Does schema markup help you get cited by AI?
There is no verified evidence that JSON-LD schema markup increases citations in AI answers, in either direction. It is widely recommended and cheap to implement, but our review of the published research found no controlled study establishing an effect on AI citation rates specifically. Treat it as low-cost structural hygiene that helps classic search features — not as an AI visibility lever.
The same absence of evidence applies to several other commonly recommended technical measures, including llms.txt files, server-side rendering choices and content chunking. Unproven is not the same as disproven, but neither justifies prioritising them over measured levers.
One technical factor is worth auditing because it is a fact rather than a correlation: crawler access. If robots.txt blocks GPTBot, PerplexityBot, ClaudeBot or Google-Extended, those engines cannot cite pages they are forbidden to fetch. This is common on client sites and usually unintentional.
If a vendor tells you schema markup "dramatically improves extractability" for AI, ask which study. We removed that exact claim from our own product after the research did not support it.
Sources
Last reviewed 2026-08-06. We revisit these when the underlying research changes.
Related
See where you actually stand
Run a free check across ChatGPT, Perplexity, Gemini, Claude and Google AI Mode.
Check your AI visibility