SEO

What Is Schema Markup? What It Does for Google and AI Search, and What It Doesn't

Schema markup is code, usually JSON-LD, that labels what a page is about in a vocabulary search engines share. It makes a page eligible for rich results such as review stars and helps search engines identify your business. It doesn't guarantee a rich result or a higher ranking, and the best study so far found no lift in AI citations.

Schema markup has had a confusing year. Google retired its FAQ rich result in May. In the same month, Ahrefs published a controlled test of whether schema raises AI citations. Meanwhile, many guides still present schema as the way into AI answers.

This guide separates what is documented from what is hoped. Every claim about a search engine below comes from that company's own documentation, checked on November 12, 2026, or from a named and dated study.

Schema markup tells machines what your page already says. It can't say anything your page doesn't.

What is schema markup?

Schema markup is code added to a web page that labels its content with terms from Schema.org, so software can tell that "4.8" is a rating, "Mon to Fri" is opening hours and "Northside Plumbing" is a business. Visitors never see it. Search engines read it.

Three facts are worth knowing:

  • The vocabulary is shared. Schema.org was founded by Google, Microsoft, Yahoo and Yandex. Its current release, version 30.1 from September 16, 2026, defines 826 types and 1,540 properties.
  • "Structured data" is the same thing in practice. Google defines structured data as "a standardized format for providing information about a page and classifying the page content." Schema markup is structured data written in the Schema.org vocabulary.
  • JSON-LD is the usual format. Google accepts JSON-LD, Microdata and RDFa, and says all three are fine when the markup is valid. It recommends JSON-LD as the easiest to implement and maintain in most cases.
A worked example

LocalBusiness markup for a plumber, line by line

{
  "@context": "https://schema.org",
  "@type": "LocalBusiness",
  "name": "Northside Plumbing",
  "address": {
    "@type": "PostalAddress",
    "streetAddress": "120 Example St",
    "addressLocality": "Columbus",
    "addressRegion": "OH",
    "postalCode": "43215",
    "addressCountry": "US"
  },
  "telephone": "+1-614-555-0100",
  "url": "https://northsideplumbing.com/",
  "openingHoursSpecification": {
    "@type": "OpeningHoursSpecification",
    "dayOfWeek": ["Monday", "Tuesday",
      "Wednesday", "Thursday", "Friday"],
    "opens": "08:00",
    "closes": "18:00"
  },
  "sameAs": [
    "https://www.facebook.com/northsideplumbing"
  ]
}
  1. The vocabulary. This line says the terms that follow come from Schema.org.
  2. The type. What the thing is. Everything below describes this one business.
  3. Required by Google. For local business results, Google requires only name and address.
  4. Recommended. Phone, URL and opening hours are optional, and they must match what the page shows.
  5. Other profiles. sameAs links the business to its profiles elsewhere, which helps tell it apart from others with a similar name.
The business, address and phone number are invented for illustration. Property requirements are from Google's local business documentation.

What does schema markup do in Google Search?

In Google Search, schema markup does two things: it helps Google understand a page, and it makes the page eligible for rich results. A rich result is a search listing with something extra, such as review stars, a price, an event date or a breadcrumb trail.

Google's search gallery lists 25 structured data features today. The ones a typical business can use include Article, Breadcrumb, Event, Local business, Organization, Product, Review snippet and Video.

Rich results can earn more clicks. Google's own introduction cites publishers that measured this: Rotten Tomatoes saw a 25% higher click-through rate on pages with structured data, and Nestlé measured an 82% higher click-through rate on pages that appeared as rich results. Those are large sites that Google chose as examples, so treat them as what is possible, not what is typical.

Eligible is not guaranteed

Google's structured data guidelines say plainly that it "does not guarantee that your structured data will show up in search results." Valid markup makes the page eligible. Google decides whether to show the feature.

Is schema markup a ranking factor?

Google doesn't describe schema markup as something that raises rankings. Its documentation presents structured data as a way to understand content and to qualify for rich results, and nothing more.

Two statements from Google point the same way:

  • When it announced in June 2025 that it was phasing out seven markup types, Google wrote: "This update won't affect how pages are ranked."
  • Its guidelines say a structured data penalty means a page loses its rich result, and that it "doesn't affect how the page ranks in Google web search."

So the honest pitch for schema markup is a better-looking, better-understood listing in the position you have already earned. Anyone who promises a ranking jump from markup alone is promising something Google doesn't.


Which rich results has Google removed?

Google has been retiring rich result types since 2023, including two widely used ones: How-to and FAQ. The table lists the main changes. If your markup plan is older than this timeline, it needs a review.

