Schema markup can help search engines understand what a page is about, and it can make the page eligible for some supported Google rich results. It is not a ranking boost, and Google does not guarantee that an eligible enhancement will appear. For most sites it is worth adding to the pages that matter, as long as the markup is accurate.

If you are deciding what to do this week, work in this order. Mark up facts that visitors can see on the page, choose the most specific schema.org type, check Google's current requirements for the feature you want, validate, publish, inspect the URL, and then watch impressions, clicks, click-through rate and average position in Search Console.

Google reduced or retired several rich results between 2023 and 2026, FAQ and HowTo among them, so the list of types worth your time is shorter than older advice suggests. AI search features add a separate question, and your own Search Console data is the measure to use for any of it.

Key takeaways

  1. Eligibility, not a ranking boostMarkup can make a page eligible for some Google rich results, but Google says structured data by itself is not a generic ranking factor.
  2. Mark up visible factsDescribe only what visitors can see, using the most specific schema.org type that fits. Hidden or misleading markup can cost a page its rich result eligibility.
  3. Check current Google featuresGoogle stopped showing FAQ rich results in May 2026, and HowTo results in 2023. FAQPage stays valid schema.org, but it no longer leads to a Google display.
  4. Keep markup for every consumerAI Overviews and AI Mode need no special markup, Google says. Treat that as a dated statement about Google, not a reason to skip accurate markup.
  5. Follow a safe orderChoose the page, mark up visible facts, check Google's current requirements for the feature you want, validate, publish and inspect the URL in Search Console.
  6. Measure on your own dataWatch impressions, clicks, click-through rate and average position in Search Console after Google recrawls the page. No percentage is a number to expect.

What schema markup can help with

Structured data is a set of labels, usually written as JSON-LD, that describe a page in a form a machine can read. Schema.org supplies the vocabulary: types such as Article, Product and Event, and properties such as headline, offers and startDate. Google's introduction to structured data says it uses markup to understand page content and to enable certain search features. For a plain definition, the guide to what schema markup is goes slower.

The first benefit is clarity. A page that sells a ticketed bread-baking class can state, in labeled fields, that it describes an event with a start date, a place and a price. A reader can see that at a glance, and a machine otherwise has to infer it from the surrounding text.

The second benefit is eligibility. Some Google features, such as breadcrumb trails, product details, event listings, recipe details, local business details and video, depend on markup. Google's general structured data guidelines say correct markup can make a page eligible for a rich result, and that eligibility does not guarantee it will appear.

The third is a possible change in engagement. When Google does show an enhancement, the listing can look different from a plain text result, and that may change how often people click. Treat that as something to measure on your own pages, not something to expect by default.

What markup cannot do

Valid markup does not move a page up the results by itself. It does not replace useful content, clear titles or the other things that decide whether a page deserves to rank. It also does not turn hidden, inaccurate or self-serving claims into acceptable ones.

Google's guidelines say markup should describe content that is visible on the page. They also say a structured data manual action can remove a page's rich result eligibility without directly lowering its web search ranking. That is a real cost, though: a page that loses eligibility loses the displays it was relying on.

Appearance also depends on things you cannot set. Google says the query, the device, the location and other ranking and quality systems all play a part in whether an eligible page gets an enhancement. A page can pass every test and still look like an ordinary result.

What Google says about rankings

Google's position is stated plainly. In an August 11, 2020 post on the Google Search Central blog, John Mueller wrote: "Structured data by itself is not a generic ranking factor." The post goes on to describe structured data as a way to help Google understand a page and make it eligible for additional search experiences.

Google repeated the point in its June 12, 2025 announcement about simplifying the results page, which says that retiring some structured data displays does not affect how pages rank.

Google describes markup as a route to understanding and eligible displays, not as a ranking factor, and it does not say markup is pointless. The benefit lives in how a page is understood and displayed, so judge it by those measures.

Rich results that remain, and features that changed

Google's structured data gallery lists the experiences Google Search currently supports. It includes Article, Breadcrumb, Event, Image metadata, Job posting, Local business, Organization, Product, Profile page, Recipe, Review snippet, Software app, Video and several others. Eligibility varies by feature, country, device and content type.

Three search result cards in a row, the first plain, the second outlined with a thumbnail, the third with extra detail rows
Figure 2. An eligible page may be shown with more detail, but whether Google shows it is Google's decision.

Several features that older guides still recommend have changed. These dates come from Google's own announcements and its documentation updates page.

