How to get your business found on ChatGPT: a practical guide

Webopics Insights cover graphic titled Can AI Find Your Business, showing a stylised browser window with a search bar and chat bubbles

The short answer. Allow OAI-SearchBot in your robots.txt and make sure your host or CDN is not blocking it, because that is the only documented control over whether you can appear in ChatGPT search. Then publish clear, accurate pages that answer real questions. OpenAI does not publish ranking factors, so everything beyond access is informed practice, not a confirmed requirement.

There is a lot of confident advice about ChatGPT visibility, and very little of it comes from the companies involved. This guide separates what is documented from what is not.

Three different things people mean by "found on ChatGPT"

Being cited as a source. ChatGPT answers a question and links to your page as supporting material. This depends on your site being reachable by OpenAI's search crawler and on your page containing a clear answer to the question asked.

Being recommended as a business. Someone asks for an interior designer in Kozhikode and ChatGPT names companies. This draws on far more than your own website: directories, reviews, press coverage and anywhere else your business is described. You control very little of it directly.

Ranking in traditional search. Position in Google's blue links. Related, because an accessible, well structured site helps everywhere, but a separate system with its own documentation.

Most businesses asking about ChatGPT want the second one. It is also the one nobody can sell you a reliable method for.

Start with access, because it is the only documented control

OpenAI runs four crawlers and they do different jobs. Confusing them is the most common and most expensive mistake on this topic.

CrawlerWhat it doesAffects ChatGPT search?
OAI-SearchBotSurfaces websites in ChatGPT's search featuresYes. This is the one
GPTBotCrawls content that may be used to train foundation modelsNo
ChatGPT-UserFetches a page when a user asks about itNo, and not an automatic crawler
OAI-AdsBotValidates the safety of pages submitted as adsNo

Blocking AI training does not remove you from ChatGPT search

This is worth stating plainly, because the opposite is widely assumed. OpenAI's documentation says that each setting is independent of the others. Disallowing GPTBot signals that your content should not be used to train models. It does not affect whether you appear in ChatGPT search answers, which is governed by OAI-SearchBot.

So if you have reservations about your content training models, you can disallow GPTBot and still allow OAI-SearchBot. You are not forced to choose. The reverse does apply: OpenAI states that sites opted out of OAI-SearchBot will not be shown in ChatGPT search answers, although they can still appear as navigational links. Changes take around 24 hours from a robots.txt update.

Check your host, CDN and firewall, not just robots.txt

A permissive robots.txt means nothing if something upstream is refusing the request before it reaches your site. Bot protection on a CDN, an aggressive firewall rule or a security plugin can all block OpenAI's crawler while your robots.txt politely invites it in.

OpenAI publishes IP ranges for each crawler as JSON files, which is how you confirm a request is genuinely theirs rather than something impersonating them. If you use Cloudflare or similar, check your bot management settings specifically. This is the easiest item on the list to get wrong without noticing, because nothing breaks visibly.

Make your pages easy to read and easy to quote

Clear service pages help. One page per service, stating plainly what it is, who it suits and what is included, is more useful than a single page listing everything you do. Accurate company information matters too: location, service areas, contact details and opening hours should be consistent wherever they appear, including your Google Business Profile.

Pages that answer questions directly are easier to quote than pages that circle a topic. If customers regularly ask what something costs, how long it takes or whether you cover their area, answering that in plain text is useful to a human reader and happens to be the format an AI system can extract. Keep those facts in text rather than inside an image or a script-rendered widget, which is standard SEO advice and applies here unchanged.

Credibility practices, clearly labelled as such

These are reasonable things to do. None is a confirmed ChatGPT ranking factor, because no such list exists publicly.

  • Original expertise. Writing about work you have actually done gives you material nobody else has.
  • Documented case studies. Specific, dated, honest about what was measured.
  • Authentic reviews on platforms customers use. Never purchased.
  • Third-party coverage. Being described accurately on sites other than your own.

The honest reasoning: AI systems describing your business draw on sources beyond your website, so existing in more credible places is more likely to help than not. That is a sensible inference, not a documented mechanism. Treat anyone who presents it as proven with some caution.