DateWhat changed
August 8, 2023Google limited FAQ rich results to well-known government and health websites
September 13, 2023How-to rich results stopped showing on desktop too, which ended the feature
June 12, 2025Google began phasing out seven types: Book Actions, Course Info, Claim Review, Estimated Salary, Learning Video, Special Announcement and Vehicle Listing
November 5, 2025Practice problem markup was deprecated. Google also clarified that Dataset markup is used only by Dataset Search, and kept Book Actions after all
May 7, 2026The FAQ rich result stopped appearing in Google Search for every site
June 15, 2026Google removed the FAQ rich result documentation

The dates from 2025 and 2026 come from Google's documentation changelog. Google's stated reason, in its November 2025 update, was that some features "aren't being used very often."

What about the FAQ markup already on your site? In its 2023 post, Google said of markup that no longer produces a feature that "there's no need to proactively remove it." FAQPage is still a valid Schema.org type. We still use it on this page, because the questions below are visible to readers and the markup describes them accurately. We don't expect a Google feature from it.

Does schema markup help with AI search?

Not in any way that has been proven. Google says its AI features need no special schema, Microsoft says structured data may help without guaranteeing anything, and the one controlled study we found showed no meaningful change in AI citations after pages added it.

One labeled web page connected by threads of light to three pedestals: a search result card with stars, a knowledge node with linked facts, and a glass speech bubble
One page, three readers: the search result, the search engine's record of your business, and the AI answer.
SourceWhat it says about schema markup
Google AI Overviews and AI ModeGoogle's AI features guide: "There's also no special schema.org structured data that you need to add." It does list keeping structured data consistent with the visible text as a best practice
Bing and CopilotBing's Webmaster Guidelines: "Structured data may support clearer grounding but does not guarantee visibility or grounding traffic"
Microsoft, October 8, 2025Krishna Madhavan, a principal product manager at Bing, wrote that schema "helps search engines and AI systems understand your content," and that nothing guarantees selection
Ahrefs study, May 11, 2026After 1,885 pages added JSON-LD, citations changed by -4.6% in AI Overviews, +2.4% in AI Mode and +2.2% in ChatGPT, compared with matched pages that didn't

The Ahrefs study, written by Louise Linehan, deserves a closer look because it compares pages that added markup with a control group that didn't. It followed pages that added JSON-LD between August 2025 and March 2026 and compared the 30 days before with the 30 days after. Ahrefs called the AI Mode and ChatGPT changes statistically indistinguishable from zero. The small AI Overviews decline was statistically significant, but those pages were already losing citations before the markup went in.

The study has limits, and Ahrefs lists them. Every page already had at least 100 AI Overview citations before the test, so it says nothing about pages AI systems haven't found yet. All schema types were pooled together. And an earlier Ahrefs analysis did find that AI-cited pages were almost three times as likely to have JSON-LD, which is a correlation: well-run sites tend to do both.

Our take

Add schema markup because it is accurate housekeeping that Google and Bing both ask for, not because it will get you into ChatGPT. What gets a page cited is still content that answers the question clearly. Our guides on structuring content for AI search and showing up in Google AI Mode cover that work.

Which schema types does a small business need?

Most small businesses need four types: Organization, LocalBusiness, Article and BreadcrumbList. Add Product or Review snippet markup only if your pages actually show products or reviews.

TypeWhere it goesWhat Google's documentation says
OrganizationHome pageHelps Google understand your organization's details and tell it apart from others. It can influence the logo and knowledge panel. No required properties
LocalBusinessHome or location pageRequires name and address. Hours, phone, URL and coordinates are recommended. Search may show a knowledge panel with the business's details
Article or BlogPostingPosts and guidesNo required properties. Recommended: headline, image, author, datePublished and dateModified
BreadcrumbListEvery page below the home pageShows the page's place in the site in the search result. Needs at least two items. Google lists the feature as available on desktop
Product, Review snippet, Event, VideoOnly pages with that contentEach has its own required properties in the search gallery. Review markup has rules against fake and undisclosed paid reviews

Organization and LocalBusiness markup do one more job. They state, in a fixed format, who you are, where you are and which other profiles belong to you. That is the raw material search engines use to treat a business as a known entity, which we explain in our guide to entity-based SEO. Local businesses will find the rest of the setup in our local SEO checklist. Online stores using Product markup for sales should read our holiday SEO checklist for U.S. online stores, which covers sale prices marked as StrikethroughPrice.

How do you add schema markup to a website?

