Why Your Site Is Invisible to ChatGPT and Claude

What it is.

A 17-source read of AI crawler behavior, built on four official crawler docs plus published crawl logs, aimed at one question: what do these bots retrieve?

GPTBot pulled JavaScript files in 11.50% of requests and ClaudeBot in 23.84%, and neither executed a line of it. Content that renders client side is content they never read.

Both crawlers waste roughly a third of their budget: 34.82% of ChatGPT's requests and 34.16% of Claude's landed on 404 pages, against 8.22% for Googlebot.

Read it at the source.

Opens on TripleDart. Free, and the figures on this page all come from it.

Read the guide

Frequently Asked Questions

Strengths

It separates training crawls from citable search traffic, which are 48% to 52% and 9% to 11% of AI crawl volume.
The crawl-to-refer ratios give you a blunt way to argue the trade: near 870 to 1 for OpenAI and near 5,100 to 1 for Anthropic.
Server-side rendering and clean redirects are the two changes it lands on, and both are things an engineer can pick up this sprint.

Limitations

This is our own analysis, and TripleDart sells technical SEO work.
The crawl-log figures come from other people's published datasets, so they carry those datasets' sampling.
Crawler behavior moves month to month, so treat the percentages as a read on this year and not a constant.

Skip this if

Skip it if you already server-render every page and your log files show no 404 spike from AI user agents.

Ideal for

SEO leads working with engineering
Anyone auditing a JavaScript-rendered site

Why Your Site Is Invisible to ChatGPT and Claude

What it is.

Read it at the source.

Frequently Asked Questions

The plays behind the reading.