Google does publish guidance on its own AI features, and it is blunt. The best practices for SEO remain relevant, there are no additional requirements or special optimisations needed, and there is no special schema.org structured data you need to add. Google also says structured data should match the visible text on the page. Use markup to describe what is genuinely there, not as a trick.

A hypothetical example

Hypothetical, not a client

A modular kitchen company in Kannur serves Kannur, Thalassery and Kozhikode. Its website is a single page, the service areas appear only in a banner image, and its CDN bot protection blocks unfamiliar crawlers. Someone asks ChatGPT for modular kitchen companies near Thalassery and the company does not come up. The fixable problems here are access and text: the crawler cannot reach the site, and the one fact that would make it relevant to the question exists only inside a graphic. Fixing both is worth doing regardless of AI, because the same two problems affect Google.

How to measure this without fooling yourself

Links from ChatGPT search results generally arrive tagged with utm_source=chatgpt.com, which is how you identify the traffic. One quirk is worth knowing: that tag sets a source but no medium, so GA4 may file these sessions somewhere other than Referral. A custom channel group matching on source will group them properly. Some visits arrive with neither a tag nor a referrer and land in Direct, so whatever number you see is a floor, not a complete count. Track enquiries separately: referral sessions tell you people arrived, and only your enquiry log tells you whether it was worth having.

One caution matters more than any of the above. Asking ChatGPT a few questions and noting who appears does not establish a ranking. Answers vary with phrasing, conversation context and features such as memory, so two people asking the same question can get different businesses. Testing a handful of prompts and concluding you rank first, or that a competitor does, is reading the weather from a single afternoon.

The checklist

  1. Confirm OAI-SearchBot is allowed in robots.txt.
  2. Check your CDN, firewall and security plugins are not blocking it separately.
  3. Decide on GPTBot deliberately. Blocking training does not affect search visibility.
  4. Give each service its own page with plain language about what it is and who it suits.
  5. Make sure location, service areas and contact details are consistent everywhere, including your Google Business Profile.
  6. Move key facts out of images and into text.
  7. Answer your most common customer questions somewhere on the site, in writing.
  8. Add structured data that matches what is visibly on the page.
  9. Set up a GA4 channel group that catches the chatgpt.com source.
  10. Log enquiries with their source, so you can tell traffic from business.

Questions business owners ask

Do I have to let OpenAI train on my content to appear in ChatGPT search?

No. OpenAI's documentation treats these as separate controls and states that each setting is independent of the others. GPTBot governs training. OAI-SearchBot governs whether a site can appear in ChatGPT search answers. You can disallow GPTBot and still allow OAI-SearchBot.

Will adding an llms.txt file get my business cited by ChatGPT?

There is no published evidence that it will. Google has said it does not support llms.txt and has no plans to. OpenAI has made no public statement that it uses the file for ranking or citation, although some site owners report seeing it fetched. It costs nothing to publish, but it should not be sold to you as a route to AI citations.

Is there special schema markup that gets you into AI answers?

Google's guidance says there is no special schema.org structured data you need to add for its AI features, and that structured data should match the visible text on the page. OpenAI publishes no ranking guidance at all. Use structured data because it describes your page accurately.

How long do robots.txt changes take to affect ChatGPT search?

OpenAI's documentation says around 24 hours from the robots.txt update. That applies to the access setting itself. It does not mean your site will start appearing in answers within 24 hours, because OpenAI publishes nothing about how content is selected.

A competitor appeared when I asked ChatGPT about my industry. Do they rank higher?

Not necessarily. Answers vary by phrasing, by conversation context and by features such as memory, so two people asking the same question can see different businesses. A handful of prompts is anecdote. Referral traffic and enquiries measured over time are the signals worth acting on.

Where this leaves you

The documented part of this is small and worth doing properly: make sure the right crawler can reach you, and decide the training question on your own terms rather than on a myth. The rest is the same work that has always made a website useful, which is why it is a reasonable bet even without a confirmed mechanism behind it.

If you would like someone to check whether your site is actually reachable and fix the parts that are not, that is part of how we build and maintain websites. You can message us on WhatsApp or call +91 70128 12012. Related reading: how we approach performance marketing and what to look for in a digital marketing agency in Kannur.