You add schema markup by placing a JSON-LD script in the page's HTML that describes what is visible on that page. The work is mostly deciding what to describe. The code is short.

  1. Check what you already have. Run your home page and one post through the Rich Results Test in the next section. Your website platform or SEO plugin may already output markup.
  2. Pick the type that matches the page. One main thing per page: the business on the home page, the article on a post, the product on a product page.
  3. Fill it from the page. Use the same name, address, hours, prices and dates a visitor can read. Google's guidelines say not to mark up content that isn't visible to readers.
  4. Add it as JSON-LD. Put the script in the page's HTML. Google says it can also read JSON-LD that JavaScript adds to the page, but plain HTML is the simplest thing to check.
  5. Test, then publish. Fix every error. Warnings usually point to recommended properties you can add later.
  6. Keep it in sync. When your hours, prices or address change on the page, change them in the markup too.

How do you test schema markup?

Test schema markup with two free tools: Google's Rich Results Test shows which Google rich results a page's markup can generate, and the Schema Markup Validator is Schema.org's own checker. This check takes about five minutes.

  1. Open the Rich Results Test and enter your page's URL. Google describes it as a way to see which rich results the page's structured data can generate.
  2. Note which items it detects, and any errors or warnings it reports.
  3. Paste the same URL into the Schema Markup Validator, which Schema.org hosts and which isn't limited to Google's features.
  4. Compare every value with the page itself. A phone number, price or rating in the markup that a visitor can't find on the page is the problem to fix first.
  5. After publishing, watch the rich result reports in Google Search Console, which Google recommends for monitoring markup across the whole site.

Markup is one line on a longer list. Our technical SEO checklist covers the crawling, indexing and speed checks that come before it.

What mistakes get schema markup ignored or penalized?

Markup gets ignored or penalized when it describes something the page doesn't show. Both Google and Bing say so directly.

  • Marking up invisible content. Google: "Don't mark up content that is not visible to readers of the page." Bing: "Structured data must accurately represent visible content."
  • Fake or paid-for reviews in review markup. In July 2026 Google added a review snippet guideline on fake and undisclosed incentivized reviews.
  • Blocking the page. Google can't use markup on a page that robots.txt or a noindex tag keeps it away from.
  • Treating a valid test as a promise. A clean test result means the page is eligible, not that the feature will show.

The cost of getting it wrong is specific. Google can issue a manual action that removes a page's rich results, and Bing lists ignored structured data among the consequences for sites that break its guidelines.

We checked the details in this guide against Google's, Microsoft's and Schema.org's documentation on November 12, 2026. Google changes its list of supported rich results often, so check the search gallery before you plan work around a specific feature.

Not sure what your markup says today?

Send us your site and we'll tell you which schema markup it has, what is missing and what isn't worth adding. Markup review is part of the technical work in our SEO services.

Ask for a Markup Check

Frequently Asked Questions

Is schema markup a ranking factor?

Google doesn't describe it as one. Its documentation says structured data helps Google understand a page and makes the page eligible for rich results. When Google began phasing out seven markup types in June 2025, it said the change wouldn't affect how pages are ranked. The benefit is a clearer, more clickable result, not a higher position.

Does schema markup help you get cited by ChatGPT or AI Overviews?

There is no proof that it does. Google says AI Overviews and AI Mode need no special schema. In May 2026, Ahrefs tracked 1,885 pages that added JSON-LD and found no meaningful rise in citations from AI Overviews, AI Mode or ChatGPT. Microsoft says structured data may support clearer grounding in Copilot, without a guarantee.

Should I remove FAQ schema now that Google has dropped FAQ rich results?

No, there's no need. Google stopped showing FAQ rich results on May 7, 2026, so the markup no longer earns a Google feature. FAQPage is still a valid Schema.org type that other systems can read. Keep it where the questions and answers are visible on the page, and don't expect extra space in Google from it.

What is the difference between schema markup and structured data?

Structured data is the general idea: information about a page in a standard format that software can read. Schema markup is structured data written with the Schema.org vocabulary, which Google, Microsoft, Yahoo and Yandex founded. In everyday SEO the two terms mean the same thing, and JSON-LD is the format most sites use to write it.

Do I need a developer to add schema markup?

Not always. First run a page through Google's Rich Results Test, because your website platform or SEO plugin may already add markup. If something is missing, a JSON-LD block is a short script that goes in the page's HTML. A developer helps when the values must come from a product catalog or a database.

Schema Markup Structured Data JSON-LD Rich Results Technical SEO AI Search