Why Does ChatGPT Not Show My Website? Six Causes
ChatGPT skips a site for six reasons: no search, a block, an unreadable page, no index entry, no page that answers, or a stronger source. Find yours.
By Alex Nazarov
On this page
ChatGPT usually skips a website for one of six reasons, in this order. It answered without searching. It was blocked from the site. It could not read the page. The page was missing from the index it searched. No page answered the question, or another source looked stronger. Check them in order and stop at the first one that fits.
The order is the useful part. The first five causes are mechanical, and they come before anything to do with how good your work is. A page that fails one of them never gets as far as being judged.
How ChatGPT chooses sources covers how it picks between pages. This page covers why an assistant never got as far as your site.
The six causes at a glance
| Cause | What it looks like to you | Who can fix it |
|---|---|---|
| 1. No search happened | The answer names nobody and lists no sources | Nobody. The assistant decides |
| 2. Your site is blocking AI | Your pages never appear in any assistant | You, in robots.txt or your host |
| 3. The page cannot be read | Your words are missing from the page source | You or your developer |
| 4. The page is not indexed | Searching your own page title in Google finds nothing | You |
| 5. No page answers the question | Competitors with plain answer pages get quoted | You |
| 6. Another source looked stronger | You appear sometimes, others more often | You, slowly |
Fixing cause five while cause two is still true changes nothing.
Cause 1: Did the AI search at all?
Often the AI never looked anything up, so no website could have been cited.
In our test of 432 creator questions, ChatGPT searched the web for only about a third of them. GEO vs SEO has the full numbers.
When an assistant answers from its training, it does not search the web. It provides no source list or links and gives no one credit. Since there was no search, no site was in the running.
How to know if the AI searched the web: look under the answer. No sources listed means no search happened. If the answer cites other sources but not yours, skip ahead. A search happened, but your site was left out for another reason.
You cannot fix this one, but it is still worth knowing. It can stop you from rebuilding your site because of an answer that was never going to cite anyone.
Cause 2: Is your site blocking AI without you knowing?
This can catch people who did everything else right. It is usually not something anyone chose on purpose.
Your robots.txt file, a small text file that tells crawlers what they may read, can block AI crawlers in a single line. Your hosting company can also block crawlers. So can your security provider or content delivery network (CDN). These blocks can stop crawlers before they ever reach robots.txt. Google’s guidance for AI features says to allow crawling in both robots.txt and your CDN or hosting setup. That is why checking robots.txt alone is not enough.
Default settings have changed too. On July 1, 2025, Cloudflare announced that new domain owners would be asked during sign-up whether to allow AI crawlers. Rush through that choice or leave the default in place, and you may end up with a block no one remembers setting.
Perplexity recommends allowing PerplexityBot in your robots.txt file. This helps your site appear in its search results. It also recommends allowing its published IP ranges.
Pages behind a login or a paywall sit here too. A crawler has no account, so it sees what a logged-out visitor sees, which is usually a prompt to subscribe.
To check if your site is blocking AI, open yoursite.com/robots.txt in a browser and read it. Then check the bot or security settings in your hosting dashboard. Do both, not just the first.
Cause 3: Can the AI read the page?
A page can be fully allowed and still arrive empty.
Many modern sites draw their text with JavaScript after the page loads. A browser runs that code and shows you a full article. A crawler that does not run it sees an almost blank document. Vercel published its research on December 17, 2024. Its tests found that major AI crawlers did not render JavaScript. The one exception was Google’s Gemini, which used Googlebot’s systems and could fully render JavaScript.
Google gives similar advice in its best practices for AI features. It says to make sure important content is available as text.
Answers that exist only inside a video, an audio file or a picture of text fall here as well. The crawler loads the page but finds no text it can quote. What each engine can read on social platforms explains this in more detail.
To test whether AI can read your page, look up how to turn off JavaScript in your browser. Once JavaScript is off, load your website in a new tab. What you see is close to what an AI crawler sees.
Cause 4: Is the page in the index the AI searches?
This section uses Google’s published rules as an example you can check. Google explains its requirements. ChatGPT does not publish how its search index is built.
Google must index your page and allow it to appear in search results with a snippet. Only then can it appear as a supporting link in AI Overviews or AI Mode. Google says there are no extra requirements or special optimizations needed.
The settings that remove a page from search remove it from these features too. Google lists four controls that limit what it can show from your pages: nosnippet, data-nosnippet, max-snippet and noindex.
This also answers a question we get often: my site is on page one of Google, so why does AI never quote it? Ranking for a phrase is not the same as having a passage that answers the exact question someone asked. The page is eligible. It just was not the best match for that question.
