Why doesn't ChatGPT recommend my brand?

First, the honest caveat: OpenAI does not publish how ChatGPT selects brands, and nobody outside the company can inspect the process. What follows is based on how these systems are built and on the patterns you can observe when you test them systematically.

ChatGPT answers in one of two modes. Without web search, it answers from its training data, a snapshot of the public web collected before a cutoff date. In that mode, brand naming is essentially a prevalence-and-consensus effect. A brand mentioned once, in one place, barely registers. A brand described the same way across hundreds of independent pages becomes something the model is confident enough to state as fact. There is no submission form and no index to appear in. What matters is how often, and how consistently, other people describe you.

With web search on, ChatGPT runs searches (Bing's index has historically done the heavy lifting), reads a shortlist of results, and composes an answer with citations. The question then becomes very concrete: when someone searches "best [your category] UK", which pages come back, and do those pages name you? If the comparison articles and listicles that rank for your category leave you out, the browsing answer leaves you out too.

In practice, missing brands almost always fall into one or more of four buckets. The first is a thin third-party footprint: your own site says excellent things about you, but few independent sites corroborate any of it. Models weight corroboration heavily, because that is what separates a real, established brand from a claim on a landing page.

The second is absence from the places recommendations actually live. "Best of" listicles, review platforms and Reddit threads are precisely the content an AI model learned category recommendations from, and the same pages a browsing model reads. If you have no reviews, no listicle placements and no community mentions, you are invisible in both modes at once.

The third is an ambiguous entity. If your brand name is a common word, or you share it with a bigger company, the model may not be able to confidently attach facts to you, so it plays safe and says nothing. The fourth is simple recency: if you launched, renamed or pivoted after the model's training cutoff, the non-browsing model cannot know you exist, no matter how good your marketing is.

There is one more failure mode worth naming: even when ChatGPT does mention a brand, what it says is often stale or wrong. In our July 2026 test of 30 brands, 27 were described with at least one materially wrong claim when the AI answered without live web search (full study). Being absent is one problem. Being misdescribed is another, and both need watching.

1. Tighten your own entity (days). Write one clear sentence describing what you do and who for, and use it verbatim on your homepage, About page, LinkedIn and every directory listing. Add schema.org Organization markup, keep your name, location and category identical everywhere, and kill any conflicting old descriptions. This is the cheapest fix and the only cure for the ambiguity problem. A model cannot recommend an entity it cannot pin down.

2. Build reviews on the platforms your category uses (weeks). Trustpilot and Google reviews for consumer brands; G2 and Capterra for software. These sites are heavily represented in training corpora and rank well in Bing, so they work in both of ChatGPT's modes. Ask real customers systematically; a steady drip of genuine reviews beats a one-off burst.

3. Get into the listicles that already rank (weeks to months). Search your category query on Bing and note the top ten comparison articles. That shortlist is roughly what a browsing ChatGPT reads. Pitch the publishers honestly: offer updated information, original data, or free access so they can review you properly. One placement in an article that ranks is worth more than ten guest posts nobody finds.

4. Show up in community conversation (months, ongoing). Reddit and niche forums feed both training data and browsing results, but they punish shortcuts. Astroturfed praise gets deleted and can permanently sour how your brand is discussed, and models learn from that. The sustainable version is founders and staff genuinely answering questions in your niche, disclosing who they are, and being useful first.

5. Earn independent coverage with original data (highest effort, biggest compounding). Publish something citable, such as a benchmark, a piece of research or an annual report on your industry, and do the digital PR to get it referenced. Every independent article that describes you accurately strengthens the consensus a model sees, and it keeps paying off in future training runs long after the coverage goes quiet.

Before spending a penny on fixes, get a baseline. Ask ChatGPT your category questions the way a customer would ("best [category] for [use case]", "alternatives to [market leader]") in both modes, with web search on and off. Vary the phrasing, note which brands get named and what is claimed about them, and repeat the exercise monthly, because answers drift as models and indexes update.

If you want the quick version, our free AI visibility checker will show you whether an AI engine recognises your brand and what it says about you. No signup, and it takes about a minute. For the deeper picture of how ChatGPT specifically handles your category, and what ongoing monitoring looks like as our engine coverage rolls out, see the ChatGPT visibility tracker page. Discoverable runs live checks on Claude today, with ChatGPT and other engines rolling out. The manual protocol above works right now and costs nothing.

Set honest expectations first. Browsing-mode answers can change within weeks of landing in the pages that rank, so that is your fastest lever. Training-data presence only updates when a new model ships, so mentions you earn now are an investment that pays out at the next release, not next Tuesday.

This week, do four things. Run the baseline checks above and screenshot the results. Write your one-sentence entity description and roll it out everywhere your brand appears. Search your main category query on Bing and list the top ten results, marking which name you and which don't. That list is your outreach pipeline. And pick the one review platform that matters most in your category, then ask your ten happiest customers to write something honest. It is not glamorous work, but it is the evidence ChatGPT looks for.