1. Entity footprint
Whether AI engines can recognise the page's underlying entity. Checks for Organization / Person schema, sameAs links to LinkedIn / Companies House / Wikidata, identifiers and logos. Without this, the engine has no entity to map citations against.
2. Schema layer
How many distinct schema.org types are present and whether they cover the page's actual content. AI engines extract from schema first, content second.
3. Content extractability
Whether the content is chunked into AI-friendly sections (150-280 words per H2 sweet spot), and whether tables / figures / structured elements are present that AI engines disproportionately quote.
4. Freshness
How recently the page was meaningfully updated. Industry analysis (PromptAlpha, Ziptie, Goodie AI) reports freshness drives 30 to 40 percent of Perplexity ranking and that content updated within 30 days gets 2 to 3x more AI citations than year-old content.
5. Named author / E-E-A-T
Whether the page has a named, verifiable author. Industry research (GEO Alliance, PresenceAI) reports a 2 to 3x Perplexity citation rate for named-expert authored content vs anonymous.
6. Citation hooks
Structures AI engines are biased toward quoting: comparison tables, blockquoted stats, FAQ blocks, speakable specifications. The more of these per major section, the higher the citation rate.