Does schema help AI citations?
No reliable evidence shows that schema markup makes AI apps cite you. In a matched Ahrefs study published 11 May 2026, adding JSON LD produced no citation growth in Google AI Mode or ChatGPT, and Google's own guide says no special schema is needed for its AI features. KailxLabs treats schema, llms.txt and FAQ markup as hygiene, never as the reason to buy.
What does the evidence say?
Five sources, from a vendor study to academic work. None shows schema getting a page cited.
| Source | Date | Sample | Finding | Type |
|---|---|---|---|---|
| Ahrefs, matched study | 11 May 2026 | Pages that added JSON LD against matched pages that did not, over 30 days; pages already had 100 or more AI Overview citations | No citation growth in AI Mode or ChatGPT. In AI Overviews, pages with new schema did 4.6% worse than their matches. | Vendor study |
| Google, AI optimization guide | Updated 10 Jul 2026 | Official guidance | No special schema is needed for generative AI search. Search ignores llms.txt and other AI text files. | Platform |
| searchVIU live fetch test, cited by Ahrefs | May 2026 | Five AI systems fetching pages live | All five read the visible HTML and ignored JSON LD. | Third party |
| CITECHOICE, arXiv 2609.15164 | 14 Sep 2026 | 113 pairs of documents with the same facts, controlled replay | A structured version of the same content drew more citations per answer once retrieved (0.50 more). Whether a source got cited at all was inconclusive. | Academic |
| Martinez, arXiv 2607.14035 | 15 Jul 2026 | Review of 45 studies | No reviewed technique showed a stable, cross platform causal effect on being found over time. | Academic |
The Ahrefs study had limits: a 30 day window, schema types pooled, and pages that were already cited often. Treat it as strong evidence of no lift, not proof of harm.
What about llms.txt and FAQs?
Two more pieces of markup are still sold as AI levers. Both have been answered by Google.
Google Search ignores llms.txt and other AI text files. It helps coding assistants read developer docs, and little else.
Google, AI optimization guide · updated 10 Jul 2026, checked 26 Sep 2026
FAQ rich results were removed on 7 May 2026. People Also Ask answers are now largely AI written.
Google Search Console data anomalies page · 7 May 2026. People Also Ask: AlsoAsked and Allintitle, reported 9 Sep 2026 · third party
When is schema worth it?
Schema is hygiene. Hygiene still matters, just for different reasons.
Product and offer markup that matches your feed and visible price supports shopping results and Merchant Center.
Video markup still earns video features in Google results.
Markup that repeats your visible facts does no harm. Markup that contradicts them adds one more wrong fact for AI to find.
What moves citations instead?
Retrieval comes first: an AI can only cite a page it found and read. These steps change what it finds.
- Crawler access for the AI search crawlers, at robots.txt and CDN level. Which crawlers to allow.
- Facts in visible text: prices, services and policies on the page, not only in markup or images.
- A place in the sources AI cites: third party lists, reviews and comparisons for your category.
- Real updates: fresh substance, not a changed date.
Hygiene included, levers first
Answer ready build
$4,995 one time. Pages built to be quoted, crawler access and consistent facts, with valid schema as hygiene.
Answer ready websiteAI visibility baseline
$1,500 one time. The sources each AI app cites for your buyer questions.
About the baselineCitation engineering
From $2,000 a month. Placements in the pages AI already cites.
Citation engineeringSchema questions, answered.
Should I remove my schema markup?
No. Keep valid schema that matches your visible text. It still supports product and video rich results and Merchant Center. Just do not pay for it as an AI citation lever.
Does Google use llms.txt?
No. Google's AI optimization guide, updated 10 Jul 2026, says Google Search ignores llms.txt and other AI text files. It can still be useful for developer documentation that coding assistants read.
Do FAQ blocks still help?
FAQ rich results ended on 7 May 2026, so FAQ markup no longer earns a special search result. A FAQ still helps if it answers questions buyers really ask, in visible text.
Does KailxLabs add schema?
Yes, as hygiene inside the $4,995 answer ready build: valid, matching the visible page, and never sold as the reason AI will cite you.
- 26 Sep 2026: page published.
Talk to the founder.
20 minutes with Kailesk Khumar. You leave knowing where AI names you today, and what it would take to become the answer.
- A live look. We ask five of your buyer questions on ChatGPT, Google and Perplexity while you watch.
- The reasons. The three biggest reasons AI names a rival instead of you.
- The plan. A clear next step and its price, whether or not you hire us.
Pick a time that suits you. The calendar invite arrives straight away. Prefer a link? Book on Cal.com.