What is Schema.org? The Vocabulary Behind All Structured Data

Last Updated: February 25, 2026 · 11 min read

When you add structured data to your website, you are not just writing code — you are speaking a specific language that search engines understand. That language is defined by Schema.org. Created jointly by Google, Bing, Yahoo, and Yandex in 2011, Schema.org is the shared dictionary that defines what terms like "Product", "Recipe", "ratingValue", and "author" mean in the context of web content.

2011
Year Schema.org launched
900+
Types defined
Google + Bing + Yahoo
Founding search engines

1. The Analogy: Schema.org as a Dictionary

Imagine you want to describe a book to someone who speaks a different language. You both agree in advance that the word liber means "book", auctor means "author", and paginae means "number of pages". Now you can communicate about books unambiguously.

Schema.org does exactly this for web content and search engines. It defines agreed-upon terms — called Types and Properties — so that when you write "@type": "Recipe" and "cookTime": "PT30M", every search engine that speaks Schema.org knows what you mean.

2. Types and Properties: The Two Building Blocks

Types (the "what")

A Type is a category of thing. Assigned via the "@type" property.

ProductRecipeArticleEventPersonOrganizationLocalBusinessJobPosting

Properties (the "details")

Properties describe specific attributes of a Type. Each Type has its own set of valid properties.

name — the name of the thing
description — a summary
image — an image URL
url — the canonical URL
author — who created it
datePublished — when published
price — for products/offers
ratingValue — numeric rating

3. Schema.org Type Hierarchy

All Schema.org types inherit from a root type called Thing. More specific types inherit all the properties of their parent — for example, BlogPosting inherits all properties from Article, which inherits from CreativeWork, which inherits from Thing.

Thing
CreativeWorkEventIntangibleOrganizationPersonPlaceProduct
CreativeWork
ArticleBookMovieMusicRecordingRecipeSoftwareApplicationVideoObjectWebPage
Article
BlogPostingNewsArticleTechArticleSocialMediaPosting
Organization
LocalBusinessCorporationEducationalOrganizationNewsMediaOrganization
LocalBusiness
FoodEstablishmentHealthAndBeautyBusinessLodgingBusinessStore

4. Schema.org vs JSON-LD: What's the Difference?

Schema.org = the vocabulary

Defines what the types and properties mean. Schema.org is a website (schema.org) and a specification — not a piece of code.

JSON-LD = the format

Defines how you write the structured data. JSON-LD is the coding syntax (based on JSON) used to express Schema.org vocabulary in your HTML.

Analogy: Schema.org is English grammar rules; JSON-LD is the pen you use to write in English. You need both.

5. Which Schema.org Types Does Google Support for Rich Results?

Schema.org TypeRich ResultCTR Benefit
RecipeRecipe card with image, time, rating🔥 +40–50%
Product + OfferPrice, availability, star ratings🔥 +30–35%
FAQPageExpandable Q&A dropdowns⭐ +20–30%
HowToNumbered step previews⭐ +15–25%
JobPostingGoogle Jobs listing⭐ High
EventEvent carousel listing⭐ +20–25%
VideoObjectVideo thumbnail + duration⭐ +25–40%
BreadcrumbListClean path in snippet✅ +5–10%
SoftwareApplicationApp ratings + price✅ Medium
CourseCourse listing with provider✅ Medium

6. How to Look Up Any Schema.org Type

1

Go to schema.org/[TypeName] — e.g. schema.org/Recipe, schema.org/Product, schema.org/Person

2

The page lists all properties for that type, inherited properties from parent types, and expected value types for each property

3

Cross-reference with Google's rich results documentation at developers.google.com/search/docs/appearance/structured-data to see which properties Google requires vs recommends

4

Build your JSON-LD using required + recommended properties, then validate at schemavalidator.org

7. How Schema.org Evolves: Who Controls It?

Schema.org is governed by a W3C Community Group — an open collaborative group that includes Google, Microsoft, Yahoo, Yandex, and community members from academia and industry. Understanding this process helps you anticipate what is coming and why certain types exist:

Proposal

Anyone can propose a new type or property by opening an issue in the Schema.org GitHub repository (github.com/schemaorg/schemaorg). Proposals typically include a use case, example markup, and evidence of usage.

Discussion

Community members and Schema.org stewards discuss the proposal. Competing approaches are considered. Proposals from major platforms (Google, Microsoft) often progress faster because they come with implementation commitments.

