Content Depth

Content Depth checks whether a page carries enough substantive content for its purpose. friendly4AI counts the words and the H2/H3 sections in the main content area, then compares both against thresholds tuned to the page's detected site category. A page passes (100) only when it meets both the full word and full section targets for its category. It scores partial (50) when it clears one but not both, and fails (0) when it falls short of both. This is an Authority and Trust signal, and its scoring has not changed since v4.0.

What does Content Depth measure?

The parameter measures two things in the substantive part of a page: word count and section structure. friendly4AI first scopes the analysis to the main content area, preferring a <main> block and stripping out <nav>, <footer>, and <noscript> noise. It then counts the substantive words and the number of H2/H3 sections.

It also works out a site category for the page — blog_docs, ecommerce, landing, or default — using explicit category artifacts, the URL path, and HTML cues. A documentation page and a landing page have wildly different ideas of reasonable depth, so judging them by the same yardstick makes no sense. One safeguard: when the main content area holds under 65% of the page text, a low-coverage fallback widens the scope to the whole page. That keeps boilerplate-heavy layouts from being penalised unfairly.

Why does content depth matter for AI-readiness?

Thin content gives AI systems — ChatGPT, Claude, Perplexity, Gemini, and AI crawlers — very little to extract, summarise, or cite. Cover a topic thoroughly and split it into clear sections, and a model has multiple passages it can pull for different questions. That raises the odds your content gets surfaced.

Depth is always judged against category, never as an absolute number. A 400-word product page can be perfectly complete, while a 400-word documentation page is plainly thin. Match real depth to what the page is for, section it properly, and your content becomes far more useful to answer engines than a shallow wall of marketing copy.

How is Content Depth scored?

Category-aware thresholds drive this parameter under the v4.5 methodology, and they have not changed since v4.0. Once the category is detected, the processor compares the word count and H2/H3 section count against that category's targets:

| Site category | Full word target | Full sections | Partial word threshold | |---|---|---|---| | default | 1500 | 3 | 500 | | blog_docs | 2000 | 4 | 800 | | ecommerce | 800 | 3 | 400 | | landing | 500 | 2 | 250 |

Scoring bands:

• Pass (100): the page meets both the full word and full section thresholds for its category.
• Partial (50): the page reaches the partial word threshold OR has at least one fewer than the full section target (minimum 2).
• Fail (0): below both the partial word threshold and the section threshold.

The threshold_applied evidence field reports which threshold name was used, so you can confirm exactly which category the scan detected.

How do I fix Content Depth issues?

• Check the detected site_category in evidence first. Depth is judged against that category's thresholds, not a single global number, so know which target you are aiming at.
• Expand the substantive explanation until you clear the word target for your category. Add real information, not boilerplate.
• Split the content into clear H2/H3 sections so the section count meets the category target. Flat pages with no subheadings score poorly. See section-length-distribution for how to size those sections.
• Trim navigation- and footer-heavy layouts. If your main content sits below 65% of page text, the low-coverage fallback widens the scope, but real depth still wins the score.
• Back claims with concrete evidence (citation-evidence-density) and clear authorship (author-authority), so the extra length is genuinely substantive.
• Re-scan, then read the word_count, section_count, site_category, and threshold_applied evidence fields.