# What is schema markup, and what does it mean in SEO?

> What schema markup means when an audit asks for it or an app promises it: which vocabulary, which format, which types, and what it gets a store today.

URL: https://clione.ai/glossary/schema-markup/
Lang: en

The phrase usually reaches you as a to-do item: "missing schema markup" in an audit, "add schema markup to your products" on an app listing, "we will implement schema markup across the whole store" in an agency proposal. It is not a product, not a plugin, and not something different from [structured data](/glossary/structured-data/), which has its own entry. It is what the industry calls **structured data written with the [Schema.org](/glossary/schema-org/) vocabulary**. The useful question for a store is a different one: what exactly are they telling you when they say _schema markup_, and what do you need to ask to know whether what they offer is worth anything?

## What each word means

**_Schema_ is the vocabulary.** Schema.org is a community project, founded by Google, Microsoft, Yahoo and Yandex, that creates and maintains schemas for structured data on the web, and its vocabulary covers entities, the relationships between them and actions.[^schema-org-home] In practice these are the fixed names of things, such as [Product](/glossary/product-schema/) or [Organization](/glossary/organization-schema/), and of their properties, such as price or brand.

**_Markup_ is the act of marking up the HTML.** Schema.org's getting-started guide explains it with a [title tag](/glossary/meta-title/): `<h1>Avatar</h1>` tells the browser how to display the text, but not what it means, and "Avatar" could be the movie or a profile picture. Marking up the page means adding tags that tell the search engine what each piece of data is about, so it can understand it and display it usefully.[^schema-org-gs] The word dates from when this was done with attributes on the visible text, known as [microdata](/glossary/microdata/). Today the same vocabulary can also be written in [RDFa](/glossary/rdfa/) or in [JSON-LD](/glossary/json-ld/),[^schema-org-gs] which sits in a block separate from the text, and the name has stuck even though nothing visible is marked up anymore.

So the phrase tells you which vocabulary, but not which format. Google does not care: it accepts all three, recommends the one that is easiest for you to implement and maintain (in most cases, JSON-LD) and says all three are equal to it if the markup is valid and follows the documentation for each feature.[^google-sd-intro] The volume gives an idea of how established it is: according to Schema.org, in 2024 **more than 45 million domains marked up their pages with more than 450 billion objects** from its vocabulary.[^schema-org-home]

## What it gets you in Google, in 2026

Google uses structured data to understand a page's content and to display it with a richer appearance in the results, which it calls [rich results](/glossary/rich-results/).[^google-gallery] There are two limits that a schema markup proposal does not always mention.

The first: **Google does not guarantee that your structured data will appear in the results, even when the markup is correct** according to its own testing tool.[^google-sd-policies] Marking up is a condition for qualifying, not a result.

The second: the catalog of rich results changes, and in the last two years it has shrunk. In September 2025 Google retired the documentation for several types it no longer showed, such as course info, estimated salary and vehicle listings. The FAQ rich result stopped appearing in Search on May 7, 2026, and in June Google also retired its documentation.[^google-updates] [FAQPage](/glossary/faqpage/) markup is still valid vocabulary, but **it no longer produces that rich result in Google**. For a store, what is still alive and growing is product markup: in July 2026 Google expanded the documentation for [merchant listings](/glossary/merchant-listings/) to accept the product category.[^google-updates]

And for the generative AI features in Search, Google is explicit: structured data is not a requirement and **there is no special Schema.org markup you need to add**; it still advises using it as part of your SEO strategy, because it helps you qualify for rich results.[^google-ai-guide]

## In a store

A proposal line like "we will implement schema markup across the whole store" says almost nothing until five questions are answered. **Fictional example**: the line is made up, and the answers in the third column are the ones you can check against Google's documentation.

| Question                       | Why it matters                                                                            | An answer you can check                                                                                                                                                                                                                                                   |
| ------------------------------ | ----------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Which types?                   | The vocabulary has hundreds; Google uses only some for its rich results.[^google-gallery] | Product with its offer on [product pages](/glossary/product-page/), BreadcrumbList on pages with [breadcrumbs](/glossary/breadcrumbs/), Organization on the home page. |
| On which pages?                | Not everything goes everywhere.                                                           | Google recommends putting organization data on the home page or on a single page that describes the organization, not on every page.[^google-organization]                                                                                                                |
| Where does the data come from? | The markup has to say what the visible page says.[^google-sd-policies]                    | From the same field that renders the price or name on the [product page](/glossary/product-page/), not from a separate copy.                                                                                                                                              |
| How many properties?           | More is not better.                                                                       | Google prefers a few recommended properties that are complete and accurate over many that are incomplete or malformed.[^google-sd-intro]                                                                                                                                  |
| How is it checked?             | What is not measured is not delivered.                                                    | The [Rich Results Test](/glossary/rich-results-test/) on a published URL, before and after.[^rich-results-test]                                                                                                                                                           |

The fourth row is the one that saves the most money. Google says you must include all the required properties for a type to qualify for the rich result, and that adding recommended ones can help, but that **it is more important to provide fewer properties that are complete and accurate than to try to provide them all** with incomplete or inaccurate data.[^google-sd-intro] A quote priced by "number of properties marked up" is measuring the wrong thing.


Believing there is a schema markup "for AI" that is different from the usual one, and paying for it separately. For the generative AI features in its search engine, Google says there is no special Schema.org markup to add.[^google-ai-guide] What does count is the usual markup, complete and faithful to the page.


## How to check

1. **See what markup you already have.** In the source of a product page, look for `application/ld+json` (JSON-LD), `itemscope` (microdata) and `typeof` (RDFa). Before you buy schema markup, know how much your theme already ships.
2. **Run the URL through the Rich Results Test.** It accepts all three formats and tells you which rich result types it finds, with their errors and suggestions.[^rich-results-test]
3. **Compare each type against Google's gallery**, the list of markup Search supports today.[^google-gallery] If a type is not there, do not expect a rich result from it in Google.
4. **Check the changelog of Google's documentation** before accepting a quote: that is where features being retired and expanded are announced.[^google-updates]
5. **Compare what the markup says with what a customer sees**, field by field. What is not visible on the page is not marked up.[^google-sd-policies]

## Related concept: structured data

Structured data is the general idea: a standardized format for providing information about a page and classifying its content.[^google-sd-intro] Schema markup is that idea built with one specific vocabulary, Schema.org, and since most of the structured data Google Search uses relies on that vocabulary,[^google-sd-intro] in a store the two names usually point to the same block of code. The difference matters when someone uses "schema" to mean only JSON-LD, or only the app that generates it: the vocabulary, the format and the tool are three different things, and a serious proposal names all three.

To go further, the [structured data reading path](/learn/structured-data/) puts the rest of the terms in order, and the guide to [auditing a product page's structured data](/guides/audit-product-page-structured-data/) applies these questions to a whole product page.