Not sure where to start? Take a guided tour.
AI Site Grader
AI Query Scanner
llms.txt GeneratorThe Anatomy of an AI Citation
What cited pages look like: schema, depth, authorship, and formatting patterns. Answer pages, reference facts, and live trackers drawn from this study.
Published 9 April 2026 · Updated 30 March 2026
22 results
Do AI-cited pages usually have schema markup?
Yes. 67.8% of cited pages in the study have schema markup, versus a 38.5% web-average benchmark, and the most-cited subset rises to 80%.
Does FAQ schema correlate with higher citation volume?
Yes, in this benchmark. Pages with FAQ Schema + FAQ Content average 36.9 citations versus 25.4 for pages with no FAQ signal, though the subgroup is small and should be read carefully.
How long are pages that get cited by AI?
They are usually long-form. Average word count is 2,289.6 words, and 78.4% of cited pages are above 1,000 words.
Which schema types over-index most on AI-cited pages?
Person schema over-indexes the most in the benchmark at 9.4x the web average, while FAQPage is also materially overrepresented at 2.4x.
Do the most-cited pages look more structured than the average web page?
Yes. They are much more likely to use schema, canonical tags, Open Graph tags, and long-form content than the average page on the web.
Does authorship look important on cited pages?
Yes. Person schema is the most overrepresented schema type in the benchmark at 9.4x the web average, which suggests cited pages often make authorship explicit.
Are canonical and Open Graph tags common on cited pages?
Yes. 91.4% of cited pages in the study have canonical tags and 89.2% include Open Graph tags.
Is schema enough on its own to earn AI citations?
No. Schema is common among cited pages, but the benchmark also shows that cited pages are long, structured, and rich in answerable content. Markup helps, but it does not replace substance.
What does the benchmark say about page design for AI citations?
It says citeable pages are explicit, structured, and evidence-dense. They tend to be long enough to resolve a topic and formatted so a model can extract answers confidently.
What is the most useful way to interpret the schema data?
Interpret it as a benchmark profile rather than a magic lever. The schema data shows what cited pages tend to have in common, which helps teams design better page templates even when it does not prove strict causality.
Why does content density show up so strongly in the citation benchmark?
Because dense pages are better at resolving ambiguity. The benchmark’s 2,289.6 average word count and 78.4% share above 1,000 words show that cited pages usually answer the full question, not just part of it.
What should you copy from pages that already win citations?
Copy the structure, not the brand halo. Use explicit answers, dense but skimmable sections, visible authorship, clear metadata, and supporting schema where it genuinely helps parseability.
More than two thirds of cited pages have schema
An analysis of 1465 AI-cited pages from The Anatomy of an AI Citation study reveals that 67.8% of these pages implement schema markup.
The most-cited pages hit an 80% schema rate
In the Anatomy of an AI Citation study, analysis of the highest-cited subset of pages reveals an 80 percent schema prevalence rate.
FAQ schema plus FAQ content pages average 36.9 citations
The FAQ-structured subgroup performs strongly in the benchmark.
Pages with no FAQ signal average 25.4 citations
The no-FAQ baseline is materially lower than the FAQ-structured subgroup.
Average word count of cited pages is over 2,000 words
Cited pages are usually dense enough to resolve the topic fully.
Nearly four in five cited pages are over 1,000 words
Long-form depth is the norm in the benchmark.
Canonical tags appear on over 90% of cited pages
Metadata hygiene is common in the cited-page benchmark.
Person schema is the most overrepresented schema type
In The Anatomy of an AI Citation study, analysis of structured data reveals that Person schema is the most overrepresented schema type on cited pages compared to the general web baseline.
Benchmark profile of AI-cited pages
Benchmark profile of AI-cited pages from the study The Anatomy of an AI Citation.
Schema types overrepresented on cited pages
Benchmark data tracking the prevalence of specific schema types on AI cited pages compared to the general web.