Boilerplate ratio

What this check looks for

This check compares the length of the readability-extracted main content against the total visible text on the page. It requires at least 200 characters of total visible text to run — thinner pages report "na". If main content makes up less than 25% of the visible text, it warns, reporting the exact character counts and percentage. At 25% or higher, it passes.

Why it matters

Every page carries some unavoidable boilerplate — navigation, footer links, cookie banners — but when that boilerplate dwarfs the actual content, search engines and readers alike have to wade through repeated site chrome to find anything unique to this specific page. A low ratio is also a duplicate-content risk in disguise: if 80% of a page's text is identical navigation repeated across thousands of pages, the small remaining slice may not be enough to make each page feel meaningfully distinct. AI extractors specifically try to isolate main content from chrome (this is what the geo.extract.readability-yield check measures too) — a low boilerplate ratio means more of the page is noise the model has to filter out before it can find your actual message.

How to fix it

Increase unique content relative to repeated structural elements:

<!-- Mostly nav/footer, minimal unique content -->
<nav>...200 links...</nav>
<main><p>Contact us for pricing.</p></main>
<footer>...150 links...</footer>

<!-- Stronger ratio: substantive main content -->
<nav>...200 links...</nav>
<main>
  <h1>Enterprise pricing</h1>
  <p>Our enterprise plan includes dedicated support, SSO, and custom SLAs...</p>
</main>
<footer>...150 links...</footer>

Where nav/footer link walls are unavoidable, wrap main content clearly in <main> so extraction tools know exactly where to look.