Skip to content
AI search visibility · ChatGPT

How to get your business mentioned in ChatGPT

By · Last updated:

ChatGPT answers with a small set of sources it can fetch and parse. Your job is to be reachable and easy to use. This guide covers the checks that live on your side of the fence, the ones you can run this afternoon, and the point where outside inspection stops.

Check what ChatGPT can reach on my site

Free scan · verified account required · no credit card

Key takeaways

  • ChatGPT search builds answers from sources it can retrieve. OpenAI documents OAI-SearchBot for search and GPTBot for training as separate agents.
  • Access is layered: robots.txt asks, and your CDN or firewall enforces. A robots.txt line does not prove a crawler is getting through.
  • You can run five checks today: robots.txt, edge rules, how the page serves, how clearly it states what you do, and what other sources say about you.
  • The free scan measures page-level signals. CDN behaviour, JavaScript rendering, and OpenAI's internal source selection need a deeper look.

How ChatGPT finds sources

ChatGPT search builds answers from a set of pages it can retrieve. OpenAI documents its crawlers separately: OAI-SearchBot for ChatGPT search, and GPTBot for model training. They are controlled independently. Blocking GPTBot does not remove you from ChatGPT search, and allowing it does not put you in.

Check 1: crawler access in robots.txt

Open robots.txt at your domain root and read the user-agent blocks. A block that keeps you out of ChatGPT search looks like this:

User-agent: OAI-SearchBot
Disallow: /

An allow looks like this:

User-agent: OAI-SearchBot
Allow: /

Some platforms, security plugins, and CDN bot presets ship crawler-blocking rules by default, and copied templates can carry blocks nobody intended. If an AI-crawler rule is present and nobody on your team remembers adding it, treat it as a real configuration to review. Verify by opening https://yourdomain.com/robots.txt in a browser.

Check 2: what sits in front of your site

robots.txt asks. Your CDN or firewall enforces, and it can block a crawler that robots.txt allows. Cloudflare has bot-management settings that can block AI crawlers at the edge, and Vercel has firewall rules. A robots.txt line does not tell you whether OAI-SearchBot is actually getting through. Confirm the setting in the CDN dashboard, or ask whoever manages it.

Check 3: what the crawler actually receives

Crawlers differ in whether they run JavaScript, and OpenAI's crawler documentation is the place to confirm that for its bots. Content that appears only after scripts execute may not be in the response a given crawler gets.

Check it yourself: open the page, then view the page source (Ctrl+U in most browsers), and search the source for a sentence you can see on screen. If it is not in the source, that points to a likely rendering problem rather than proof that no crawler can see the page. Some crawlers do run scripts, so treat the missing text as a signal to investigate, not a verdict.

Ecommerce example: a store that renders product descriptions and specifications client-side shows a full page to shoppers and may show a near-empty one to a crawler that does not run scripts. Moving the description into server-rendered HTML closes the gap for the crawler and, as a side effect, speeds the page up for shoppers on slow connections.

Check 4: is the page clear about your business

An answer engine can only reuse a sentence that stands on its own. State what you sell, who you serve, and where you operate in the first screen. Name the product or service the way customers do.

Check 5: do other sources describe you

Answers often draw on more than your own site. Reviews, directories, and press contribute slowly. Worth knowing this signal exists. Not worth chasing this week.

What the free scan measures, and what it cannot

The free scan reviews one public page and reports page-level signals: 18 SEO checks, 11 GEO readiness findings, and 17 AEO Readiness (Beta) checks, plus a separate review of 9 crawler identities. It reads robots.txt and the directives on the page.

Three things sit outside it. It cannot see your CDN or firewall rules, so it cannot confirm that OAI-SearchBot reaches you through the edge. It does not render JavaScript the way Google does. It cannot see which sources OpenAI chose for an answer, because that happens inside OpenAI. Those questions need a deeper look at your CDN configuration, your rendering, and the page itself.

Run the checks in order

  1. Fetch robots.txt and look for OAI-SearchBot and GPTBot.
  2. Confirm your CDN or firewall bot rules.
  3. View source and confirm your key text is in the HTML.
  4. Ask five real customer questions in ChatGPT and note whether you appear and whether the facts are right.
  5. Run the free scan on your main page for the on-page signals.

See what your page exposes

The free scan checks one public page for the signals above, in a few seconds. You need a verified account, and there is no credit card. Results are directional: they describe what the page exposes, not what OpenAI will do with it.

If you would rather have the access path and the page reviewed by a person, with a prioritized fix plan across your main page and up to four more, the done-for-you audit is the $1,995 human-reviewed audit.

Frequently asked questions

Does ChatGPT use my website?

It can, when it runs a search and your page is among the sources it can fetch and parse. It does not read every site.

Why is my business not in ChatGPT?

Start with access. A blocked crawler, content that is not in the served HTML, or a page that is hard to parse cover the common causes. Selection beyond that happens inside OpenAI.

How do I get ChatGPT to recommend my business?

You cannot compel a mention. Removing access problems and writing clear, factual pages improves your odds.

Is this the same as SEO?

Related, not identical. ChatGPT builds answers from sources it can parse, so access and clarity matter alongside ranking position.

Do FAQ questions give me an AI ranking advantage?

No. They make the page easier for machines to read. No markup makes a system cite you.

By · Last updated: