TechmadeazyA buyer asks ChatGPT for the best running shoes for flat feet and gets three names. Your brand is not one of them. Here is what goes into that answer, and the parts you can change.
Two ways an assistant knows about you
Assistants draw on two sources. The first is what the model learned during training, a broad picture of brands, categories and reputations built from a large snapshot of the web. The second is live retrieval: many answers today are grounded in pages the assistant looks up at the moment the question is asked, often shown as citations. Training shapes the assistant's general sense of your brand. Retrieval decides whether your pages, and pages about you, get pulled into a specific answer.
You cannot edit a model's training data, but you can influence both over time, because what gets published and cited today feeds tomorrow's answers and future training.
What tends to get a brand named
1. The assistant is confident about who you are
If your website, marketplace listings, social profiles and directory entries describe you differently, the assistant has conflicting evidence. Consistent names, categories, product facts and claims make you easier to recommend with confidence.
2. Your content answers the exact question
Assistants favour pages that state an answer plainly: who a product is for, what it does better, how it compares, what it costs. Pages built around a keyword, with the important facts locked in images, give the assistant little to quote.
3. Other sources vouch for you
For product questions, assistants lean on reviews, round-up articles, forums, communities and marketplaces. These are the evidence behind a recommendation. A brand that appears in the sources cited for its category has a much better chance of being named.
4. The assistant can read your site
If AI crawlers are blocked in your robots.txt, or your key facts only appear after scripts run, some assistants may not see your pages at all. This is the quickest problem to rule out.
What does not work
Stuffing pages with the names of AI tools, publishing large volumes of thin AI-written content, or buying listicle placements on sites nobody reads rarely changes answers. Assistants are looking for clear facts and credible corroboration, the same things a careful human researcher would look for.
Where to start
- Write down ten questions a buyer asks before choosing a product in your category.
- Ask them in ChatGPT, Gemini and Perplexity in a logged-out window and note who gets named and which sources are cited.
- Compare those sources with where your brand currently appears.
Our free self-test writes the questions for you, and the EASY framework explains how we work through each of the four areas above.
