Structured Data (Schema.org / JSON-LD)

Structured data is a machine-readable description of a page's content that tells search engines and other systems what the content means - that this is a recipe, this is a product with a price, this is an article with an author and a date, this is a business with an address and opening hours. The vocabulary standard is Schema.org, a joint project of Google, Microsoft, Yahoo, and Yandex. The recommended format today is JSON-LD, a block of JavaScript in the page head that's separate from the visible content and easy to maintain.

Why do it: without structured data, a search engine has to guess the meaning from the text. With it, the meaning is explicit. That's the prerequisite for rich results - star ratings in search, product prices, FAQ dropdowns, breadcrumb navigation, event cards. These elements take up more space in the results and lift click-through rates, even though they don't directly change ranking.

The most-used types: Article and BlogPosting for content, Product and Offer for e-shops, LocalBusiness and Organization for companies, FAQPage for common questions, HowTo for guides, BreadcrumbList for navigation. The sameAs property matters too - it links an entity to its profiles elsewhere (Wikipedia, LinkedIn, social networks), which helps both search engines and AI models correctly identify who or what you are.

A few rules from practice: structured data has to match what's actually visible on the page, or you risk a penalty for misleading markup. It's worth checking in Google's Rich Results Test. And in the age of AI search it has a double benefit - it helps not only classic rich results but also helps language models understand and cite you correctly.