Why Isn't My Website Showing Up in ChatGPT?

Ask ChatGPT about your business. If the answer is empty, vague, or wrong, one of five causes is almost always behind it. This guide walks through them in the order you should check, cheapest first.

The stakes are simple to state. As of mid-2026, 51 percent of B2B buyers start their research in an AI chatbot. When the assistant has nothing to say about you, that first conversation happens without you, and you never see it happen.

Work through the causes in order. Most sites fail on the first two.

Cause 1: Your site blocks AI crawlers

ChatGPT learns about the web through a crawler called GPTBot. If your robots.txt file tells GPTBot to go away, it goes away. The same applies to ClaudeBot and PerplexityBot.

Sites rarely block these on purpose. It happens through old robots.txt rules copied from a template, a security plugin's defaults, or a "block all bots" setting in your CDN. Cloudflare and similar services can also challenge crawlers with checks a bot cannot pass, which blocks them just as effectively without any rule you can read.

Check it:

curl -s https://your-site.com/robots.txt

Look for Disallow: / under User-agent: GPTBot, or under User-agent: * with no exceptions. Then check your CDN or firewall dashboard for bot-blocking rules.

Fix it: remove the disallow lines for the crawlers you want, or add explicit allows:

User-agent: GPTBot
Allow: /

User-agent: ClaudeBot
Allow: /

User-agent: PerplexityBot
Allow: /

One caution. Only unblock what you mean to unblock. If you have paid content or internal pages, keep those paths disallowed and open the rest.

Cause 2: Your content only exists after JavaScript runs

This is the silent one. Your page looks complete in every browser, but the HTML your server actually sends is a near-empty shell that scripts fill in later. AI crawlers do not run those scripts. They read the shell.

Check it: open your most important page, right-click, choose "View page source", and search for a sentence from your main content. Missing there means missing for every AI crawler. For the full walkthrough, including how to fetch your page exactly as GPTBot does, read Does ChatGPT read JavaScript?

Fix it: get the text into the server's HTML. On frameworks like Next.js or Nuxt that is a rendering setting. On single-page apps it is a bigger job, so start with the one page that earns the most and work down.

Cause 3: Your facts are not machine-readable

An AI engine that has to infer your prices from marketing prose will get them wrong or skip them. Structured data removes the guesswork. It is a small block of JSON in your page that states the facts plainly:

{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Your Product",
  "offers": { "@type": "Offer", "price": "49", "priceCurrency": "EUR" }
}

Check it: view your page source and search for application/ld+json. If it is absent, or only describes your cookie banner, this cause applies.

Fix it: add schema.org markup for your organization, products, prices, and FAQs. Most platforms have plugins for this, and the markup itself is plain enough to write by hand.

Cause 4: Your answers are buried

Generative engines quote passages. They favor pages where a question is a heading and the answer follows immediately, in plain sentences. If your pricing page opens with three screens of brand story, the engine may read your page and still find nothing quotable.

Check it: for each question a buyer would ask, find where your site answers it. Count how far into the page the actual answer sits.

Fix it: restructure your key pages so every important question gets a direct answer within the first sentence or two of its section. Answer first, elaborate after. This also happens to be what impatient humans prefer.

Cause 5: Your site is too new or too thin to trust

Engines weigh evidence. A site with no history, few pages, and no mentions elsewhere gives them little to stand on, even when it is perfectly readable. This cause is real but overdiagnosed. Check the first four before you settle on this one, because they are faster to fix and more often guilty.

Fix it: publish genuinely useful pages around the questions your buyers ask, and earn mentions on sites the engines already read. This is the slow part of GEO, and there is no shortcut worth your money.

Diagnose in this order

Order	Cause	Time to check	Time to fix
1	Crawlers blocked	2 minutes	Minutes
2	JavaScript-only content	5 minutes	Days to weeks
3	No structured data	5 minutes	Hours
4	Buried answers	An hour	Days
5	Thin authority	Honest reflection	Months

A note on re-testing. Ask the assistant the way a stranger would, in a fresh chat: "What does Acme do and what does it cost?", with your company in place of Acme. Your own chat history can color the answer, so a clean session is the honest test. Test in ChatGPT and Perplexity both, because they read the web on different schedules, and write down the date and the answer so you can tell whether your fixes changed anything.

Fix one cause at a time and re-test, because the symptoms look identical from the outside. This checklist is essentially what our scan automates, six areas, graded from evidence, with the scoring approach published.

Or start by seeing which of the five causes applies to your site right now.