How to appear in ChatGPT: the short answer
To understand how to appear in ChatGPT, start from how these assistants work: they recommend what they can read, understand and verify. A company improves its chances with a clear page for each service, direct answers to its customers’ real questions, consistent facts across the web and correct crawling permissions. Nobody can guarantee a mention.
How AI assistants decide what to mention
ChatGPT, Gemini, Perplexity and other assistants combine two sources: what they learned during training and what they look up on the web when answering. When they search, they read pages, summarize and cite. That is why generative engine optimization is close to traditional search optimization, with one difference: the assistant does not show ten results, it picks a few and explains them. The content that answers best and with the least ambiguity wins.
One clear page per service
If a company offers five services and describes them all on one generic page, the assistant struggles to know what to recommend it for. Each service needs its own page, with a title that says exactly what it is, who it is for and where it is offered. The words must be the customer’s words, not internal jargon: "WhatsApp chatbot" rather than "omnichannel conversational platform".
Direct answers and FAQs
The first paragraph of each page should answer the main question in two or three sentences, without a warm-up. Detail comes after. A FAQ section with real questions (what it is, how much it costs, how long it takes, what it includes, what it does not do) gives the assistant ready-to-quote passages. It also helps traditional search and, above all, the customer.
AI search visibility: structured data and consistency
AI search visibility also depends on technical signals. Structured data (schema.org markup for organization, service, FAQ and article) states explicitly who the company is and what it offers. Consistency matters too: name, address, field and services should say the same thing on the website, Google Business Profile, LinkedIn, directories and anywhere else the company appears. Contradictions lower the assistant’s confidence.
What llms.txt is and what it is not
The llms.txt file is a convention proposed in 2024: a plain text file at the root of the site that summarizes what the company is and which pages matter. It is useful as a clean index, but it is not an official standard or a guarantee: not every assistant reads it, and none commits to using it. It is worth having, as long as the underlying content is good.
Crawling permissions
Many sites block assistant crawlers without knowing it. Each AI company documents its crawler names; OpenAI, for example, separates OAI-SearchBot, used by its search, from GPTBot, used for training. A robots.txt file can allow one and block the other. It is each company’s decision, but it should be made deliberately and not left to an inherited setting.
Mistakes that reduce visibility in AI assistants
- Hidden content: key information only in images, PDF files or behind logins that crawlers cannot read well.
- Pages without a concrete answer: brand copy about values and innovation that never says what the company does, for whom and where.
- Contradictory facts: different services or addresses on the website, the Google profile and social networks, which make the assistant unsure which information is correct.
- Unsupported claims: statements such as market leader with no evidence. Assistants tend to prefer sources that explain and show proof.
- Slow or broken sites: pages that load slowly or return errors are crawled less often and refreshed less in indexes.
- Duplicate content: many near-identical pages that only swap a city name, which both Google and assistants treat with suspicion.
How to measure it
- Ask the assistants regularly about your field and area, with the phrases customers use, and record whether the company appears and how it is described.
- Check site analytics for visits coming from chatgpt.com, perplexity.ai and other assistants.
- Watch service pages in Search Console, because much of what assistants read comes from the same search results.
We explain how we work on these signals in generative engine optimization. If the company also wants its own assistant that answers with its real information, see AI chatbot for business, and to decide where to start with artificial intelligence in general, AI consulting. In every case the foundation is the same: content that is accurate, specific and easy to verify, because that is what both search engines and assistants end up rewarding.
It is also what convinces the customer once they finally reach the site. A website that explains clearly what the company does, for whom, where and under which conditions works for all three audiences at once: the search engine, the AI assistant and the person deciding whether to get in touch. Treat visibility in AI answers as a consequence of that clarity, not as a separate trick.