To test if Google has indexed your page, search Google for the page’s exact title in quotation marks, plus your website’s name. If your page does not show up, it is most likely not indexed, and none of the later steps can happen. If you do not see it, Google Search Console can show you why.
Cause 5: Does any page actually answer the question?
You have likely answered a question a hundred times, in reels, in episodes and in replies. Somebody asks an assistant that exact question. The assistant searches and finds a competitor’s page. The title matches the question, and the first two sentences answer it. The assistant quotes that page. Your answer may have been better, but it was not on a page, so it was never a candidate.
People often blame a niche subject, but the usual cause is having no page that answers real questions about it. The narrower the subject, the less competition that page will face once you create it.
Create one page for each question. Answer it in the opening lines, then add the details only you can provide. iSonic does this from content you already made: its GEO Questions turn your videos and posts into one page per question.
Cause 6: Did another source look stronger or clearer?
This is the first cause where the assistant judges your page against others.
If the assistant searched, check your page. Could it reach and read it? Was the page indexed and relevant to the question? If all of that is true and it still did not cite you, it chose another source. Why AI cites some websites and skips others explains how it makes that choice. Read it next if you have ruled out the first five reasons.
Name confusion belongs here too. An assistant may confuse you with another person or business that shares your name. It may also give a cautious answer rather than a clear one. Old descriptions of you on other sites can cause the same mix-up in a quieter way.
How you know this is you: you do appear sometimes, for some questions, on some assistants. Partial visibility means the technical setup works. Other sources are being chosen ahead of yours.
What is probably not the cause?
Being small or new does not rule you out of being cited.
Google’s best practices for AI features include giving users a great page experience. Page speed is part of that. It is a best practice, but Google does not require it. If a slow page is crawlable, readable and indexed, the slowness may not be the cause.
How do you find why ChatGPT does not show your website?
Five checks, in order, stopping at the first failure.
- Ask ChatGPT a common question that your content answers, and look for a source list. No sources means the AI did not perform an internet search for its answer.
- Open yoursite.com/robots.txt and your host's bot settings. A block means your site is not letting the AI read it.
- View your page source and search for a sentence you can see on screen. If you can’t find it in the source, the AI is not able to find it.
- Search Google for your page title in quotation marks. If nothing comes back, Google has likely not indexed that page.
- Look for a page of yours that answers the exact question in its first two sentences. If none does, that is your answer.
If all five pass, you are at cause six, which has no quick check and no quick fix.
What can nobody promise?
There is no certainty or guarantee your work will be cited.
Each AI assistant decides whether to search and which sources to credit. No tool, agency or platform, iSonic included, sits inside that decision.
Where does iSonic fit?
iSonic is a GEO platform that creates structured pages from a creator’s existing content. AI assistants can read these pages, find the content and credit the creator. The pages are structured with best practices in mind. They sit on iSonic's site, which allows the major AI crawlers in its robots.txt file. Each page answers one question, which makes it easier for AI to find and cite. iSonic builds these pages from your website and social media content, including video and audio that AI cannot read as text.
Quick answers
Why does ChatGPT recommend my competitors instead of me?
Usually, they have a page that answers the question and you do not. This does not mean the assistant judged your expertise. Check the six causes first. Your page may be blocked, unreadable or missing from the index. Any of these can rule it out before the assistant compares sources. If you pass all five checks, the comparison itself is the next thing to look at.
Why can't ChatGPT access some websites?
Most often a deliberate or inherited block. A robots.txt rule can block AI crawlers from your page. Your hosting service, security settings or content delivery network can also block them. Some services, such as Cloudflare, now ask whether to allow AI crawlers during sign-up. Pages behind a login or paywall are also out of reach because crawlers have no account.
Do paywalls stop AI from citing my work?
Generally yes, for the part behind the paywall. A crawler sees what a logged-out visitor sees, which is usually a headline and a prompt to subscribe. Anything you want quoted needs to sit in front of the paywall as readable text. A free summary page answering the question is the common workaround.
Does a slow site stop AI from citing me?
Not on its own. Google lists a good page experience as a best practice for its AI features. It is not a requirement. A slow page that crawlers can reach, read and index remains eligible. Speed is worth fixing for your readers, but it is unlikely to be the reason you are missing.
How long after fixing these will AI mention me?
Nobody can give you a reliable timeline. A page must be crawled before it can be used. Each assistant then decides whether to search and which sources to cite. Treat the fixes as removing obstacles, not as starting a clock.