is a retail media network

A retail media network is an advertising platform operated by a retailer, enabling brands to reach shoppers using the retailer's first-party data and owned media inventory. Retailers know what customers search for, consider, and purchase. This makes their advertising uniquely accountable through closed-loop measurement that connects ad exposure directly to transactions.

Unlike traditional digital advertising where conversion tracking requires complex attribution, retail media operates within a closed system. A customer sees a sponsored product on Walmart.com, adds it to their cart, and completes the purchase. The retailer observes the entire journey and can attribute the sale to the ad impression with certainty.

The Amazon origin story (2012-2015)

2012: Three services launch the category

In 2012, Amazon introduced three interconnected advertising services: Amazon Media Group (AMG), Amazon Advertising Platform (AAP), and Amazon Marketing Services (AMS). These services gave marketplace sellers the ability to promote products and compete for visibility against thousands of competing vendors selling identical or similar items.

The initial focus was banner advertisements and search result highlighting within Amazon's own properties. But Amazon had something no other advertising platform could match: data showing exactly what customers searched for, browsed, added to carts, and purchased.

‍

This closed-loop advantage proved decisive. Advertisers could measure return on ad spend with precision impossible on search or social platforms.

2013: Off-site expansion

After demonstrating viability on its own properties, Amazon recognized that its purchase data had value beyond Amazon.com. In 2013, the company launched place-targeted advertising campaigns on external websites, using its first-party data to reach potential customers across the broader internet.

This expansion established a pattern that would define retail media's evolution: onsite advertising within the retailer's owned properties, followed by offsite advertising using retail data to target customers across third-party publishers.

2014-2015: Infrastructure investment

The 2014-2015 period appeared quiet on the surface. Amazon's advertising revenue growth seemed stagnant. But the company was investing heavily in technological infrastructure: more sophisticated keyword analysis, machine learning for targeting optimization, and expanded advertising surfaces.

Amazon also introduced the Echo in November 2014, opening entirely new channels for retail media through voice-enabled devices. By 2015, Amazon's advertising business crossed $1 billion in annual revenue, validating that e-commerce platforms could monetize user attention as effectively as search engines or social networks.

The scale era (2016-2019)

Sponsored products transform the model

In September 2016, Amazon launched its dedicated Advertising Blog, signaling organizational commitment to advertising as a growth priority. More significantly, Amazon introduced sponsored products designed to help sellers boost visibility during peak shopping periods.

Sponsored products proved transformative. The format created a direct connection between product discovery and advertising placement, enabling sellers to promote specific merchandise within search results and category pages. By 2017, Amazon expanded to personalized display advertisements (driving 23% higher purchase rates) and video advertisements.

2018: Consolidation and $10 billion

In 2018, Amazon consolidated its fragmented advertising services into a single entity: Amazon Advertising. The separate brands for AMG, AMS, and AAP were retired. Headline Search became Sponsored Brands. The Amazon Advertising Platform became Amazon DSP.

The financial results validated the reorganization. Amazon's full-year advertising revenue reached $10.1 billion, establishing the company as the third-largest digital advertising platform globally behind Google and Facebook. This was achieved in just six years.

Amazon's revenue trajectory (2015-2019)

Other retailers take notice

Amazon's success forced competitors to evaluate their own advertising potential. Walmart, Target, and Kroger possessed similar assets: massive customer traffic, detailed purchase data, and owned digital properties. The critical variable was organizational ability and technical competence to build sophisticated platforms.

Walmart had been developing advertising capabilities throughout the late 2010s, though formal repositioning came later. Target initiated digital advertising capabilities in 2007 but operated at modest scale. Kroger launched Kroger Precision Marketing in 2017, powered by its data analytics subsidiary 84.51°.

COVID acceleration (2020-2021)

The pandemic compresses digital transformation

The COVID-19 pandemic unexpectedly accelerated retail media adoption, compressing years of digital transformation into months. As lockdown measures closed physical stores and consumers shifted shopping online, traffic flowing through e-commerce platforms and their advertising opportunities expanded dramatically.

Online retail sales across seven major economies increased from approximately $2 trillion in 2019 to $2.5 trillion in 2020, then $2.9 trillion in 2021. The share of online sales in total retail jumped from 16% to 19%. Consumers who discovered online shopping convenience maintained these habits even after restrictions lifted.

Instacart's rapid emergence

The pandemic particularly accelerated grocery delivery. Instacart, founded in 2012 but relatively niche until 2020, experienced explosive growth as lockdown-bound consumers sought alternatives to in-store shopping. US adults reporting increased online grocery shopping jumped from 11% in early March 2020 to 37% by late March.

Instacart launched a self-serve advertising platform in May 2020 to monetize this surge. Major CPG brands including Procter & Gamble, Unilever, and PepsiCo tested the platform immediately. Instacart's advertising business generated $300 million in 2020 despite launching mid-year.

