In the last few weeks, Amazon embedded an AI shopping assistant into its search bar, OpenAI launched product feed ads inside ChatGPT, and the largest AI platforms in the world signaled that retail media is no longer a category they plan to watch from the outside.
Amazon’s replacement of Rufus with Alexa for Shopping put AI directly into the highest-intent real estate on its site. OpenAI followed with product feed ads that allow retailers to generate ChatGPT campaigns directly from their product catalogs, alongside a self-serve Ads Manager with cost-per-click bidding and new measurement tools. OpenAI’s ad pilot has already reached $100M in annualized revenue, with Williams-Sonoma, Target and The Knot among the early testers.
What these announcements collectively reveal is that AI is no longer an experimental layer on top of the shopping experience — it is becoming the shopping experience itself, and the most valuable real estate in retail is shifting in ways the existing retail media playbook was not designed for.
The Transaction Moment is Where This is Being Decided
For the last decade, retail media’s center of gravity has been discovery: sponsored listings, on-site banners, search placements designed to capture intent earlier in the funnel. That strategy fit the era of unlimited inventory expansion, but it is not the strategy that will define what comes next.
What Amazon is demonstrating, and what OpenAI is focusing on from a different angle, is a working model for what retail media looks like when AI sits at the moment of purchase rather than upstream of it. Amazon’s competitive advantage is rooted in the depth of its first-party data — what customers bought, what they viewed, what similar customers purchased, where they ship items and how their accounts are configured. That data is most valuable in the seconds before, during and after a transaction actually happens, where intent is no longer a signal but a fact.
The OpenAI moves reinforce the point. OpenAI is serving ads based on what users are actively asking in conversation rather than signals from search behavior or marketplace browsing, and is moving away from taking a cut of transactions toward capturing the ad budgets that ecommerce brands already spend trying to reach shoppers. The largest AI players in the world are converging on the same conclusion: the highest-value real estate in retail is no longer where the shopper begins, but where the shopper decides.
The Strategic Questions for Other Networks
For retailers building or running their own retail media networks, this collective moment surfaces three questions that are difficult but unavoidable.
The first concerns data position. Amazon’s advantage is rooted in the depth of its first-party data and the breadth of its catalog, and most retailers cannot match either. The strategic response is not to attempt replication but to be clear-eyed about what the network’s own data position genuinely supports — and to focus that data where it is most valuable, at the Transaction Moment when the right offer can deepen the customer relationship rather than interrupt it.
The second concerns reach. Any single retailer’s network can only reach the shopper while they are inside that retailer’s environment, and the networks positioning themselves best for the next phase are those building ecosystem relationships that extend their first-party data and shopper intelligence into Transaction Moments across other commerce environments. In an AI-mediated shopping world, the shopper journey crosses many surfaces, and the value of a network is increasingly defined by how far its intelligence can travel.
The third concerns advertiser trust. The third-party sellers reacting to Amazon’s move are voicing a concern that has been building across the broader retail media industry, as recent Harvard Business Review research found that suppliers have begun to view their investment less as a strategic lever and more as a toll they are compelled to pay.
The networks that come through this next phase strongest will be the ones which have advertiser relationships durable enough to absorb significant change without breaking trust — and with performance that can be measured at the level that actually matters, which is the transaction itself.
What Comes Next
One reading of this moment is that the largest AI players are pulling further ahead. A more useful reading is that the AI shopping race validates a thesis building underneath retail media for some time: the value of a network is determined by its ability to deliver relevance at the moment of transaction, not by the volume of impressions surrounding it.
The retail media networks that can clarify their position with specificity, deepen the relationships they have built and double down on the part of the funnel where their differentiation is most defensible will not lose ground to the AI shopping race. They will use this moment to define the next chapter of the category.Â
As SVP of Retail Partnerships, Laura Cosgrove leads the strategy and execution of Rokt’s most significant retail clients across North America and Europe. Since joining the team in 2019, she has held pivotal leadership roles, helping evolve Rokt from a fast-growing startup into a global leader in ecommerce technology. Cosgrove partners with executive leaders at global giants like Macy’s, Best Buy and Fanatics to revolutionize how they engage customers during the checkout journey. By leveraging Rokt’s machine learning capabilities, she enables retailers to turn every transaction into a relevant, high-value interaction. Prior to Rokt, Cosgrove honed her strategic acumen in the broadcast media and investment sectors, building a foundation in media-for-equity and strategic brand positioning.
