The US retail media market is projected to reach $69.33 billion in 2026, outpacing the growth of the broader digital advertising sector. Yet this expansion is running headlong into a disruptive force that threatens to fundamentally alter how consumers discover and purchase products: agentic commerce.

Agentic commerce the practice of AI agents performing end-to-end shopping tasks on behalf of consumers, from research to transaction is no longer a futuristic concept. One-quarter of shoppers are expected to use AI chatbots for purchasing decisions in 2026, and nearly a third of US consumers are now open to AI making purchases for them entirely.

For retail media networks that depend on consumers browsing product pages, clicking sponsored listings, and navigating traditional shopping funnels, this shift represents an existential challenge.

The traffic diversion problem

Retail media networks (RMNs) the advertising platforms operated by retailers like Amazon, Walmart, Target, and Instacart — generate revenue primarily from sponsored product placements that appear when consumers search for or browse products on their platforms.

The economic model relies on a simple equation: more shoppers on the platform equals more ad impressions, which equals more revenue. AI agents threaten to break this equation by intercepting the shopping journey before consumers ever reach a retail website.

When a consumer asks an AI assistant to “find the best noise-cancelling headphones under $150,” the AI agent researches options, compares prices, and presents a recommendation — or completes the purchase directly — without the consumer visiting Amazon, Best Buy, or any other retailer’s website. The sponsored product listings on those platforms never get seen.

Google’s Universal Commerce Protocol, launched within AI Mode, already allows complete purchases within the chat interface without visiting external websites, with early integrations live for Etsy and Wayfair.

Amazon’s AI-powered counter-strategy

Amazon, the dominant force in retail media with the largest share of US ad revenue, is responding aggressively with its own AI infrastructure.

Amazon’s AI shopping assistant, Rufus, is projected to generate over $12 billion in incremental sales annually. Rather than fighting the AI agent trend, Amazon is embedding AI directly into its own shopping experience, attempting to ensure that AI-assisted purchases still happen within the Amazon ecosystem.

The strategy extends to Amazon Ads, which has identified agentic AI and retail media as its two primary growth vectors for 2026. By integrating AI into its advertising products, Amazon aims to create an AI-curated marketplace where product data quality and semantic attributes — rather than traditional keyword bidding — determine advertising visibility.

This approach requires brands to fundamentally rethink their Amazon advertising strategies. Clear, structured product data and consistent attributes are becoming more critical than traditional keyword-focused campaigns.

The measurement crisis

Beyond the traffic diversion threat, retail media networks face a persistent measurement problem that AI disruption is amplifying.

The sector continues to struggle with fragmented measurement standards, inconsistent attribution methodologies, and rising costs. Different retail media platforms use different metrics, making cross-platform comparison nearly impossible for brands running campaigns across multiple retailers.

AI-driven shopping introduces an additional layer of complexity. When an AI agent recommends a product based on a combination of product data, reviews, and contextual signals, how should that interaction be attributed? If the consumer never saw a sponsored listing but purchased a product that appeared in an AI-curated recommendation, does the retail media network deserve credit?

These questions remain largely unanswered, creating uncertainty for the brands investing billions in retail media budgets.

What brands and retailers must do

The retailers and brands that will thrive in the agentic commerce era share several strategic imperatives:

• Invest in product data infrastructure. AI agents make recommendations based on product attributes, reviews, and structured data — not banner ads. Brands with clean, comprehensive product data will be favoured by AI recommendation engines over those relying solely on paid placements.
• Adopt semantic modelling. Move beyond keyword-based advertising to semantic product descriptions that AI agents can parse and compare. This is the product advertising equivalent of the Answer Engine Optimisation shift happening in content marketing.
• Build unified data platforms. Retailers operating their own media networks must provide transparent, standardised measurement that allows brands to evaluate performance consistently across platforms. Those that fail to do so will lose advertiser budgets to platforms with better attribution.
• Experiment with conversational commerce formats. The advertising formats that will dominate the next decade are conversational, contextual, and integrated into AI interfaces. Brands testing these formats now — across Google AI Mode, ChatGPT, and emerging platforms — will have a significant head start.