Why doesn't ChatGPT mention my brand? Causes and fixes

Which brands do AI assistants mention?

When an AI assistant answers a question such as "which agencies in Istanbul do health tourism marketing?", it draws on two sources. The first is the text the model saw during training; the more often and more consistently your brand was mentioned there, the better the model "knows" you. The second is a live web search carried out while the answer is generated; tools such as ChatGPT's search feature, Perplexity and Gemini scan current pages, pick a few and base the answer on them.

A brand missing from the answer points to a gap in one of these two channels, and often in both. The reasons below are the patterns we meet most often in our work; they are not a published list of algorithm factors but an observation-based diagnostic framework. Most brands show several of them at once rather than just one, so the diagnosis needs to go item by item and the fixes need to be put in order of importance.

Is your brand mentioned enough on other sites?

Generative engines mostly learn about a brand not from its own website but from what others write about it. Industry publications, comparison articles, member lists of professional bodies, event pages, interviews and forum discussions make up this "third-party" layer. However well you describe yourself on your own site, if your name appears nowhere else the model struggles to see you as an option.

What to do:

• Identify the publications and lists frequently cited in your category; these are usually the sites that appear as sources in AI answers.
• Offer these publications content that shows your expertise: analyses that share data, expert commentary, guides.
• Make sure your real activities, such as events, awards, memberships and partnerships, are recorded on the relevant sites.
• Stay away from shortcuts such as bought links or fake reviews; they breach search engine policies and weaken trust signals.

Is your business information consistent across directories and profiles?

A model recognises a brand as an "entity": its name, what it does, where it is and who founded it. If this information is written one way on your Google Business Profile, another way on LinkedIn and a third way in industry directories, the model cannot be sure they all belong to the same company. An old address, a closed branch, a changed phone number or a differently spelt brand name are the most common sources of this confusion.

• Fix your brand's official name, short description, address and contact details in a single reference document.
• Update Google Business Profile, Bing Places, Apple Business Connect, LinkedIn and the key directories in your sector against that document.
• Close or update closed branches and old profiles; merge duplicate listings.
• Make sure the company name appears the same way on the profiles of founders and executives.

Does your website state what you do in clear sentences?

A sentence like "we take your brand into the future with innovative solutions" tells even a human reader little, and tells a language model nothing. Generative engines look for clear definitional sentences to quote: "X is a firm based in this city that provides service Z to companies in sector Y." If your site has no such sentence, the model cannot find the material to match you to a question.

• Define your brand in one sentence on the home page and the about page: who, what, for whom, where.
• In the first paragraph of each service page, state directly what the service is and who it is for.
• Turn the questions your customers actually ask into headings and put the answer immediately below each heading.
• Fill your FAQ section not with generic statements but with the questions you hear in sales conversations.

Do you use structured data?

Structured data (schema.org markup) is a machine-readable copy of the information on the page. Types such as Organization, LocalBusiness, Service, Person and FAQPage define your brand's name, logo, address, social profiles and services in explicit fields. Structured data alone will not put you in an AI answer, but it helps keep your entity information consistent and verifiable.

The point to watch is this: the information in the markup must match the visible text on the page exactly. Adding a service, review or rating to the markup that is not on the page breaches policies and damages trust. Linking your official social profiles and key directory listings through the sameAs property helps pull scattered entity information together into a single identity.

Can crawlers reach your website?

AI tools that search live cannot cite your page if they cannot read it. Technical barriers that are often overlooked include:

• robots.txt rules: Some sites block every AI-related crawler wholesale. Crawlers that collect data for model training and crawlers that fetch pages for search results can be different; OpenAI, for example, defines OAI-SearchBot for search and GPTBot for model training separately. Decide knowingly what you allow and what you block.
• Firewall and content delivery network settings: Some services can block bot traffic by default. Even if robots.txt allows it, the request may never reach your server.
• Client-side rendered content: Pages that build their text only with scripts running in the browser can look empty to tools that do not run scripts. Serving important content as ready-made HTML from the server is the safe route.
• Pages that are not indexed: Many generative search experiences draw on classic search indexes. If your page is not indexed in Google or Bing, it is unlikely to appear in an AI answer either.

Does your category suit AI answers?

Not every question is answered with a list of brands. For questions asking for recommendations, such as "the best" or "which company", models are more inclined to mention brands; for definition or how-to questions they usually answer without naming any. So the first task is to identify the set of questions where it actually makes sense for your brand to be mentioned. If your category is very new or very niche, the model may not yet have enough material; in that case producing third-party content matters even more.

Geography matters too. When the same question is asked in different languages and from different countries, different brands can come to the fore. A business serving customers abroad stays visible only to Turkish-language questions unless it has definitional content in its target markets' languages and mentions in publications read in those markets.

How do you measure a brand's visibility in AI answers?

Improvement without measurement is guesswork. A simple method that works:

1. 1.Build a realistic set of questions your customers might ask, including recommendation, comparison, pricing and location questions.
2. 2.Ask these questions the same way in ChatGPT, Gemini, Perplexity and Copilot at regular intervals.
3. 3.For each answer, record whether your brand is mentioned, in what context, and which sources are shown.
4. 4.Examine the sources shown: on which sites do your competitors appear, and on which are you missing?
5. 5.Track the relationship between your changes and your mentions over time; a single measurement can mislead, because answers vary.

Which service solves which problem?

Each of these causes falls into a different area of work. Our GEO service covers the full effort to be mentioned by generative engines such as ChatGPT, Gemini and Perplexity. Direct answer surfaces such as featured snippets, the "People also ask" box and voice assistants fall under AEO. Visibility in Google's own AI Overviews is the subject of our AIO service. We explain the differences between the concepts in detail in our article on SEO, AEO, GEO, AIO and GEM.