Pending Status

Accepted proposals move into https://pending.schema.org — a staging area for new types that are approved in principle but not yet fully stable. You can use pending types in production, but they may change.

Release

Stable types graduate to the main schema.org namespace. Schema.org has a public changelog at schema.org/docs/releases.html. Major releases are numbered (currently at 27.x as of 2026).

8. Critical Properties You Should Know But Often Miss

Beyond the obvious type-specific properties, three Schema.org properties appear in advanced implementations across every schema type:

{
  "@context": "https://schema.org",
  "@type": "Article",
  "@id": "https://example.com/article-slug#article",
  // @id creates a globally unique identifier for this entity.
  // It anchors the entity in Google's Knowledge Graph.
  // Use a fragment URL (#article, #org, #person) to distinguish
  // multiple entities on the same page URL.

  "sameAs": [
    "https://en.wikipedia.org/wiki/Author_Name",
    "https://www.wikidata.org/wiki/Q12345",
    "https://twitter.com/authorhandle"
  ],
  // sameAs links this entity to authoritative external representations.
  // Google uses sameAs to confirm you are talking about a known
  // real-world entity, not just a name string. It is the most
  // important property for entity disambiguation.

  "mainEntityOfPage": "https://example.com/article-slug"
  // mainEntityOfPage declares that THIS page primarily exists to
  // describe this entity. Use it on your most important entities.
  // Helpful for Google to understand page-entity relationships.
}

9. Common Misconceptions About Schema.org

❌ Myth: Schema.org markup is only for Google

✅ Reality: False. Bing, Yandex, Yahoo, Pinterest, LinkedIn, and Slack all use Schema.org structured data. More importantly, AI systems including ChatGPT, Perplexity, and Google's AI Overviews use Schema.org signals to understand your content.

❌ Myth: If a type exists on Schema.org, Google supports it

✅ Reality: False. Google supports a subset of Schema.org types for rich results. See schema.org/docs/gs.html for Google's supported types. Many legitimate Schema.org types (like SoftwareSourceCode, Dataset) are crawled and understood but do not generate a visible rich result.

❌ Myth: You need to implement every property of a type

✅ Reality: False. Schema.org types have required fields (for Google to generate rich results), recommended fields, and optional fields. Only required properties are mandatory. Adding more recommended properties improves the quality of rich results but is not required.

❌ Myth: Schema.org and JSON-LD are the same thing

✅ Reality: Different layers. Schema.org is the vocabulary — the list of types and properties. JSON-LD is the serialisation format — how you express Schema.org vocabulary in code. You can also express Schema.org using Microdata or RDFa, but JSON-LD is the recommended format.

Frequently Asked Questions

Is Schema.org free to use?

Yes. Schema.org is an open community project. The vocabulary, documentation, and all schemas are freely available under a Creative Commons licence. There is no cost to use Schema.org types.

Do I need to implement every Schema.org property for a type?

No. Schema.org types have required, recommended, and optional properties. You only need required properties to be eligible for rich results. Adding recommended properties improves quality but is not mandatory.

Can I mix multiple Schema.org types on one page?

Yes — and it is encouraged. Most pages benefit from a BreadcrumbList plus a content-specific type (Article, Product, etc.) plus an Organization block. Use separate <script> tags or a JSON array for multiple types.

What is the difference between a Schema.org Type and a Property?

A Type categorises the thing (e.g. Recipe, Product, Person). A Property describes a detail of that thing (e.g. cookTime, price, name). Every JSON-LD block has one @type and any number of properties.

How often does Schema.org release new types?

Schema.org releases updates several times a year. Major version numbers (currently 27.x) group significant changes. Check schema.org/docs/releases.html for the changelog. New types appear first as "pending" at pending.schema.org.

Does Schema.org markup directly improve Google rankings?

Not directly — it is not a confirmed ranking factor. However, rich results from schema.org markup improve click-through rates significantly (often +20–50%), which improves traffic and indirectly signals relevance.

What happens if I use a Schema.org property that does not exist?

Google ignores unknown properties silently. It will not penalise you, but the data provides no benefit. Stick to documented types and properties from schema.org.

Can I use Schema.org on non-English websites?

Yes. Schema.org is language-neutral. Property values (like name and description) can be in any language. You can also use inLanguage to declare the content language. Google processes structured data in all supported languages.

Validate Your Schema.org Markup

Check any page or JSON-LD snippet against Schema.org and Google's rich result requirements.

Validate Schema →