Section Length Distribution

What this parameter measures

Are your H2 sections scoped to a size AI systems prefer to cite? That is the question behind this parameter. friendly4AI splits the main content region by H2 boundaries (treating any content before the first H2 as an implicit section), counts the words in each section, and computes the arithmetic mean across all of them. A page scores best when that average lands in the 120-180 word window: long enough to carry a substantive answer, short enough to stay focused on one subtopic.

Why it matters for AI-readiness

Answer engines extract and cite at the section level, mapping an H2 heading to the passage beneath it. Sections that are too short read as thin and offer nothing worth quoting; sections that run hundreds of words dilute the core answer and force a model to guess which sentence to lift. A consistent 120-180 word section gives ChatGPT, Gemini, and Perplexity a clean, self-contained unit to surface, which raises your citation odds in answer-style results.

How we score it

Under the v4.4 methodology, this Content Structure parameter is a gradient driven by the average section length. The processor awards 100 when averageWords is in the inclusive [120, 180] range. Outside that band it computes a deviation (120 - avg when too short, or avg - 180 when too long) and deducts round(20 * deviation / 50), clamped to 0-100. So a page passes with an average inside the optimal range, earns a partial when the average drifts moderately and the penalty bites without zeroing the score, and fails when the average is far outside the range or the page has fewer than two sections (an automatic 0, since a single block cannot be assessed for distribution). The bands reflect the processor's proportional penalty rather than fixed cutoffs.

How to fix common issues

• Aim for each H2 section to carry roughly 120-180 words of substantive content.
• Split sprawling sections that run well past 180 words into smaller, focused sub-sections with their own H2s.
• Expand thin sections under 120 words with concrete examples, data, or evidence rather than filler.
• Make sure the page has at least two real H2 sections so the distribution can be measured at all.
• Re-scan and check the averageWords and sectionCount evidence fields to confirm the average moved into range.