What structured data is, and what it does for a page

Search engines read pages the way a very fast, very literal person would: words, links, headings. Structured data is a second layer, written for machines, that says what the page is — a product with this price, a business at this address open these hours, a recipe taking this long — using a vocabulary called schema.org. It's how results get star ratings, FAQ dropdowns, event dates and prices in the listing. This guide explains the pieces; the schema generator writes the code.

What structured data is

Schema.org is a shared vocabulary (launched in 2011 by Google, Bing, Yahoo and Yandex) of types — Product, Article, Event, Person, Organization, LocalBusiness, FAQPage, HowTo, Recipe and hundreds more — each with named properties. Marking up a page means stating, in that vocabulary, the facts already visible on it: this Product has name "X", offers price 9.99 in GBP, availability InStock. Search engines use those facts to understand the page, to build knowledge panels, and to decorate results. It doesn't improve ranking directly; it improves how the result looks and how well the engine understands what to rank you for, which in practice raises clicks.

JSON-LD: the format Google prefers

There are three syntaxes. Microdata and RDFa wrap the visible HTML in attributes, which is fragile and hard to maintain. JSON-LD is a script block — <script type="application/ld+json"> — containing a JSON object with @context (schema.org) and @type, placed anywhere in the page and independent of the layout. Google recommends it, it's what content systems inject, and it's what you can paste in without touching the design. Nesting expresses relationships: an Offer inside a Product, a PostalAddress inside a LocalBusiness, a Person as an Article's author. What JSON is, and where it trips people up covers the syntax if it's new.

The types that earn rich results

• FAQPage — questions and answers on the page can appear as expandable dropdowns under the result (Google now limits these to authoritative government and health sites, but the markup still helps understanding).
• Product with Offer and, if genuine, AggregateRating — price, availability, stars in the listing.
• Article / NewsArticle / BlogPosting — headline, author, dates, image; feeds Top Stories and Discover eligibility.
• LocalBusiness (or a subtype like Restaurant, Dentist) — address, hours, phone, price range; feeds the map panel.
• Event — dates and venue as an event listing.
• HowTo and Recipe — steps, times, ingredients.
• Organization and Person — for the site as a whole: logo, social profiles, the author behind the bylines.

Rules and the ways it goes wrong

The one rule that matters: markup must reflect content visible on the page. FAQs that aren't shown, reviews that aren't displayed, an availability you don't have, a recipe on a page that isn't a recipe — Google's structured-data guidelines call this spam and the penalty is a manual action removing rich results site-wide. The technical failures are quieter: a missing required property (a Product without an Offer), a date in the wrong format (use ISO 8601: 2026-09-16), a price as "£9.99" instead of a number and a currency, or a JSON syntax error that silently disables the whole block. Generators exist to prevent the second kind; only honesty prevents the first.

Testing and what to expect

Paste the page URL or the code into Google's Rich Results Test — it reports errors, warnings and which enhancements the page is eligible for. Then wait: rich results appear when Google recrawls, and only when it chooses to; eligibility is not a guarantee, and Google shows them per query. Track impressions and click-through in Search Console's enhancement reports. Alongside markup, the other half of how a page is understood is its text: the title (How long should a title tag be? Pixels, not characters), the description and clear headings. The meta tag generator handles those; the sitemap generator helps the crawler find the page at all.