Study · 2026
The state of llms.txt & AI-citability
We scanned 57 homepages of leading SaaS and tech companies with our citability engine. llms.txt adoption has crossed the majority — but the Q&A structured data that actually gets a page quoted word-for-word is almost entirely missing.
have an llms.txt file
have schema.org markup
have Q&A markup (FAQ/QA)
average citability score /100
What the leaders have — and don't
The real gap: Q&A markup
An llms.txt helps AI read your site. But what makes them quote you word-for-word is Q&A schema.org markup (FAQPage / QAPage). Only 4% of the sites scanned have any — even among those that already have an llms.txt and generic schema. That's exactly the piece Citeable adds.
The most citable in the sample
Methodology
Sample: 60 homepages of recognized SaaS / dev-tool / no-code / AI companies, scanned on 2026-07-13 (57 reachable; 3 blocked by an anti-bot firewall). Engine: the same deterministic checks as Citeable's free scan (llms.txt quality, schema.org presence, meta description, heading structure, content depth). “Q&A markup” = a FAQPage, QAPage or Question @type present in the homepage JSON-LD. Nothing is estimated: every value comes from a real crawl.
Is your site citable? Free scan→