How to fix it
Look for inlined data, hidden markup, or a component that renders far more than the visitor sees. Move data to separate requests and paginate long lists.
When it is not a problem
A long documentation page can be legitimately large. The thing to check is whether the size is text a reader sees or state a framework serialised into the page.
How Rankelle handles it
- SeverityInformational. Often correct as it stands, and reported so you can see the pattern rather than so you can clear it.
- DetectionDetected on a single page while crawling it, then grouped with the other pages under the same URL pattern.
- GroupingEvery page with this finding under the same URL pattern becomes one cause, with a confidence taken from how much of the pattern is affected. 400 of 400 is a template; 6 of 14 is phrased as a question. Causes are ranked by the impressions behind their pages.
- LandingThe affected URLs stay under cheap watch. When this finding is gone from all of them, the fix is marked landed without anyone reporting the deploy.
- VerdictAt landing, comparable untouched URLs are frozen as a control set. Twenty-eight days later the two are compared, and the fix can come back as having done nothing — all four verdicts reach the report.
Why the rule says this: Googlebot reads the first 15 MB of a page; most crawlers stop far earlier, and this one reads 1 MB. Expiry condition: Crawlers publish materially larger limits.
Try it on your own site, free
The whole audit of every website you add, no card and no expiry. Pay only when you want the proof.