What is structured data?

Structured data adds a machine-readable description to a page. It lets a system distinguish a company name from ordinary text, associate a product with its offer, identify an article's author, or reconstruct a breadcrumb trail.

The markup does not replace the page or create information that is not already there. It makes published facts and relationships less ambiguous for search engines and other services that process web content.

How do structured data, Schema.org, and JSON-LD differ?

Structured data is the general concept. Schema.org supplies the vocabulary of types and properties, such as Organization, Article, name, and datePublished. JSON-LD is one format used to publish that vocabulary on a page.

Microdata and RDFa embed properties into HTML. JSON-LD keeps the data in a separate block and is often easier to generate and maintain. In precise English usage, Microdata is one syntax rather than a synonym for all structured data.

What does a simple JSON-LD object look like?

This abbreviated object describes an organization. A production implementation should draw its values from the same source used for the visible page.

JSON-LD
Documentation 
{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Company name",
  "url": "https://example.com"
}

The object is normally placed inside a script element with the type application/ld+json. Visitors do not see it as a separate content block, but a system can read the entity type and its properties.

Which structured data type belongs on each page?

The type should follow the page's real purpose rather than the search appearance a publisher hopes to obtain. One website will normally use several types across different sections.

  • Organization describes a company, while WebSite describes the site.
  • BreadcrumbList represents the breadcrumb sequence.
  • Article suits editorial content with an author and publication date.
  • Product and Offer describe a product and its price, currency, availability, and other offer details.
  • Service can describe a service, although it does not create a dedicated Google rich result by itself.
  • FAQPage should be used only for visible questions and answers and according to the current rules of the target search engine.
  • DefinedTerm and DefinedTermSet suit an individual definition and a knowledge base.

How can structured data affect search appearance?

A search engine may use the markup to understand the page and power supported rich results. Valid markup does not guarantee a rich result, a ranking, or indexing. Schema.org includes more types than any individual search engine supports, so a valid object may describe a page correctly without producing a special Google feature.

Why must markup agree with visible content?

A site should not mark up an invented rating, a hidden review, an outdated price, or details that a visitor cannot find on the page. Structured data is not a second promotional version of the content.

Dynamic properties should come from the same source as the visible values whenever possible. This keeps prices, dates, authors, availability, and addresses synchronized and avoids conflicting descriptions of the same entity.

Which implementation mistakes are common?

  • choosing a type for the desired rich result rather than the actual page;
  • omitting a required property or using the wrong value format;
  • marking up information that is not visible to the visitor;
  • letting prices, dates, availability, or addresses become stale;
  • publishing conflicting JSON-LD objects from several plugins or components;
  • checking syntax without reviewing the meaning of the entity relationships.

How is structured data validated?

Schema Markup Validator checks an object against the Schema.org vocabulary. Google Rich Results Test focuses on markup connected to Google-supported rich results. Search Console can report issues and appearances for supported types after publication.

Passing a syntax test is only the first check. The type still needs to fit the page, each property must have the intended meaning, and every value should agree with the visible content.

Structured data can make a company, author, article, or product easier to identify. There is no special “AI schema,” however, and Schema.org markup cannot guarantee a citation in a generative answer.

Google says its normal search requirements apply to AI Overviews and AI Mode and that no special schema.org markup or AI file is required. Structured data should be implemented to describe the site accurately, not as a shortcut into an AI response.

Sources