  • FAQ: Google limited FAQ rich results in August 2023, stopped showing them in Google Search in May 2026, and removed the feature documentation in June 2026.
  • HowTo: Google stopped showing how-to rich results on mobile in 2023 and extended that to desktop on September 13, 2023.
  • Sitelinks search box: Google announced on October 21, 2024 that it would remove the feature from November 21, 2024, and said ordinary sitelinks and rankings were not affected (announcement).
  • Course info, estimated salary, learning video, special announcement and vehicle listing: documentation removed in September 2025 because the features were no longer shown.
  • Practice problem: documentation removed in January 2026 for the same reason.
  • Dataset: Google clarified that Dataset markup is used by Dataset Search, not by ordinary Google Search results.

FAQPage is still a valid schema.org type, and it can still describe real questions and answers on a page accurately. It is no longer a route to a Google display, so do not add it expecting one. Choose markup by looking at the page's actual content and at Google's current feature documentation, not at an old list of types. The guide to types of schema markup helps with that choice.

Google says structured data is not required to appear in AI Overviews or AI Mode, and that there is no special schema.org markup to add for them. That appears in its AI features documentation and its guide to optimizing for generative AI features.

The same pages say a page must be indexed and eligible to appear in Google Search with a snippet before it can be a supporting link, that structured data should match the visible text, and that Google does not guarantee crawling, indexing, inclusion, citations or traffic. Google also says traffic from AI features is reported inside the overall Web search type in Search Console.

Treat that as a dated statement about Google's AI features, not as a reason to skip markup. Accurate markup still gives machines clear context about the organization, product, event or article on a page, and it still serves the Google features listed above. No schema type guarantees inclusion or a citation, so treat any promise of a fixed lift from one type with caution.

Google is not the only consumer. Bing's documentation on structured data says markup helps Bing understand the type of content on a page, and it does not promise a ranking increase. The practical advice is the same for every consumer: keep markup accurate and matched to what visitors see. For more on this question, see whether AI uses schema markup and schema markup for AI search.

Add markup in a safe order

The order matters more than the code. Each step protects the next one, and the last two are the ones people skip.

  1. Choose the page and decide what its main entity is: an article, a product, an event, a business.
  2. Mark up only information a visitor can see on that page.
  3. Pick the most specific schema.org type that fits.
  4. Check Google's current documentation for the feature you want and note its required properties.
  5. Write the JSON-LD and validate it.
  6. Publish, then use URL Inspection in Search Console to confirm Google can reach the page.
  7. Measure after Google has recrawled it.

Google recommends JSON-LD in most cases, though valid JSON-LD, Microdata and RDFa are all supported formats. If you are choosing between them, JSON-LD, Microdata and RDFa compared lays out the trade-offs, and how to add schema markup covers installation by platform.

A safe setup sequence

Valid syntax does not guarantee a visible search enhancement, so the sequence ends with measurement.

  1. Choose the pageDecide the page and its main entity before writing any code.
  2. Match visible factsMark up only information visitors can see on the page.
  3. Pick the specific typeUse the most specific schema.org type that fits the content.
  4. Add JSON-LDWrite complete markup using Google's current requirements for the feature.
  5. ValidateUse the Rich Results Test and the Schema Markup Validator.
  6. Publish, inspect, measureUse URL Inspection, then watch Search Console and conversions after recrawling.

Read a complete example before you write your own

The example below describes a fictional bakery article with its organization and breadcrumb trail. Replace the name, URLs, dates, image and author with the real, visible facts for your page. Keep @context in place, give each item an accurate @type, and make headline match the heading visitors read.

json
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Organization",
      "@id": "https://example.com/#organization",
      "name": "Example Bakery",
      "url": "https://example.com/",
      "logo": {
        "@type": "ImageObject",
        "url": "https://example.com/images/example-bakery-logo.png",
        "width": 512,
        "height": 512
      }
    },
    {
      "@type": "Article",
      "@id": "https://example.com/blog/sourdough-starter-basics/#article",
      "mainEntityOfPage": "https://example.com/blog/sourdough-starter-basics/",
      "headline": "Sourdough Starter Basics for Home Bakers",
      "image": ["https://example.com/images/sourdough-starter.jpg"],
      "datePublished": "2026-05-11",
      "dateModified": "2026-05-11",
      "author": {
        "@type": "Person",
        "name": "Maria Example",
        "url": "https://example.com/about/maria-example/"
      },
      "publisher": { "@id": "https://example.com/#organization" }
    },
    {
      "@type": "BreadcrumbList",
      "@id": "https://example.com/blog/sourdough-starter-basics/#breadcrumb",
      "itemListElement": [
        {
          "@type": "ListItem",
          "position": 1,
          "name": "Blog",
          "item": "https://example.com/blog/"
        },
        {
          "@type": "ListItem",
          "position": 2,
          "name": "Sourdough Starter Basics for Home Bakers"
        }
      ]
    }
  ]
}

On a page, this JSON sits inside a script element with type="application/ld+json". The publisher property points back to the organization by @id instead of repeating it, which keeps the graph consistent if you later change the logo. Set dateModified to a new date only when the article changes in a way readers would notice. If you would rather fill in fields than write JSON, the Article schema generator, the breadcrumb schema generator and the organization schema generator produce the same kinds of blocks.

