What does structured data do for AI search, and what doesn’t it?
Structured data states a business’s facts in a form machines read without guessing, which helps search engines tie your website, profiles and listings to one entity and can earn rich results in Google. It is not a reliable way to get cited by AI assistants.
In an Ahrefs study published in May 2026, 1,885 pages that added JSON-LD were compared with matched pages that did not. Citations in ChatGPT and Google AI Mode changed by amounts indistinguishable from zero. A companion test cited in the study found that ChatGPT, Claude, Perplexity, Gemini and Google AI Mode read the visible HTML when fetching a page live and ignored JSON-LD. Every page studied was already heavily cited, so the study cannot say whether markup helps a page AI systems do not yet see.
The practical rule: put every fact in the visible text of the page, and treat the markup as a parallel, machine-readable copy of it. Google asks that the two match.
What does this generator build?
A single @graph containing your business node, one Person node per person with worksFor pointing back to the business, and one Service node per service with provider pointing back to the business. Identifiers are built from your website address, such as https://www.example.com/#organization, so other pages can reference the same nodes instead of repeating them.
What are the common schema mistakes?
- Names that differ from your listings. “Marlow Reyes Realty” in markup and “Marlow & Reyes Realty LLC” on the licence record undoes the point of the markup.
- sameAs links to profiles you have not checked. A directory page showing an old address tells systems your own markup disagrees with itself.
- Review stars about yourself. Ratings a business marks up about itself on its own site are not eligible for review rich results in Google.
- Duplicate @id values across templates, which happens when markup is copied without changing identifiers.