Citegrove · AI brand visibility monitoring
Citability Score
ChatGPT, Perplexity, and Gemini don't cite whole pages — they pull specific paragraphs. The most-cited paragraphs are 134-167 words long, fact-rich, and self-contained. Paste any URL and we'll score how well-shaped your content is for AI citation, surface your top 3 most-citable passages, and tell you exactly what to fix.
FAQ
Frequently asked
What does "citability" mean?↓
How likely an AI engine is to quote a specific passage from your page when answering a user. AI engines prefer self-contained, fact-rich paragraphs of 134-167 words — long enough to make a complete point, short enough to quote whole.
Where does the 134-167 word band come from?↓
It's the band research from Anthropic citations docs, OpenAI passage-ranking benchmarks, and the geo-seo-claude project all converge on. Below 80 words a passage lacks context; above 220 it gets truncated.
How is the document score calculated?↓
50% from what fraction of your paragraphs land in the ideal 134-167 word band, 25% from heading structure (H1 + multi-level), 15% from JSON-LD schema presence, 10% from average passage length being workable.
What does the "fact density" number mean?↓
How many tokens per passage look fact-like — digits, percentages, dollar amounts, dates, named entities. Higher fact density = more citable. AI engines preferentially cite passages with concrete facts over pure prose.
Can you score a page behind a login?↓
No. We fetch via Jina Reader which only sees public content. Public docs, blogs, marketing pages all work; logged-in app pages don't.
Is this related to traditional SEO?↓
Adjacent. Traditional SEO optimizes for ranking on search-result pages. Citability optimizes for being quoted INSIDE AI answers. The two strategies overlap on schema and headings but diverge on passage construction.
Do you store the URL or content?↓
We log a hashed IP for rate limiting (10 audits per IP per 24 hours). URLs and content aren't persisted.