AI Shopping Agents Threaten $38B Retail Media Market

Something quietly radical is happening to how people shop online. Instead of going to Walmart's website or browsing Target's app, a growing number of consumers are simply asking ChatGPT or Google's Gemini to find them the best product. AI does the searching. AI makes the recommendation. In some cases, AI even places the order.

For the companies that built billion-dollar businesses selling ads on their retail websites, this is not a minor trend to monitor. It's a direct attack on their revenue model.

The Business Model Under Pressure

Retail media networks are advertising businesses that sit inside large retailers. Walmart has one called Walmart Connect. Target has Roundel. Macy's runs the Macy's Media Network. The basic idea: brands pay to show ads to shoppers already browsing a retailer's website, where purchase intent is high and conversions are strong. These networks carry margins of 70%–80%, making them among the most profitable parts of modern retail.

The whole model depends on shoppers visiting retailer websites. US retail media search ad spending hit nearly US$38 billion in 2025, with search ads representing roughly 60% of total retail media revenue. That's the revenue stream most exposed to what's happening now.

According to Adobe's 2026 AI and Digital Trends Report, 55% of consumers now use AI for shopping research, 47% for product recommendations, and 43% to find deals. Meanwhile, Gartner predicts traditional search engine volume will fall 25% in 2026. Fewer searches on retailer sites means fewer ad impressions to sell.

Tyler Murray, Chief Enterprise Solutions Officer at VML North America, put it bluntly: "AI agents really become not just a threat to retail media, but truly an existential threat."

Two Camps, Two Survival Strategies

The retail media industry has split into two responses.

Why AI Ad Agents Are Collapsing the Economics of Video Production

AI ad agents now produce ads for $5-$8 in minutes, collapsing traditional production costs. APAC marketers must accelerate AI-native creative workflows to stay competitive.

Mission MediaMission Media

The first group is following consumers to where they are now spending attention: AI platforms. Target made headlines in February 2026 when its Roundel network launched contextual ads directly inside ChatGPT, becoming one of the first major retailers to advertise within a large language model at scale. The ads appear labeled and separately from ChatGPT's responses, triggered by the content of the shopping conversation. ChatGPT-driven traffic to Target's website is already growing 40% month-on-month.

Walmart Connect is expanding aggressively off its own platform, rolling out self-service retail media campaigns on Meta with TikTok coming next, plus beta-testing an "Add to Cart" feature across Meta, TikTok, and Pinterest. Off-site inventory is growing at more than double the rate of on-site. Macy's took the more surprising route, partnering with Amazon's advertising technology in what one analyst called "coopetition," trading some independence for access to Amazon's measurement infrastructure.

The second group is doubling down on what AI agents cannot touch: physical stores. In-store retail media, from shelf displays to checkout screen ads, exists in a space no AI shopping assistant can disintermediate. Albertsons Media Collective launched a formal measurement program for in-store campaigns in early 2026, with a Mondelez pilot delivering US$2.41 in incremental return for every ad dollar spent and a 14% in-store sales lift across 116 locations.

The Attribution Crisis No One Has Solved

Both strategies run into the same problem: measurement is breaking down.

86% of commerce media decision-makers rate strengthening attribution as a high or critical priority, according to Forrester's State of Retail Media 2025. In a traditional digital advertising environment, tracking a customer from ad exposure to purchase is imperfect but workable. In an AI-mediated shopping journey, it nearly collapses. A consumer might see a ChatGPT-surfaced product recommendation on Monday, ask follow-up questions on Wednesday, and purchase on Friday from what looks like direct website traffic in the analytics. The brand gets no credit for the AI touchpoint that drove the decision.

This is the deeper risk inside the shift. Even if retail media networks successfully place ads inside LLM platforms, they may lack the ability to prove those ads work. Without proof, advertisers pull back.

What This Means for APAC

For marketing leaders in Asia, the timing creates a specific dilemma. Southeast Asia retail media networks are projected to reach US$4.7 billion by 2030, growing 11% annually. Remarkably, 99% of Asia marketers plan to increase their retail media network spending in the next 12 months. Budgets are still flowing toward the category even as its foundations are being questioned in Western markets.

AI Overviews Now Capture 58% of Clicks from Top Google Rankings

Google's AI Overviews suppress 58% of organic search clicks—up from 34.5% eight months ago. CMOs must move beyond traditional ranking metrics.

Mission MediaMission Media

The window for strategic repositioning may be shorter than it appears. Deloitte Asia Pacific reports only 29% of Asia Pacific consumer businesses have adopted agentic AI today, but projects that figure will reach 76% within two years. The disruption that took hold in the US first is heading to APAC on a compressed timeline.

The brands and agencies that treat retail media as a stable, high-margin channel may find they are allocating budgets based on a model that the next wave of consumer behavior will render obsolete. As Deloitte put it: "In 2026, many purchases will begin as suggestions rather than responses to searches."