Can ChatGPT read your product page?

The crawlers behind ChatGPT, Claude and Perplexity do not run JavaScript. They read the HTML your server sends, and they follow your robots.txt. Paste one product URL and see what they get.

The result appears on this page in about ten seconds. Free, no account. Use a single product page, not your home page.

What the check looks at, and why.

Five checks, in the order of the talk
  • Product in the raw HTML. We fetch the page once, without running JavaScript, and look for the product name, price and description from your own metadata in the visible text.
  • Product structured data. A Product block in JSON-LD or Microdata, with offers.price, priceCurrency and availability.
  • Reviews and /review/. Whether the page loads its reviews from review/product/listAjax, and whether your robots.txt blocks that path.
  • AI crawlers. Your robots.txt rules for this exact product URL, read per crawler as RFC 9309 prescribes, split into search, user-request and training crawlers.
  • Agentic commerce. Whether a UCP profile is published at /.well-known/ucp. Information only.
What we measured before building it

We fetched five Magento demo storefronts without JavaScript on 9 October 2026. Luma, Hyvä and Breeze deliver name, price and description in the raw HTML. The PWA Studio demo Venia delivers 71,267 bytes of code, and the only readable sentence is «Oops! JavaScript is disabled». None of the five outputs Product JSON-LD.

Luma loads review texts from review/product/listAjax when the tab opens, and Magento's default robots.txt disallows /review/.

On the same day we read the robots.txt of 125 German online shops. 87 of the 102 we could read block none of the 15 AI crawlers we checked, and none of the 22 Magento and Hyvä shops blocks an AI search crawler. The bigger risk is not the robots.txt, it is what the crawler finds once it is allowed in.

Questions

Does the check run JavaScript?

No, on purpose. GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot and PerplexityBot did not execute JavaScript in the crawler study by Vercel and MERJ. Google renders JavaScript, so a page can look fine in Google and still be empty for ChatGPT.

Does it work for Shopware, Shopify or other platforms?

Yes. The check reads plain HTML and robots.txt, which every shop has. Only the review check looks for Magento's review/product/listAjax pattern. On other platforms it tells you whether /review/ is blocked and stays quiet otherwise.

Will Product JSON-LD get my shop cited by ChatGPT?

Nobody can promise that. In a study by Ahrefs, pages that added JSON-LD were not cited measurably more often by ChatGPT afterwards. Structured data makes price and availability unambiguous for machines. Whether an AI recommends your shop depends mostly on sources you do not own: marketplaces, comparison sites, reviews.

Our firewall blocks the check. What does that mean?

Our scanner identifies itself honestly and does not run JavaScript, like most AI crawlers. If your WAF, CDN or bot protection refuses us, check whether it treats GPTBot, OAI-SearchBot, ClaudeBot or PerplexityBot the same way. On Adobe Commerce Cloud, bot protection is the merchant's responsibility.

What happens with my email address and the URL?

We store your email address, the checked URL and the time of the check to send you the result, and we may contact you once about it. No newsletter, no sharing with third parties. The site is hosted in Switzerland. Ask us to delete your data at any time: kontakt@bombelli-analytics.ch.

Questions about your result?

The check covers your own product page. Which sources an AI actually uses when it recommends shops in your category is a different measurement, and that is what we do for clients.

Connect on LinkedIn Book an intro call

Saw the talk in Mainz? Get in touch for the slides or your questions.