What is structured data?
Structured data is code you add to a page that describes its content in a standardized way, so search engines and AI can understand exactly what the page is about. Instead of guessing from the text, an engine reads labelled facts, for example that this is an article, by this author, published on this date, about this topic. Most structured data uses the shared Schema.org vocabulary, written in a format called JSON-LD.
Where plain HTML tells a browser how to display content, structured data tells machines what the content means. That distinction has become critical in 2026, because the same signals that earn rich results in Google now help AI engines decide what to cite.
How structured data works (JSON-LD and Schema.org)
Structured data is expressed with the Schema.org vocabulary, a shared dictionary of types like Article, Product, Organization, and Review. The preferred format is JSON-LD (JavaScript Object Notation for Linked Data), a small script you place in the page head that lists the facts about the page as name-value pairs. Google, Bing, Perplexity, and ChatGPT all rely on JSON-LD to extract structured signals (Digital Applied, 2026).
The schema types that matter most in 2026 are Organization, Article or BlogPosting, Product, LocalBusiness, and FAQPage (with the caveats below). Mark up what genuinely appears on the page; do not describe content a user cannot see.
Why structured data matters for SEO and AI search
Structured data does two jobs. First, it unlocks rich results: the stars, prices, and other enhancements that make a listing stand out and win more of the SERP features. Pages with structured data get about 30% more clicks than plain results (BrightEdge). Second, and increasingly, it makes a page machine-readable for AI. With AI Overviews now appearing on 50 to 60% of US searches (Digital Applied, 2026), the entity signals in your schema help engines like Gemini, ChatGPT, and Perplexity recognise your brand and cite your page. Clear structured data also reduces the factual ambiguity that makes AI skip or misattribute a source. It is a core part of generative engine optimization, and it pairs with fundamentals like headings and crawlability.
How to implement structured data
- Choose the right type. Pick the Schema.org type that matches the page (Article, Product, Organization, and so on).
- Use JSON-LD. Add it as a script in the page head; it is the easiest format to maintain and the one engines prefer.
- Describe what is really there. Only mark up visible, accurate content, or you risk a manual action.
- Test it. Validate with Google's Rich Results Test and the Schema.org validator before and after publishing.
- Keep it current. Update dates, prices, and details as the page changes.
Structured data and FAQ rich results (2026 update)
Google has wound down some rich result types. FAQ and HowTo rich results are no longer shown for most sites, so FAQPage markup rarely produces a visible enhancement now. A visible FAQ section on the page still helps users and gives AI clean question-and-answer passages to cite, so keep the FAQ; just do not rely on FAQPage schema to earn a rich result. Focus schema effort on Organization, Article, and Product, where the payoff is real.
Common structured data mistakes to avoid
- Marking up hidden content. Schema must describe what is visible on the page.
- Wrong or bloated types. Use the type that fits; do not stuff irrelevant schema.
- Never validating. Small syntax errors stop rich results entirely.
- Set and forget. Stale schema, such as old prices or dates, can mislead engines.
Structured data FAQs
What is structured data in simple terms?
It is a small block of code that labels the facts on a page (what it is, who wrote it, what it costs) in a standard format, so search engines and AI can understand the page precisely instead of guessing from the words. It usually uses the Schema.org vocabulary written in JSON-LD.
What is the difference between structured data and schema markup?
They are used almost interchangeably. Structured data is the general concept of labelling content for machines; schema markup is structured data written with the Schema.org vocabulary specifically. In practice, when people say schema markup they mean structured data based on Schema.org.
What is JSON-LD?
JSON-LD (JavaScript Object Notation for Linked Data) is the recommended format for structured data. It is a small script, usually placed in the page head, that lists facts about the page as name-value pairs. Google and the major AI engines prefer it because it is easy to read and keeps the markup separate from the visible HTML.
Does structured data help with AI search?
Yes. The entity signals in your schema help AI engines recognise what and who a page is about, which supports citations in AI Overviews and chatbots and reduces factual ambiguity. As AI Overviews now appear on a majority of searches, structured data has become a core part of generative engine optimization.
Do FAQ rich results still work in 2026?
Mostly no. Google has discontinued FAQ (and HowTo) rich results for most sites, so FAQPage markup rarely shows a visible enhancement. Keep a visible FAQ section for users and AI, but focus schema effort on types that still earn results, such as Organization, Article, and Product.
How do you test structured data?
Use Google's Rich Results Test to see whether a page is eligible for rich results, and the Schema.org validator to check the markup is valid. Test before and after publishing, and re-check when you change prices, dates, or other marked-up details.