- Methodology
- Parameters
- Answer Block Shape
Answer Block Shape
stableCategory: content-structure · Methodology v4.5
Does your page open with a direct, appropriately sized answer near the top?
Signal Source
- Source
https://{domain}- Kind
- html_dom
Score Bands
| Verdict | Condition |
|---|---|
| Pass | A direct answer paragraph of 40-60 words is present within the first-screen window (the first ~200 body words) — verdict present, score 100 |
| Partial | An answer-like block is detected but sub-optimal: too_long (over 60 words, score 60), too_short (under 40 words, score 50), or mis_placed (in range but beyond the first-screen window, score 40) |
| Fail | No answer block detected anywhere on the page, or the body is gated behind a paywall/login signal — verdict missing, score 20 |
Description
What this parameter measures
Does your page open with a direct, appropriately sized answer near the top? This parameter checks exactly that. friendly4AI strips nav, header, footer, and aside, then walks the body in document order looking for an answer cue: a question heading (H2/H3 ending in ?), a <dt> / <dfn> definition term, or a "{X} is / means / refers to" sentence opener. The first paragraph after a cue is measured by word count and by its running-word position, producing one of five verdicts: present, too_long, too_short, mis_placed, or missing.
Why it matters for AI-readiness
AI systems favour pages that lead with a clear answer to the likely user query. A tight answer paragraph in the first screen is the chunk most likely to be lifted into an AI response. A page that buries its answer below long introductions, pads it past 60 words, or stubs it under 40 words gives the model a weaker target, and a page with no detectable answer block at all rarely earns an answer-block citation. Leading with the answer directly raises your citation rate.
How we score it
The v4.4 methodology scores this Content Structure parameter as a five-state gradient. The scanner reads the answer paragraph's word count against the target range (40-60 words by default) and its position against the first-screen window (the first 200 body words). A present verdict (in range and inside the window) scores 100 and passes. A too_long (over 60 words, 60), too_short (under 40 words, 50), or mis_placed (in range but past the 200-word window, 40) verdict is a partial. Note that too_short is penalised harder than too_long, since a stub answer is less useful than a verbose one. A missing verdict scores 20 and fails; a paywall or login signal in the first 500 body characters forces missing too. The product's failAction copy cites a wider 80-200 word guide, but the processor's default target range is 40-60 words. Aim for that to clear the pass band.
How to fix common issues
- Add a concise 40-60 word answer paragraph near the top of the page that directly addresses the main user query.
- Position it within the first ~200 body words, before lengthy introductions or navigation, so it is not flagged
mis_placed. - If the block is
too_long, trim it to the core answer; iftoo_short, add one or two supporting sentences. - Precede the answer with a clear cue — a question heading, a definition term, or a
"X is …"opener — so the scanner recognises it. - Re-scan and check the
answer_block_verdict,answer_block_word_count, andanswer_block_position_wordevidence fields to confirm the fix.
Version History
- Introduced
- v4.4
- Last changed
- v4.4
Key takeaways
- Signal: https://{domain}
- Category: Content Structure
- Passes when: A direct answer paragraph of 40-60 words is present within the first-screen w…