Definition:
Schema markup is an implementation of structured data that uses the Schema.org vocabulary to describe entities, properties, and relationships found on a web page. It helps search engines and other systems interpret explicitly what each piece of data represents.
Schema.org is the shared vocabulary; structured data is the organised information; and markup is the code used to add it to a page. These concepts are related but not identical, even though they are often used interchangeably.
Índice de contenidos
How it works
Markup creates a machine-readable representation of elements that should also correspond with visible content. For example, a product page can identify its name, brand, image, offer, and availability; an article can declare its headline, author, and publication date.
Each element is expressed through types and properties. A type defines the entity class, such as Product, Article, or Organization. Properties describe its attributes or relationships, such as name, author, offers, or address. Entities can be connected to form a coherent graph.
This description complements HTML, but it does not correct inaccurate content or replace understandable architecture. Markup must represent what actually exists on the page and maintain visible consistency.
How to implement it
Schema.org can be expressed through JSON-LD, Microdata, or RDFa. Google generally recommends JSON-LD where possible because it keeps the structured block separate from visual markup and is often easier to generate, review, and update.
In JSON-LD, entities are commonly included in an application/ld+json script. Microdata adds attributes directly to HTML elements, while RDFa uses HTML-compatible attributes to describe resources and relationships. All three formats can be valid when their syntax is correct.
Implementation may be manual or generated through a template, plugin, or CMS. Large websites should produce markup from a reliable source, such as the product, event, or author database, to avoid discrepancies between displayed content and structured values.
Common types
Schema.org contains many types, but selection should reflect the real entities on each page. The following are among the common types:
- Organization: Identifies an organisation, its name, URL, logo, and corporate relationships.
- LocalBusiness: Describes a local business and may include its address, telephone, opening hours, or service area.
- Product: Represents a product and can connect it with offers, brand, identifiers, and availability.
- Article: Describes articles, news stories, or posts through properties such as headline, author, and image.
- BreadcrumbList: Expresses the hierarchical navigation sequence leading to the page.
- Event: Identifies an event, its dates, location, attendance mode, and ticket offers.
- JobPosting: Represents a job vacancy with its role, organisation, location, and conditions.
- VideoObject: Describes a video, its thumbnail, duration, date, and content location.
- Recipe: Structures a recipe with ingredients, times, instructions, and yield.
- DefinedTerm: Defines a term and can connect it with a reference set or glossary.
Choosing the most attractive type is not enough. A page should use the specific type that matches its main entity and include only truthful properties. A plugin that inserts many nodes does not guarantee a better representation.
Rich results
Search engines may use certain structured data to generate special presentations known as rich results. Eligibility depends on the type, required properties, search engine policies, and visible content.
Valid markup does not guarantee display. The search engine decides whether to show a rich result according to the query, device, page quality, and other factors. Some Schema.org classes also have no specific rich feature in Google while still describing an entity correctly.
Markup is not by itself a direct ranking factor that ensures better positions. It can support interpretation and make a page eligible for certain features, but it cannot replace relevance, quality, links, or technical accessibility. Its relationship with semantic SEO lies in the explicit description of entities and relationships, not an automatic visibility promise.
How to validate it
Validation should check both syntax and meaning. The Schema Markup Validator reviews general vocabulary usage, while Google Search documentation and tests focus on requirements for its rich features.
A technical review can follow these steps:
- Check syntax: Verify that the code can be parsed and uses valid types and properties.
- Review entities: Confirm that nodes represent real elements and connect through consistent identifiers.
- Compare content: Ensure that names, prices, dates, images, and statuses match visible information.
- Assess requirements: Review required and recommended properties and policies for the corresponding rich result.
- Monitor changes: Detect errors after modifying templates, plugins, catalogues, or content models.
Reports in Google Search Console may show errors and improvements for supported types, but they do not replace an audit of the complete graph. Correct markup should be valid, truthful, maintainable.
When a URL, image, author, price, or other structured value changes, its representation must also be updated. Schema quality depends on keeping the code synchronised with content and avoiding duplicate or contradictory nodes.