August 19, 2026
Not all AI agents work in the same way. “Some are built by the technology companies that sit between you and your customer, and those companies have their own commercial interests that don’t always align with yours or the customer’s.”
The agents most relevant here aren’t the ones you build for yourself; those are more like advocates, with a clear mandate and a bounded scope. The commercially ambiguous ones are built by the platforms between you and your customers, such as Google’s shopping agents, ChatGPT in discovery and Amazon’s Alexa for Shopping. They claim to represent the customer. But these platforms have investors who expect returns and advertisers who pay for placement – paying the platform, not the retailer whose products the agent is selecting between.
Google Shopping makes the conflict visible rather than hidden, which makes it the clearest illustration available. Its agentic checkout runs on the same surface – AI Mode in Search – that carries Direct Offers: paid placements shown to shoppers ready to buy. Organic recommendations and paid placements coexist in the same interface, governed by the same algorithm, with no clear signal to the customer about which is which. The agent that claims to find the best product is running on infrastructure that’s also selling shelf space to the products it recommends. By April 2026, ads were appearing in roughly a quarter of AI Mode results. None of this is necessarily malicious – it’s the structural condition of any agent built by a platform whose primary business is advertising.
This is why a protocol integration with a platform agent should be treated as a commercial relationship, not a plug-in. The book reduces it to a single accountability question:
“Answer the accountability question, inside and out. For every platform agent your products are accessible through, ask what data it accesses, how it ranks your products, and who is accountable when something goes wrong. Then ask the internal version of the same question: when an agent acting on a customer’s behalf initiates a transaction and something goes wrong, who does the customer call, who authorises the refund, and who bears the cost?”
The commercial apparatus of retail has always carried some version of this tension – advertising, promotions, shelf placement. The difference in an agentic world is that the tension is hidden. Name it, ask the three questions before you integrate, and you keep your eyes open about whose interests the agent is really serving.
The full book maps all agent types in detail, how platform agents really rank you, and how to stay the merchant of record when the front door is no longer yours.
Paul Sims is a Retail Technology Strategy Consultant, who recently joined Equal Experts after spending the last 20+ years working for retailers across the globe, including positions as CTO at Halfords and Chief Architect at New Look, Primark, Marks & Spencer, and Argos in Shanghai. Connect with Paul on LinkedIn.
All business models have their pros and cons. But, when you consider the type of problems we help our clients to solve at Equal Experts, it’s worth thinking about the level of experience and the best consultancy approach to solve them.
If you’d like to find out more about working with us – get in touch. We’d love to hear from you.
