Why is my website not showing on ChatGPT?

ChatGPT uses three different crawlers, and they are not interchangeable

This trips up almost everyone. OpenAI does not have one bot; it has three, each with a separate job and a separate name in robots.txt. Blocking the wrong one produces exactly the outcome people complain about.

A great many sites added User-agent: GPTBot / Disallow: / during the training-data backlash and then blocked all three by copying a stricter snippet from a blog post. If you want the traffic but not the training, that distinction is the whole game.

Check what your robots.txt actually says

Not what you think it says — what it says. Open yoursite.com/robots.txt in a browser. Then read it carefully, because two rules commonly bite:

• A Disallow: / under User-agent: * blocks every crawler that does not have its own more specific group. Adding User-agent: OAI-SearchBot / Allow: / fixes it, because the most specific matching group wins.
• Some hosting platforms and CDNs inject or override robots.txt at the edge. If the file you uploaded is not the file being served, that is your answer. Fetch it and compare.

If the page needs JavaScript, assume it is invisible

Ask for your page the way a crawler does, with no browser involved:

curl -sL https://yoursite.com | grep -i "some text from your page"

If your actual content does not come back, it is not in the HTML — it is assembled in the browser afterwards. Googlebot renders JavaScript, slowly and selectively. AI crawlers overwhelmingly do not. A React or Vue app with client-side rendering serves an empty <div id="root"> to every one of them.

The fix is server-side rendering or static generation, which every major framework supports. This is the single highest-impact change for AI visibility and it is invisible in a browser, which is why it goes unnoticed for years.

Being crawlable is not the same as being cited

Once the mechanics are right, you are competing on merit. What answer engines pull from, in rough order of influence:

• Directly answering a specific question. Pages structured as question-and-answer get quoted because a passage can be lifted whole. A page that buries the answer in paragraph nine does not.
• Being the primary source. Original data, first-hand detail and specifics get cited. Summaries of other people's articles get skipped in favour of the article.
• Corroboration elsewhere. Models weigh whether other sources agree with you. A claim that appears nowhere else reads as unreliable, however true.
• Clear attribution. A named author, a visible organisation, a date. Anonymous content is harder to cite responsibly, so it gets cited less.

Work through it in this order

1. Fetch yoursite.com/robots.txt and confirm OAI-SearchBot and ChatGPT-User are not blocked, by name or by a catch-all.
2. Fetch the page with curl and confirm your real content is in the HTML.
3. Search for a distinctive sentence from your page in Bing. ChatGPT search leans on that index; absence there is a strong signal.
4. Ask ChatGPT directly about a topic you cover and see who it does cite. That tells you what a citation-worthy source looks like in your area.
5. Only then change the writing.

Steps one and two take ten minutes and account for most cases. Do not rewrite anything before you have done them.