Walmart Connect arrives

On January 28, 2021, Walmart officially announced its strategic repositioning as Walmart Connect. The announcement communicated Walmart's commitment to building an advertising business leveraging the company's unique omnichannel capabilities.

Walmart Connect focused on three pillars:

1. Digital properties: Holistic campaigns combining search and display across Walmart.com, pickup/delivery, and the mobile app
2. In-store experiences: Media activations on approximately 170,000 digital screens across 4,500 stores
3. Offsite media: A partnership with The Trade Desk to leverage Walmart's shopper data across the broader internet

The in-store capability represented a distinctive advantage. Pure-play e-commerce platforms like Amazon lacked comparable physical retail infrastructure. Walmart could deliver advertising experiences spanning digital and physical environments that Amazon simply could not replicate.

Network proliferation begins

The 2020-2021 period witnessed numerous additional retail media network launches:

The pandemic demonstrated that any retailer with substantial customer traffic, loyalty program data, and digital properties could establish a retail media network capable of generating meaningful revenue.

The fragmentation era (2022-2024)

200+ networks emerge

By 2024, the retail media ecosystem encompassed over 200 distinct networks ranging from massive platforms operated by Amazon and Walmart to specialized networks serving specific customer segments. This proliferation reflected proven profitability and relatively low barriers to entry for retailers possessing customer data and digital properties.

The diversification extended beyond traditional retail. Financial institutions including JPMorgan Chase and PayPal launched commerce media platforms, recognizing that their transaction data possessed advertising value. Specialty food delivery companies, quick commerce platforms, and convenience store chains established retail media offerings. "Retail media" was evolving into "commerce media."

Concentration despite fragmentation

Despite the proliferation, market share concentrated heavily among leaders. By 2024:

Amazon and Walmart collectively commanded approximately 84% of US retail media spending, leaving the remaining 16% fragmented across 200+ networks competing for marginal advertiser budgets.

Measurement chaos emerges

As networks proliferated, a critical gap emerged: the absence of standardized measurement methodologies across platforms. Different networks employed distinct attribution models, conversion definitions, and performance metrics, making meaningful comparison across networks impossible.

A typical CPG brand might operate campaigns simultaneously across Amazon, Walmart Connect, Target Roundel, Kroger Precision Marketing, Instacart, and numerous smaller networks. Each platform reported results differently. Consolidating performance data to assess overall return on investment required significant manual reconciliation.

The Interactive Advertising Bureau (IAB) and Media Rating Council (MRC) initiated standardization efforts. In September 2023, they released draft retail media measurement guidelines. The final guidelines arrived in January 2024, establishing consensus on impression definitions, viewability standards, and incrementality measurement approaches.

Adoption remained uneven. Major platforms committed to implementing standards, but many smaller networks proceeded independently. Complete standardization would require extended implementation periods.

The advertiser experience

By 2024, approximately 85% of CPG brands allocated budgets across four or more retail media networks. Portfolio diversification had become standard practice despite the complexity this fragmentation created.

Advertisers reported consistent challenges:

Many brands relied on platform-provided attribution metrics rather than independent third-party measurement, recognizing that networks possessed superior data but potentially creating measurement bias.

The current state (2025)

Market scale and growth trajectory

US retail media advertising reached approximately $58.79 billion in 2025, with projections of $69.33 billion in 2026 and $100 billion by 2028. Retail media is growing at 20% annually compared to 4.3% for the overall advertising market.

Three-quarters of advertisers plan to increase retail media spending, with nearly one-third targeting double-digit increases. The channel has established itself as the third wave of digital advertising, following search and social.

The agentic commerce acceleration

A new variable entered the equation in 2025: AI shopping agents. According to recent agentic commerce research, 39% of consumers now use AI for product discovery, 23% of Americans bought something via AI in the past month, and AI traffic to US retail sites increased 805% year-over-year on Black Friday 2025.

Bain projects 15-25% of e-commerce will flow through agentic channels by 2030. McKinsey estimates the category could reach $5 trillion globally.

This raises a difficult question for retail media networks: what happens when AI agents bypass traditional advertising surfaces entirely? Agents do not see sponsored product placements. They make decisions based on structured data, real-time availability, and API responses. The measurement and identity infrastructure that RMNs have struggled to standardize becomes even more critical when agent transactions arrive without the customer journey context retailers are accustomed to.

Leader performance

Amazon reported $56.21 billion in global advertising revenue during 2024, representing approximately 20% year-over-year growth. Growth rates have moderated from earlier years but remain robust.

Walmart Connect demonstrated stronger momentum, generating $4.4 billion globally during fiscal 2024 with 27% year-over-year growth. The US business grew 31% in the most recent quarter, and advertising grew at six times the rate of overall retail sales.

Target Roundel generated approximately $2 billion in annual value by 2024-2025 with 20% growth. The company projects Roundel's value will double over the next five years.

Kroger Precision Marketing serves thousands of brands using shopper data from over 60 million households. In 2024, Kroger unified retail media, consumer insights, and loyalty marketing under KPM leadership.

Instacart Ads reached $1.18 billion in 2024 with 25.5% growth, projecting $1.45 billion in 2025.

Offsite expansion accelerates

Offsite retail media, where retailers use their data to target customers across external publishers, represents the fastest-growing segment. Offsite spending is projected to grow 42.1% in 2025 compared to 15.3% for onsite. By 2027, US offsite advertising could reach $29 billion.

This expansion brings retail media into competition with established digital channels. Brands can now allocate budgets across social, search, connected television, and retail media within unified campaigns leveraging consistent audience definitions derived from purchase data.

Where infrastructure struggles to keep pace

The data quality crisis

The infrastructure supporting retail media networks has not kept pace with demand. 42% of customers abandon purchases due to insufficient product information. Businesses lose an average of $15 million annually to poor data quality. These problems existed before agentic commerce, but agents amplify them: AI systems inherit data quality evaluation criteria and apply them at scale.

The data fragmentation problem

The rapid proliferation of retail media networks has outpaced the infrastructure required to operate across them efficiently. Brands managing campaigns across 10+ networks face a fundamental challenge: each platform is a data silo with distinct identifiers, attribution models, and reporting formats.

Connecting performance data across networks to understand true incrementality requires:

• Separate data pipelines extracting metrics from each platform
• Manual reconciliation when attribution windows and conversion definitions differ
• Significant analytical resources to normalize and compare results
• Technology investments that disadvantage smaller brands

The IAB standards address some of these issues, but implementation varies. What's measured, how it's measured, and when it's reported remain inconsistent across networks.

Identity resolution across surfaces

Retail media networks increasingly span digital and physical surfaces: website, mobile app, in-store screens, connected TV, and offsite publisher inventory. Measuring advertising impact across these surfaces requires identity resolution that most networks have not fully solved.

A customer sees a sponsored product on their phone, researches on desktop, and purchases in-store three days later. Connecting these touchpoints to a single customer journey, and attributing the sale to the original mobile impression, requires identity infrastructure that persists across devices and channels.

Safari's Intelligent Tracking Prevention limits client-side cookies to seven days. Cross-device identification through cookies alone fails when customers take longer than a week between impression and purchase. The networks that can resolve identity across surfaces and persist it beyond browser limitations will deliver superior attribution accuracy.

Real-time data requirements

Retail media optimization depends on fresh data: current inventory levels, real-time pricing, product availability by location. When a network promotes a product that's out of stock or displays the wrong price, the result is wasted ad spend and poor customer experience.

First-mile data infrastructure, the systems that capture and route signals from the point of customer interaction, determines whether downstream advertising systems receive accurate, timely information. Many networks built advertising capabilities on top of legacy data infrastructure designed for different purposes. The result is latency: inventory changes that take hours to propagate, pricing updates that lag behind reality.

The measurement credibility gap

Retailers report that proving ROI to advertisers represents a significant challenge. Brands face parallel struggles assessing incremental impact versus baseline purchasing patterns. Many campaigns reach customers who would have purchased anyway, making true incrementality difficult to isolate.

Independent third-party measurement is emerging as a priority, but most brands lack the analytical infrastructure to implement sophisticated econometric models or experimental designs. The networks that can demonstrate true incrementality, not just last-touch attribution, will capture disproportionate advertiser investment.

The infrastructure opportunity

Retail media works. First-party data, high-intent audiences, and closed-loop measurement create advertising that outperforms traditional digital channels. The market will reach $100 billion by 2028.

But the infrastructure supporting these networks has not kept pace with demand. Data fragmentation across 200+ networks creates operational complexity. Identity resolution across surfaces remains unsolved for most platforms. Real-time data sync between commerce systems and advertising platforms introduces latency that degrades campaign performance.

The agentic commerce era compounds these challenges. Research on agentic commerce trends shows 4.4x higher conversion rates for AI-generated product recommendations versus traditional search, but only for merchants whose infrastructure can support agent requirements. Networks that cannot prove influence on agent recommendations will lose advertiser confidence as agents capture an increasing share of commerce.

The next phase of retail media growth depends on infrastructure that can:

• Capture customer signals at the point of interaction with consistent data quality
• Resolve identity across devices, browsers, physical locations, and agent surfaces
• Persist identity beyond browser limitations (Safari's 7-day cap, ad blockers)
• Sync inventory, pricing, and availability in real-time to advertising and agent systems
• Enable cross-network measurement and attribution across traditional and agentic channels

The retailers and infrastructure providers that solve these challenges will define retail media's next decade.

‍

MetaRouter provides first-mile data infrastructure for retail media networks. Server-side collection captures customer signals with 40% more data than client-side alternatives, while Sync Injector extends identity persistence to 365 days in Safari. For RMNs seeking infrastructure that matches their advertising ambitions, see how MetaRouter works.

‍