Improve AI search readiness (GEO)
The AI Search Readiness tab scores how clearly AI search engines and large language models can parse and cite each published page. The score is also called GEO (Generative Engine Optimization). Each page gets a 0 to 10 score across six factors, with a worklist of the pages most worth fixing.
Before you begin
Scores are computed when you publish a site. Publish your content first. Pages with no published version have no score.
Scores are recalculated at publish time. After you improve a page, re-publish to see its score update.
Open the tab
Go to Analytics in the sidebar.
Open the AI Search Readiness tab.
Read the readiness breakdown
The breakdown shows the average score across all scored pages, plus how many pages fall into each band:
Band | Score | What it means |
|---|---|---|
AI-ready | 7 to 10 | Clear enough to be cited |
Partially understood | 4 to 6 | Parseable but not optimized |
Likely ignored | 0 to 3 | Too thin or unstructured |
The six factors
Each page's score is a weighted blend of six factors, each scored 0 to 10:
Structural clarity: heading hierarchy, paragraph length (about 20 to 80 words is ideal), and list usage.
Semantic completeness: whether the page actually answers the question its title implies.
Schema potential: whether the topic type maps to structured data. Task, how-to, FAQ, tutorial, and troubleshooting types score highest.
Answer density: how many direct, factual statements there are versus filler and preamble.
Internal linking: cross-reference density to related topics; roughly four to eight links is ideal.
Fetch size: how large the page's Markdown is. Pages over about 50,000 characters are silently truncated by AI crawlers mid-read, so an oversized page scores low no matter how well it is written.
Work the worklist
The Pages to improve worklist lists every page scoring below AI-ready (under 7), worst first. Each row shows:
The page (click to open it in the editor).
Its score.
Its biggest opportunity: the lowest-scoring factor and a concrete fix for it.
Work top to bottom: open the page, apply the suggested fix for its weakest factor, then re-publish. Because each factor is a deterministic lever, the fix shown maps directly to how that factor is scored:
Weak factor | Suggested fix |
|---|---|
Schema potential | Set a topic type: task, how-to, or FAQ score highest. |
Structural clarity | Add headings and a list; keep paragraphs around 20 to 80 words. |
Internal linking | Add cross-references to related topics (4 to 8 is ideal). |
Answer density | Lead with the answer; trim filler and hedging. |
Semantic completeness | Make sure the page answers the question its title implies. |
Fetch size | Split the page into focused topics under about 50,000 characters. |
When every page scores 7 or higher, the worklist reports that there is nothing left to fix.
See also
Triage content gaps: find content readers want but cannot find
AI queries and crawler activity: what readers ask and which AI crawlers ingest your docs
AI-ready published sites: the endpoints AI engines read
Published site SEO: schema.org markup and canonical URLs