Render-parity diff

A render-parity diff compares two versions of the same page: the raw HTML returned by the server, and the DOM after a real browser has executed the page's JavaScript. The bigger the difference, the more content depends on client-side rendering, and the less a non-JS crawler can read.

This matters because the AI crawlers (GPTBot, ClaudeBot, PerplexityBot, and friends) fetch the raw HTML. Classic Googlebot does render JavaScript on a second pass, but the AI crawlers, in their documented production paths, do not. So a site can pass a Google render check and still be invisible to ChatGPT and Perplexity. The render-parity diff catches exactly that case.

To compute it, we fetch the raw HTML, render the same URL in a headless browser, and compare the extracted body text and the heading set. We report a parity score and flag pages where the raw version is missing the main content. The remediation is to server-render the primary content. The diff is also a useful regression guard: a framework upgrade or a new client-only component can quietly drop a page below the readability threshold, and the diff surfaces it.