ChatGPT · Diagnostic guide
Why Isn't My Site Showing Up in ChatGPT?
Google visibility does not guarantee a ChatGPT citation. The obstacle may be crawler access, page structure or content clarity.
AI systems need to reach your pages, read the content and understand it before it can be considered for an answer.
By @buildwithdylanUpdated
Find the first obstacle
Check your homepage’s technical signals before changing your content.
The short answer
A site can work perfectly in Chrome and appear in Google while being difficult for another crawler to access or parse. Check each step before assuming your content is the problem.
- 01
Access
Can the crawler reach the page?
- 02
Render
Is useful content in the response?
- 03
Understand
Are the structure and facts clear?
- 04
Answer
Does it help answer this query?
Crawlable → parseable → understandable → eligible to appear. Being selected as a source is still not guaranteed.
Enter your URL to check common technical obstacles and structural signals that can make a site difficult for AI systems to read.
Reason 1: Your robots.txt is blocking OpenAI crawlers
Your robots.txt tells compliant crawlers which paths they may request. Different OpenAI agents have different purposes:
- GPTBot crawls content that may be used to train OpenAI’s foundation models.
- OAI-SearchBot supports discovery in ChatGPT’s search features.
These permissions are independent. Blocking GPTBot does not automatically block ChatGPT search. Blocking the relevant crawler restricts access for that purpose. See OpenAI’s crawler documentation.
User-agent: OAI-SearchBot
Allow: /
User-agent: GPTBot
Disallow: /Review existing rules before editing. Firewalls and login screens can still prevent access. Check whether GPTBot is blocked or analyse your robots.txt for AI crawlers.
Reason 2: Your important content only appears after JavaScript runs
A page can look complete while the original HTML contains little information. Your browser runs JavaScript and fills in the rest afterwards. Other automated systems may not process it the same way.
Keep product descriptions, pricing, FAQs, headings, reviews and business information accessible in the initial response where possible. Heavy browser-side rendering can make extraction less reliable; JavaScript does not make every website invisible.
Reason 3: Your page structure makes the facts difficult to extract
Words need context. Use one clear H1, descriptive H2s and H3s, meaningful titles and concise factual sections. Name your company and products explicitly. Put important text on the page, not inside images.
Structured data supplies machine-readable descriptions of entities and content. Organization, Product, Article, FAQPage, LocalBusiness or SoftwareApplication may fit, depending on the page.
Only describe content that is really there. Schema is neither a citation requirement nor a shortcut to being selected. Check your structured data for a starting point.
Reason 4: Your content is too thin or vague to answer anything
A perfectly crawlable page may still offer little worth citing. Compare these descriptions:
Vague
“We provide innovative solutions that help modern businesses succeed.”
Specific
“CanAIRead.me checks whether AI crawlers can access your website, whether important content is visible in the HTML, and whether common machine-readable signals are present.”
Give explicit answers, definitions, comparisons and relevant numbers. Include product details, locations and service information. Explain relationships: who makes the product, who it serves and what it does.
Make the page easier for people to understand and machines usually benefit too.
What about llms.txt?
The llms.txt proposal offers a cleaner map of important content for systems that use it. It is an emerging convention, not a guaranteed ranking or citation mechanism.
It does not replace crawlability, readable HTML or useful content. Treat it as an extra after fixing those foundations. Check your llms.txt.
How to check which problem is affecting your site
- 01
Crawler access
Can the site be reached? Review search and training permissions separately.
- 02
robots.txt
Are relevant crawlers explicitly blocked?
- 03
Content rendering
Does meaningful content exist before browser-side JavaScript runs?
- 04
Structured data
Are key entities and page types described clearly?
- 05
Content clarity
Does the page answer real questions with specific facts?
CanAIRead.me combines technical checks and structural clues in one homepage scan. Review factual usefulness yourself; the scan does not evaluate every page or measure actual citations. Apply a fix, rescan, then compare findings.
Instant preview · Full report free by email
Frequently asked questions
Does ChatGPT know about my website?
There is no simple public database that confirms whether ChatGPT knows your entire website. Check whether your pages are accessible and contain useful information. A ChatGPT website check can identify technical obstacles; it cannot inspect a model’s training data.
Is my website readable by AI?
AI readability is not a single yes-or-no property. Crawler access, rendering, semantic HTML, structured data and clear content all contribute. CanAIRead.me tests common technical signals; you still need to assess whether the content answers a reader’s question.
How do I test if ChatGPT can read my site?
Check robots.txt and the permissions for GPTBot and search crawlers. Compare the original HTML with the browser-rendered page, inspect structured data, then review whether the page answers meaningful questions. Start with the free scanner to collect the technical findings.
Why isn't my content cited by AI answers?
Crawlability does not guarantee citation. Other sources may answer the question more directly, offer stronger evidence, clearer explanations or better topic coverage, and provide more trustworthy or corroborated information. These are useful editorial checks, not a description of ChatGPT’s private ranking algorithm.
Start with what you can check
See what AI crawlers can actually read
Run your site through CanAIRead.me and check crawler access, robots.txt, page structure, structured data and other common AI-readability issues.
Instant preview · Full report free by email
Technical accessibility is the starting point. Passing a scan never guarantees a ChatGPT citation.