Technical checks
Whether your pages are readable to a client that does not run JavaScript — the precondition for everything else.
Many AI crawlers fetch a URL and read the HTML that comes back. They do not wait for a JavaScript bundle to boot and render. If your content only exists after that bundle runs, those crawlers see an empty page — and an empty page cannot be cited, quoted or summarised.
The Technical sub-tab measures exactly that, page by page.
What is checked
Server-rendered content. How much of the page's actual content is present in the HTML the server returns. This is the headline check: a page that is blank without JavaScript is invisible to a large share of AI crawlers.
Client-side rendering. Whether the page depends on JavaScript to produce its main content, and whether it offers any fallback for clients that do not run it.
Page weight. The size of the HTML and how much JavaScript the page pulls in. Heavy pages are fetched less thoroughly and less often.
Language declaration. Whether the page states its language — needed by anything deciding which market and audience the page belongs to.
Transport security. Whether the page is served over HTTPS.
Reading the results
The findings distinguish two very different situations that look similar in a browser:
- The page is server-rendered. Content is in the HTML; a JavaScript-dependent enhancement on top is fine. Fallback markup is not needed and is not demanded here.
- The page is blank without JavaScript. This is the finding that matters. Everything else on the page — structured data, copy quality, internal links — is downstream of fixing it.
That distinction is why a "no fallback content" finding appears on some pages and not others: on a server-rendered page it would be busywork.
Fixing, in order
- Get the content into the server response. Server-side rendering, static generation, or pre-rendering the routes that matter. This single change moves more than everything else combined.
- Cut the weight on pages that are heavy enough to be truncated or deprioritised.
- Declare the language correctly, including on translated variants.
- Serve everything over HTTPS, including assets and redirect chains.
Re-run the audit after deploying — results persist, so the comparison is direct.
How this connects
| Symptom | Reading |
|---|---|
| Crawlers visit but you are never cited | Likely a server-rendering problem — check here first |
| Crawlers never visit | Access problem — Robots.txt audit |
| Cited but described inaccurately | Legibility problem — Schema audit |
The Bots tab tells you whether crawlers arrive; this tab tells you what they find when they do.
Related
- Report tab — where the audits live
- Schema audit — the layer above this one
Updated 2026-08-03