The prevailing wisdom, just months ago, was that retail media networks (RMNs) were an unstoppable force. Their integration into the shopper journey, driven by first-party data and the promise of precise targeting, seemed like a strong defense against the looming cookieless future. Retailers from Walmart to Instacart were not just selling products; they were building sophisticated advertising businesses, carving out a lucrative slice of the ad tech pie. Everyone expected RMNs to mature, expand into off-site channels, and become a cornerstone of brand advertising.
But here’s the thing: the ground has shifted beneath our feet, and the tectonic plates of e-commerce are grinding in a new direction. The real disruption isn’t coming from another privacy regulation or a clever new targeting algorithm. It’s coming from artificial intelligence, specifically from the emergence of agentic commerce – where AI agents actively shop for consumers.
Digiday executive editor Tim Peterson, a man who openly admits he’s “very close to actually having AI do my grocery shopping,” articulated the core problem. His own agentic meal planning system, powered by tools like Notion and Claude, illustrates the emerging paradigm. This isn’t just about getting inspiration; it’s about delegation. When AI becomes the first stop for inspiration, planning, and even execution of shopping tasks, where does that leave the established retail media networks?
“If you’ve got more people that are starting their search on ChatGPT, Gemini, Claude and these other LLMs, what then is the case for a retail media network,” said Kimeko McCoy, Digiday senior marketing reporter.
This is the existential question facing RMNs today. For years, these networks thrived because retailers drove search traffic directly to their own sites. This on-site traffic provided valuable inventory for advertisers. But if the initial eyeballs are no longer on Walmart.com or Macys.com, but rather on a Gemini or ChatGPT interface, the entire value proposition begins to fray. The limited on-site ad inventory that RMNs historically relied upon might become a relic.
The push towards off-site channels—like Walmart Connect acquiring Vizio or Instacart partnering with Roku—was a strategic attempt to follow shoppers beyond the retailer’s own digital walls. It was a logical evolution, attempting to capture attention on streaming services and social platforms. Target’s Roundel, in a move that suggests they saw the writing on the wall even before it was fully legible, has reportedly begun promoting its business and partners within LLMs. This is a bold, albeit early, bet on where the future interaction might lie.
The fundamental challenge for RMNs in this new agentic world is performance and attribution. Can an ad placed through an RMN hold its value or demonstrate true performance when the purchase journey is mediated by an AI agent? Peterson himself raised this concern, questioning, “The challenge there is does the value or really the performance of retail media advertising hold up in these environments?”
Agentic commerce compounds an already tricky problem: incrementality. Proving that an ad campaign actually drove new sales rather than just influencing a purchase that would have happened anyway has always been a headache for RMNs and their advertisers. When an AI agent is making the decision based on a multitude of factors — price, availability, user preference, perhaps even the AI’s own learning — disentangling the causal link between an RMN ad and a sale becomes astronomically harder. The AI might fulfill a request without ever explicitly surfacing or prioritizing the RMN’s ad inventory.
This isn’t a minor architectural tweak; it’s a potential paradigm shift. We’re moving from a world where brands pay to reach consumers on specific platforms to one where consumers (via their AI agents) might dictate where and how they engage with advertising, effectively controlling the flow of attention and, by extension, the value of inventory. The data RMNs have meticulously collected becomes less valuable if the AI agent is the one doing the filtering and decisioning, potentially bypassing the RMN’s curated environment.
Brands like Duluth are reportedly trying to navigate this by leveraging AI for bidding while keeping their brand voice human-led. It’s a delicate dance. But the core question remains: Can the infrastructure built around the traditional shopper journey—an infrastructure that underpins the entire RMN model—survive a future where the shopper’s primary interface is an intelligent agent?
What this implies is that RMNs might need to fundamentally rethink their offerings. Instead of focusing solely on on-site or even off-site inventory, they might need to become facilitators for AI agents, embedding their data and ad capabilities directly into the AI’s decision-making processes. This is a far more complex technical and business challenge than simply extending reach to CTV. It requires a deep integration into the AI’s reasoning engine, a space that is still very much up for grabs.
The historical parallel here isn’t entirely unprecedented. Think back to the early days of search engines. Brands that once relied on direct traffic and printed directories had to rapidly adapt to the rise of Google. Those that didn’t invest in understanding and optimizing for search quickly faded. Retail media networks are at a similar inflection point, but the technology at play—generative AI and autonomous agents—is evolving at an exponentially faster pace. The window for adaptation might be much narrower.
Is Agentic Commerce a Real Threat to Retail Media Networks?
Yes, the emergence of agentic commerce presents a significant, potentially existential, threat to the current retail media network model. RMNs are built on the premise of capturing shopper attention and intent on retailer-owned or closely affiliated platforms. Agentic commerce shifts the primary point of engagement for shopping decisions away from these platforms and towards AI interfaces. If AI agents become the default starting point for discovery and purchase, the direct traffic and on-site inventory that fuel RMNs could diminish dramatically. Furthermore, the ability to accurately attribute ad performance and prove incrementality becomes far more complex when an AI intermediary is involved.
How Can Retail Media Networks Adapt?
Adaptation for retail media networks will likely require a fundamental re-architecting of their value proposition. Instead of focusing solely on selling ad space, RMNs may need to explore:
- AI Integration: Developing capabilities to integrate their data and ad offerings directly into AI agent decision-making processes. This means becoming a valuable data source or recommendation engine for the AI itself, rather than just a place to serve ads.
- Performance Verification for Agents: Creating new measurement frameworks that can prove the value of advertising within an agentic context, addressing the incrementality and attribution challenges head-on.
- Data Democratization for AI: Rethinking how their first-party data is structured and exposed to ensure it’s easily consumable and valuable for AI agents to use in making purchase recommendations.
- Partnerships with AI Developers: Collaborating with AI platform creators to ensure RMN data and advertising capabilities are considered in the development of future shopping agents.
Failure to address these shifts could see RMNs become less relevant as consumer behavior moves towards AI-driven shopping experiences.
🧬 Related Insights
- Read more: Incrementality Testing: How to Measure True Ad Campaign Impact
- Read more: AI isn’t coming, it’s here.
Frequently Asked Questions
What is agentic commerce? Agentic commerce refers to a future where artificial intelligence agents autonomously perform shopping tasks on behalf of consumers, from product discovery and comparison to purchasing and reordering.
Will AI replace online shopping websites? AI agents won’t necessarily replace online shopping websites entirely, but they are poised to become the primary interface for many shopping journeys, influencing how and where consumers interact with products and brands.
What does this mean for advertisers? Advertisers will need to rethink their strategies to ensure their products and brands are discoverable and recommended by AI shopping agents. This might involve optimizing product information for AI consumption and exploring new ways to integrate their offerings into AI-driven shopping workflows.