Check what your platform already outputs

Before you add anything, look at the rendered page. Many platforms and SEO plugins already emit some markup, and a second block that describes the same item can conflict with the first.

An older woman with a silver bob and reading glasses studies a laptop at a cluttered desk in a garden center back office
Figure 1. Look at the rendered page and at what the platform already outputs before adding any markup of your own.

On WordPress, an SEO or schema plugin may already output Article, Organization, WebSite or Breadcrumb data, so check for duplicates before adding a snippet. WordPress's learning material shows schema selection through common SEO plugins, and the guide to schema markup on WordPress goes through the install routes.

Shopify says its themes include structured data in their markup, so inspect the product page before adding custom code (Shopify SEO overview). Wix generates preset markup for several page types and lets you customize JSON-LD in its structured data settings. Webflow lets you add schema in page settings or custom code, and warns against putting the same markup in more than one place. Squarespace and other hosted builders may also generate markup for common page types, so look at the live page there too.

On a Next.js site, the Next.js documentation recommends rendering JSON-LD in a script tag in layout.js or page.js, sanitizing serialized strings and validating the output.

Property status belongs to a specific Google feature, not to schema.org as a whole. A property can be valid vocabulary without being required, or even used, by a Google rich result. That is why a page can be valid in one sense and still missing something a feature asks for.

In Google's documentation, a required property is necessary for eligibility, and a recommended property can improve the quality of a result. Article markup lists no required properties, and recommends author, headline, image, datePublished, dateModified and publisher (Article documentation). Breadcrumb markup requires itemListElement with at least two ListItem entries (Breadcrumb documentation).

Organization markup lists no required properties, and Google recommends relevant ones such as name, url, logo, sameAs, address and contact details (Organization documentation). Product snippets require name and one of review, aggregateRating or offers (Product snippet documentation). Local business markup requires name and address in the current documentation (Local business documentation).

Those pages were updated in 2026, and Google changes this kind of documentation, so open the one for your feature before you publish.

Validate, inspect and measure

Each of these needs its own check: the JSON can be valid, the vocabulary can match schema.org, the markup can meet the requirements of a specific Google feature, and the page can actually show an enhancement. Passing one does not mean passing the next.

For syntax and vocabulary, use the Schema Markup Validator or the Schema Nova validator. For Google feature eligibility, use the Rich Results Test, which Google describes in its help page for the tool. Neither one promises that Google will show anything.

After publishing, use URL Inspection to confirm Google can reach the page. Then measure once it has been recrawled: impressions for the eligible appearance, clicks, click-through rate, average position, and conversions or revenue in your analytics. Compare similar pages before and after the change where you can, and write down any other content, template or linking changes made around the same time.

Results differ by page type, feature and site, so no percentage is a number to expect. Your own Search Console data over several weeks is the more useful guide to whether the markup earned its place.

Questions people ask

Should every page have structured data?

Not necessarily. Start with pages where a type matches a documented Google feature or describes a clear entity: the home page, articles, products, events and local business pages. Add markup elsewhere only when it describes something visitors can see.

Can structured data describe something that is not visible on the page?

No. Google's guidelines say markup should represent visible, accurate content. Hidden or misleading markup can lead to a manual action that removes rich result eligibility, so match the markup to what the page shows.

Is FAQPage markup still worth keeping?

Google stopped showing FAQ rich results in May 2026, and FAQPage remains a valid schema.org type. If your page shows the questions and answers and the markup matches them, leave it as it is. Do not add it expecting a Google display.

Why does my markup pass but no rich result appears?

Passing a test means the code is readable and meets a feature's requirements. Google decides whether to show an enhancement, and the query, device and location all play a part. Check that the page is indexed, then compare your Search Console data over time.

How will I know whether markup helped?

Compare impressions, clicks, click-through rate and average position for similar pages before and after the change, and add conversions from your analytics. Note other changes made at the same time so you do not credit markup for them.

Quick check: ranking, eligibility and appearance

Choose an answer to see feedback. Nothing is scored or saved.

Question 1 of 3

Does valid schema automatically move a page higher in Google results?

Question 2 of 3

Does passing the Rich Results Test guarantee a rich result?

Question 3 of 3

What does Google say about markup for AI Overviews and AI Mode?

Schema and SEO terms

Select a term to see its definition.

Structured data. Machine-readable information that describes a page or entity.