Structured Data

Structured data lets you spell out what a page means in a format machines read without guessing: this is a product, that is its price, this is the author, that is the rating. It earns you rich results and helps engines extract facts. What it does not do is raise your rankings, and confusing those two is the most common structured-data mistake.

What structured data is

Structured data is machine-readable markup describing a page's content using the schema.org vocabulary. Google's recommended format is JSON-LD, a <script type="application/ld+json"> block you drop into the page. Microdata and RDFa are alternatives that inline the markup into the HTML, but JSON-LD keeps the annotation separate from the visible markup, which is easier to manage, so it is the default choice.

The mental model: your prose is written for humans, and Google parses it as best it can. Structured data is a second, explicit channel that states the same facts in a form a machine cannot misread. It removes ambiguity for the data points that matter most.

JSON-LD's separation from the visible markup is a practical advantage. Because the block sits in its own script rather than being woven through your HTML, you can generate it from the same data that renders the page, template it, and update it without disturbing layout. Microdata and RDFa force the annotations into the visible tags, which is harder to maintain and easier to break during a redesign. For almost every case, JSON-LD is the format to reach for.

What it does and does not do

Structured data makes a page eligible for rich results: review stars, FAQ accordions, recipe cards, product prices, event details, breadcrumbs. It also helps engines extract facts cleanly (price, availability, author, dates) and clarify entities and their relationships for knowledge-graph understanding.

Here is the boundary. Structured data is not a direct ranking factor. It can win rich results, which often lift click-through because the listing takes more space and looks more useful, but the markup itself does not move you up the results. It does not rescue thin content, and it must match the visible content on the page. Mark up a rating the page does not show and Google can ignore the markup or treat it as a policy violation. One caution on volatility: Google reduced FAQ and HowTo rich results for most sites in 2023, so treat those two as unreliable sources of extra visibility rather than a guarantee. That change is a moving target, not a fixed rule.

Structured data buys eligibility for rich results and cleaner fact extraction, not rankings. It must mirror what the page actually shows, or Google can disregard it.

The types worth knowing

Most sites need only a handful. The common ones:

• Article for blog posts and news.
• Product with Review and AggregateRating for ecommerce.
• FAQPage and HowTo, with the 2023 reduction caveat above.
• BreadcrumbList for the breadcrumb trail in results.
• Organization and LocalBusiness for entity and local signals.
• Event, Recipe, and Video for their respective content.

Match the type to what the page genuinely is. Organization and Person markup in particular feed how engines understand your brand as an entity, which is central to entity-level optimization. Entity schema for brand and author → On-page markup is one piece of the broader on-page toolkit. On-page SEO →

Structured data and AI

For Google's own AI features, schema is not required to appear; Google has said you do not need to create special markup for AI surfaces. Where structured data clearly helps is data that is hard to read from prose, price, stock, shipping, where an explicit field beats a sentence a model has to parse. Beyond that, the story is mixed. Studies conflict on whether structured data correlates with being cited by AI answer engines, and Google's own spokespeople describe the benefit as "it depends" by feature.

So treat the AI-citation benefit as plausible and low-cost, not proven. Machine-readable facts are cheap to add and can only help a machine that is trying to read you, which is reason enough for the data-heavy types. Just do not build your AI-visibility strategy on schema alone. Machine-readable signals for AI →

Validate it

Use two tools, because they answer different questions. The Rich Results Test tells you whether Google considers the page eligible for a specific rich result. The Schema Markup Validator at validator.schema.org checks general validity against the vocabulary, whether your markup is well-formed regardless of Google's rich-result support. Run both.

One rendering check applies here as everywhere: if your structured data is injected by JavaScript, confirm it survives in the rendered HTML, because markup that only exists after a script runs is markup Google might miss. After deploying, monitor the Enhancements reports in Search Console for errors Google finds in the wild. Next, we move from describing pages to helping engines discover them at scale. Sitemaps next →