AI search
How to appear in ChatGPT answers
ChatGPT names businesses when people ask it for recommendations, comparisons and prices. Nobody can force their way into that answer. You can make your business easy for it to find and safe for it to quote.
How does ChatGPT decide which businesses to name?
ChatGPT answers in one of two ways. Sometimes it replies from what the model learned in training, with no links. For many questions, and especially recommendations, local questions and anything recent, it searches the web first, reads a few pages and writes an answer with cited sources.
The second mode is the one you can work on. OpenAI’s documentation says its search crawler, OAI-SearchBot, is what surfaces websites in ChatGPT’s search features. If that crawler can read your pages, and your pages answer the question clearly, you are among the candidates. Whether you get picked then depends on how well your pages answer and on what other sources say about you.
The answer also shifts with the wording, the language and what ChatGPT knows about the person asking. Two people asking for a good dentist in the same city can see different names. There is no position one to hold. The goal is to be named often, across many phrasings. That is one part of generative engine optimization.
Which OpenAI crawlers should you allow?
OpenAI uses separate user agents for separate jobs, and each robots.txt setting is independent of the others (OpenAI crawler documentation).
| User agent | What it does | What to decide |
|---|---|---|
| OAI-SearchBot | Crawls pages so they can appear in ChatGPT search results | Allow it if you want to be cited. OpenAI says sites that block it are not shown in ChatGPT search answers |
| GPTBot | Collects content that may be used to train OpenAI’s models | Your call. Blocking it does not block OAI-SearchBot |
| ChatGPT-User | Opens a page when a user asks ChatGPT or a custom GPT to | It does not crawl on its own, and OpenAI says robots.txt rules may not apply to it |
Can OpenAI’s crawler actually reach your pages?
Open yoursite.com/robots.txt in a browser. Look for a block that starts with “User-agent: OAI-SearchBot”, or a general “User-agent: *” block, followed by “Disallow: /”. Either one shuts ChatGPT’s search crawler out of the whole site. If you want to be cited, remove that rule or add a separate block for OAI-SearchBot with “Allow: /”. OpenAI says robots.txt changes take about a day to reach its search systems.
robots.txt is only the first gate. Many hosts and CDNs have bot protection that can turn AI crawlers away before they read anything. Ask whoever runs your hosting to check the firewall rules, and look in your server logs for OAI-SearchBot requests. OpenAI publishes the IP ranges it crawls from, so a real visit can be told apart from an impostor.
Then check that your pages are in the search indexes. Use Google Search Console and Bing Webmaster Tools, submit your sitemap to both, and fix the pages marked as not indexed. Copilot is built on Bing and Google’s AI features on Google Search, so this work serves more than ChatGPT.
Last, make sure key facts are in the HTML your server sends. Not every crawler runs JavaScript. A price or an address that only appears after a script loads may never be read.
What does a page ChatGPT can quote look like?
ChatGPT writes its answer from passages it has read. Give it a passage that answers the question on its own, in plain words, with specifics.
For a business, that usually means one page per service and one per location, each opening with what you do, for whom, where, and what it costs or what the cost depends on. Add the questions buyers ask before they call, with short direct answers. The brand story can sit further down.
- The first two sentences answer the page’s question directly
- Prices, or a clear explanation of what the price depends on
- Areas served, opening hours and how to book, written as text
- Short answers to the questions people ask before calling
- Organisation and local business structured data that matches the page
- One clear page per service and per city, with no near-duplicates
How do you get the rest of the web to agree with you?
Your site is one source among several. An answer that recommends a business usually rests on what a few independent sources say, so those sources need to say the right things.
Start with the listings you control: Google Business Profile, Bing Places, Apple Business Connect, and the directories that matter in your sector and country. Make the name, address, phone, hours and categories identical everywhere. Then work on reviews. Ask every satisfied customer, answer every review, and keep them coming at a steady pace, which is what review generation is about.
Mentions in local press, professional associations, partner sites and comparison articles come next. They take longer, and they are worth it, because they are exactly the kind of independent source an assistant can check you against. Digital PR is the name for earning them on purpose.
Skip the shortcuts. Fake reviews and planted forum posts break the rules of the platforms that host them, and when they are removed, the trust goes with them.
A monthly check you can run yourself
Run it on the same day each month, with the same questions, so the results can be compared.
- A fixed list of 15 to 30 buyer questions, in each language you sell in
- Each question asked in a temporary chat, so past conversations do not colour the answer
- For each answer: named or not, linked or not, and which competitors appear
- Anything false noted: old prices, a wrong address, services you do not offer
- The pages ChatGPT cites, and whether they are the ones you want cited
- Visits from chatgpt.com in your analytics, set up as in our guide to tracking AI traffic
- Enquiries where the customer says they found you through ChatGPT
- robots.txt and firewall rules checked again after any site or hosting change
Frequently asked questions
Does blocking GPTBot remove me from ChatGPT answers?
No. GPTBot collects content that may be used for training, while OAI-SearchBot is the crawler behind ChatGPT search. OpenAI treats each robots.txt setting independently, so you can block GPTBot and still allow OAI-SearchBot. Decide each one on purpose.
How long does it take to appear in ChatGPT?
OpenAI says a robots.txt change takes about a day to reach its search systems. Rewritten pages count once they are crawled again, often within weeks. Reviews and outside mentions build over months. There is no guaranteed date, and anyone who offers one is guessing.
Why does ChatGPT say wrong things about my business?
Because it found them somewhere, or it filled a gap. An old directory listing, an outdated page on your own site or a missing fact can each lead to a wrong answer. Search for the wrong detail, correct it at the source, and make the right fact easy to find on your site.
Do I need an llms.txt file to appear in ChatGPT?
OpenAI has not documented that ChatGPT reads llms.txt files on business websites; its crawler documentation deals with robots.txt. A file does no harm, but it will not make up for pages that are blocked or unclear. Our llms.txt guide explains when it is worth the effort.
Can you guarantee my business will appear in ChatGPT?
No one can. The model writes each answer, and it changes with the question and the person asking. Our ChatGPT optimization service commits to the work above and to a monthly record of what the answers say about you.
Last updated · By the Napexy team