--- title: "Why Strong Brands Can Lose When AI Agents Compare Suppliers" slug: "strong-brands-ai-agents-compare-suppliers" url: "https://renascor.pro/insights/strong-brands-ai-agents-compare-suppliers" category: "Strategy" excerpt: "Buyers now ask AI agents such as ChatGPT to pick a few suppliers. The agent goes by what it can find in writing. If what customers value most about you lives in your people, it isn’t there, and neither are you." coverImage: "https://renascor.pro/objects/articles/18/cover/d4d65ecf-dceb-49cf-a535-a09533ec6067" publishedAt: "2026-11-13T18:10:44.556Z" updatedAt: "2026-11-13T20:08:31.162Z" --- ![Cover image]() A buyer looking for a new supplier can now ask an AI agent (an assistant such as ChatGPT, Gemini or Claude that compares options for a person and, in some cases, buys for them) for the best options in a category. The agent comes back with a shortlist of a few names, each with its reasons. When Kearney surveyed US consumers in 2025, [about 60% said they were at least moderately likely](https://www.kearney.com/documents/d/asset-library-291362522/agentic-commerce-retail-disruption-pdf) to try an AI agent for shopping within a year. How fast business buyers will follow is still an open question, but purchasing teams already use agents to request and compare quotes. Your offer has to be readable to get on that shortlist. OpenAI publishes a [product feed specification](https://searchengineland.com/optimizing-chatgpt-shopping-463316) (the format of a file that lists a merchant's products, prices and specifications), and ChatGPT indexes and ranks the data in it. [Shopify passes its merchants' catalogues to ChatGPT](https://lseo.com/generative-engine-optimization/chatgpt-shopping-results-how-merchants-can-improve-discoverability/) the same way. The obvious response is a data project, and if your company has started one, it probably sits with e-commerce or marketing. A clean feed gets your offer read, but it gives the agent no reason to pick you over a cheaper supplier. The feed holds price, specifications, delivery time and ratings. Think of why your best customers stay: perhaps the engineer who turns up on a Saturday, or the account manager who corrects a billing mistake before the customer notices it. Those reasons sit in no field. --- ## Your brand now has two audiences Brands were built for people. A customer chose you because of what working with you was like and what colleagues said about you. An agent has none of that. It ranks you on what it already holds about your brand and on what it can find and check today: public terms, specifications, prices, reviews, independent tests. dunnhumby, which advises retailers on loyalty, warns that agents [optimise for objective criteria](https://www.dunnhumby.com/resources/blog/loyalty-personalisation/en/agentic-commerce-how-to-win-and-retain-customer-loyalty) such as price, delivery speed and specifications, which can make suppliers look interchangeable. Put the two audiences side by side, and a company sits somewhere on two axes: how strongly people choose it, and how much an agent can verify about it. Four positions follow: - *Chosen by both.* Each reason customers give has a proof an agent can check. - *Loved but unseen.* Customers stay for reasons that live in your staff and your contracts, while the agent sees mostly your product data. - *Picked on price.* Complete data and a low price get you on the shortlist, and little keeps the customer once the order is placed. It is a relative of [Efficient Irrelevance](https://renascor.pro/insights/efficient-irrelevance-hidden-failure-mode), where a company runs well in a direction customers no longer choose. - *Out of sight*, for companies that neither people nor agents have much reason to pick. **Strong brands are the exposed ones here, because their advantages were never written down as proof.** Their leadership invested in how customers feel, then gave agent preparation to the team that manages product data. Kearney tells retailers to aim for being ["agent-preferred"](https://www.kearney.com/documents/d/asset-library-291362522/agentic-commerce-retail-disruption-pdf), the supplier algorithms select first. A clean feed won't get a company there. New customers are where the gap shows first, because an existing customer can sign in with you once and let their agent act for them, so the agent arrives with their history and their terms ([how that works](https://renascor.pro/insights/ai-squeezes-customer-relationship-two-sides)). Someone who has never bought from you sends an agent that knows only what is public. In business-to-business markets, purchasing teams now send agents to request quotes from many suppliers at once, and [Monetizely describes](https://www.getmonetizely.com/articles/will-ai-agents-make-b2b-pricing-more-competitive) how that turns occasional buying into continuous comparison. Each quote request starts from zero. What is public, and where, depends on the question the buyer asks. When an AI answer is about a named company, it draws mostly on that company's website and business listings: Yext, which sells listings management, found that [86% of 6.8 million citations](https://searchengineland.com/ai-search-citations-brand-controlled-sources-463166) across ChatGPT, Gemini and Perplexity came from sources brands own or manage. When the buyer asks which supplier to choose, the picture turns around. In a study of purchase-stage questions by the content agency [Siege Media](https://www.siegemedia.com/?p=44129), review sites, comparison pages, publishers and forums supplied 80 to 95% of citations, depending on the industry. A new customer asks the second kind of question, so a proof that exists only on your own terms page may never reach their agent. --- ## Turning a brand promise into proof is an execution job Take two familiar reasons to believe (the evidence a brand offers for its promise): "service you can count on" and "partners, not suppliers". An agent can check neither. It can check "an engineer on site within 24 hours, written into our terms", or a renewal price existing customers can see. > A reason to believe is written for people. An agent needs the proof behind it. ![](https://renascor.pro/objects/articles/18/content/3a58e552-e769-41c6-8a5f-ef8b6405b490) Marketing writes the promise, **but the proof belongs to the teams that deliver it:** operations sets response times, legal writes warranty terms, pricing decides renewal prices, and service handles complaints. Making a brand promise checkable runs into the hand-off every strategy faces, between the people who decide and the people who deliver. And once an agent can check a promise, it has to hold every time, because reviews and complaint records will show the cases where it didn't. Four steps make a reasonable start: 1. Collect the reasons customers give for choosing you, in their own words, from win/loss reviews and from your sales team. 2. Ask two or three agents, such as ChatGPT, Gemini or Claude, to recommend suppliers in your category the way a buyer would, and read what they say about you and your competitors, and which sources they cite. Whether you appear matters less than the reasons they give. 3. Pick the strongest reason that nothing public proves. Write it into your terms as a commitment, and give it to the team that has to deliver it. Then put it where buyers' agents look: in your product data next to the specifications, and in front of the customers who can describe it on the review and comparison sites your category uses. 4. Some lines will lose on price whatever you do. Decide now how low you will go on those, because an agent comparing quotes will find your lowest price anyway. --- ## Your own agents will pull toward the shortlist Marketing teams that run campaigns with AI agents have a second problem, and it sits inside the company. Those agents optimise for what they can measure fast: how often the brand appears in AI answers and shortlists, and how many sales come from agent traffic. Why customers choose you changes slowly and is hard to count. **An agent scored only on the fast numbers will keep improving your product data and your price position, while nobody measures the reasons customers stay.** Over a year that can move a company from chosen by both to picked on price, without anyone deciding it. Write both sides into your own agents’ instructions: what customers value, and what customers’ agents need to find. Then give them at least one measure from the human side that they can’t trade away, such as the reasons in win/loss reviews, and keep the price floor and the list of promises they may make with a named person. ![](https://renascor.pro/objects/articles/18/content/59dc1fdb-1024-4c83-bb0f-917d862f46d5) The executive team has to decide which reasons for choosing you the company will commit to in writing, and who answers for each one. Left open, the ranking comes down to your product data alone. ## Decision Capture What do customers choose us for that we’ve never written down or made public?