AEO Content Scorecard

AI engines can extract the lead as a citation snippet without context-hunting deeper into the page.

The page may still be indexed, but the citation will likely go to a competitor whose lead paragraph is more extractable.

The H1 acts as a strong relevance signal for both Google and AI engines doing passage extraction.

The page is competing with itself — Google has to guess which query the page actually serves.

AI engines extract H2-level headings as candidate facets to cite or quote.

Decorative or marketing-style H2s ('Discover the magic of...') signal low information density.

Structured data extracts cleanly into AI engine summaries and Google AIO snippets.

Buried-in-prose stats are systematically harder to cite than the same stats in a bulleted list or table.

FAQPage schema becomes legitimately applicable; AI engines preferentially cite FAQ-marked answers.

If you fabricate FAQs to qualify for FAQPage schema, that's structured-data spam — Google demotes it.

The page can plausibly displace the existing top-3 in the SERP it's competing in.

Underbuilt content cannot displace overbuilt content, regardless of how clean the technical SEO is.

E-E-A-T signal is present and consistent. AI engines disproportionately cite authored content.

Anonymous content is increasingly demoted in AI citation ranking. 'By the Editor' is a red flag.

The Person entity consolidates across the site, building cumulative E-E-A-T weight.

Fragmented author entities dilute E-E-A-T across multiple synthetic authors.

Fact-checking signal is present in schema and visible on-page; YMYL content particularly benefits.

On YMYL queries (medical, legal, financial), the citation gap to fact-checked content is widening.

AI engines preferentially cite content that itself cites verifiable primary sources.

Uncited claims look like AI-generated filler — and are increasingly demoted in citation rankings.

Information gain is high; AI engines prefer to cite content that adds something not findable elsewhere.

Pure synthesis of competitor content is structurally lower in citation rate than original-data content.

Structured data is consolidated; AI engines can resolve the entity graph cleanly.

Schema fragmentation or absence caps citation eligibility regardless of content quality.

Question/answer pairs are individually citation-eligible by AI engines.

Eligible FAQ content is going un-cited.

Step-by-step content is eligible for procedural rich results and AI procedure citations.

Either eligible HowTo content is missing markup, or non-procedural content is being marked up (don't do that).

Image-search and AI-engine image understanding can use the page as a citation surface.

Image-driven citation surfaces (Google Lens, AI image queries) are leaving signal on the table.

AI engines reward reciprocal citation behaviour with higher in-citation rates.

Citation manipulation patterns ('source: Wikipedia' with no link) are increasingly detected.

ChatGPT-User, OAI-SearchBot, PerplexityBot, ClaudeBot, GoogleOther are not blocked at the path level.

Default Cloudflare and CMS hardening configs frequently block AI crawlers — you may be invisible.

AI crawlers (which often don't execute JavaScript) can read your content.

JS-rendered content is systematically under-cited by AI engines that don't render JS reliably.

AI crawlers (which have shorter timeouts than Googlebot) can complete the fetch.

Slow pages get systematically skipped from AI engine indexes.

Crawl frequency and citation eligibility both correlate with internal link depth.

Orphaned or deeply-buried pages cite at a fraction of the rate of well-linked pages of comparable quality.