Structured Data for SEO: What Schema Does Today on Google and in AI

Reading time: 11 minutes

Structured data for SEO is schema markup that describes a page to Google, but it now earns fewer rich results than before. Google dropped FAQ and HowTo from search. It also says AI needs no special schema. This guide shows what still works and how to test it.

What is structured data and what does it do for SEO?

Structured data markup is a standard format for describing what a page contains. Google uses it to understand the page and, at times, to show it as a rich result. In structured data for SEO, the schema vocabulary comes from schema.org. However, Google decides whether to show the result, and it gives no guarantee.

Schema.org is an open vocabulary founded by Google, Microsoft, Yahoo and Yandex. It supports RDFa, Microdata and JSON-LD. It also has more types than Google uses. So Google says its own documentation, not schema.org, is the reference for Search.

Google cites companies that gained clicks and visits from structured data. However, each case takes one line, with no method on the page. So the documentation suggests measuring the effect of structured data markup on your own site. First, pick stable pages with no markup and several months of history in Search Console. Then add the markup and confirm that Google found it. Finally, compare performance over a few months.

Which structured data does Google still use in 2026?

Google’s gallery of features lists 25 structured data features, such as Article, Breadcrumb, Organization, Local business, Product, Event, Recipe and Video. By contrast, the current list has no FAQ or HowTo. Each feature has its own rules, and Google decides when to show the result.

The gallery groups features by site type: ecommerce, organizations, news, jobs, entertainment, food and drink, education and science, and sports. So a blog or a business uses only a small part of it. Some features also serve few sites. Dataset, for example, serves Dataset Search, not Google Search.

The general guidelines warn that correct markup is not enough. Schema structured data lets a feature appear, but it does not guarantee that it will. That happens because the algorithm weighs search history, location and device. Sometimes it prefers a text result.

Which rich results did Google cut or remove?

Google cut FAQ and HowTo in 2023 and removed HowTo from search in September of that year. On May 7, 2026, it stopped showing FAQ. In 2025, it also announced the end of seven little-used types. So FAQ and HowTo structured data markup no longer earn a rich result.

When	What Google did	Feature
August 8, 2023	Limited FAQ to government and health sites and HowTo to desktop	FAQ and HowTo
September 13, 2023	Stopped showing HowTo on desktop too	HowTo
November 2024	Stopped showing the feature	Sitelinks search box
June 12, 2025	Announced the end of seven little-used types	See the list below
November 5, 2025	Announced the end of the feature	Practice problem
May 7, 2026	Stopped showing the result	FAQ

Why did Google cut these structured data features?

Google says it wanted to simplify the results. In 2023, it spoke of a cleaner and more consistent search experience. In 2025, according to Google, the seven types it announced were rarely used and no longer added significant value. They were Book Actions, Course Info, Claim Review, Estimated Salary, Learning Video, Special Announcement and Vehicle Listing. The change, by the way, does not affect rankings. In November 2025, Google kept Book actions, because a Search feature still uses the markup.

The update log shows that the FAQ documentation came down in June 2026. Since 2023, however, the result had applied only to well-known government and health sites. So for SEO, FAQ and HowTo structured data no longer earn a rich result on Google. For readers, though, an FAQ is still useful, as the guide on how to create an FAQ shows.

Which structured data fits a blog or a business?

For the SEO of a blog or a business, we recommend Article (or BlogPosting) on posts and BreadcrumbList on inner pages. On the home page, add Organization and WebSite. This choice is ours, made from the gallery, because Google does not define a basic set. If the business has an address, add LocalBusiness. If it sells online, add Product.

Structured data	What Google says	Where it goes
Article	Helps show title, images and date	Each article
BreadcrumbList	Shows the page’s place in the site	Inner pages
Organization	Helps tell the organization apart	Home page
WebSite	Shows the preferred site name	Home page
LocalBusiness	Shows business details in the panel	Location pages
Product	Shows price, stock and reviews	Product pages

In the Article documentation, Google says no property is required. However, it recommends author (with name and url), headline, image, datePublished and dateModified. Breadcrumb, in turn, shows on desktop, according to Google. The site name comes from WebSite markup on the home page, which Google weighs together with the title and og:site_name.

Google recommends putting Organization on the home page or on a page that describes the organization. It requires no property. For a business with a physical location, the local SEO guide completes the picture.

Example: structured data on a WordPress blog

At NotHuman, the blog runs on WordPress, and Yoast generates BlogPosting and BreadcrumbList schema for the posts. FAQPage appears only in posts that have an FAQ. So this markup brings no rich result in search, since Google stopped showing FAQ results.

JSON-LD, Microdata or RDFa: which format should you use?

For structured data for SEO, use JSON-LD, the format Google recommends. All three formats are equally fine for Google, as long as the markup is valid. JSON-LD, however, is the easiest to implement and maintain at scale. Also, Google reads it even when a script injects it into the page.

JSON-LD sits in a separate <script> block, apart from the text. Microdata and RDFa, by contrast, use attributes inside the visible HTML. In a CMS such as WordPress, Wix or Shopify, a plugin or an SEO screen can generate the structured data markup. Below is a minimal BlogPosting example, adapted from Google’s documentation. All values are made up:

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "BlogPosting",
  "headline": "Article title",
  "image": ["https://example.com/photos/16x9/photo.jpg"],
  "datePublished": "2025-01-05T08:00:00-03:00",
  "dateModified": "2025-02-05T09:20:00-03:00",
  "author": [{
    "@type": "Person",
    "name": "Author name",
    "url": "https://example.com/author/name"
  }]
}
</script>

If the code comes from JavaScript, Google asks for care. The JavaScript documentation mentions Google Tag Manager and custom JavaScript, and it recommends reading page data through variables. That is because copying the data into the tag raises the risk that the structured data markup drifts from the text. For AI crawlers, see the JavaScript SEO guide.

How do you test and validate structured data?

Test structured data for SEO with two official tools. Google’s Rich Results Test shows which results a page can generate. The Schema Markup Validator, from Schema.org, checks schema without Google’s rules. Then use the URL Inspection tool and the Search Console reports.

1. Open the Rich Results Test and paste the URL or the code. Fix critical errors first. Non-critical warnings help improve the markup, but they do not block eligibility.
2. Validate the syntax in the Schema Markup Validator when the type is not a Google feature. It comes from the old Structured Data Testing Tool, which Google moved to Schema.org and which no longer checks Search features.
3. Publish a few pages and use URL Inspection to see how Google reads the page. Also confirm that robots.txt, noindex or a login does not block access.
4. Request a recrawl and submit the sitemap, because crawling can take days.
5. Track the rich result reports in Search Console, since the structured data markup can break later because of page templates.

So structured data tests belong in the technical SEO audit, along with indexing and rendering.

Does structured data help you show up in AI?

According to Google, structured data is not required to appear in AI. To appear in AI Overviews and AI Mode, a page must be indexed and eligible for a snippet. There is no special schema.org markup and no AI file to create. Even so, Google recommends keeping structured data markup in your SEO, because it qualifies the page for rich results.

The AI features and your website page lists, among best practices, checking that structured data matches the visible text. The generative AI guide puts overfocusing on structured data on its list of things you can ignore. In the same guide, Google says that content people find unique, compelling and useful will likely influence AI presence more in the long run than any other suggestion.

That applies to Google, since the documentation does not cover other assistants. In its introduction, Google says it can make general use of sameAs. It also says some schema.org structured data may serve future Search features, if they prove useful. To understand the effect on answers, see the SEO and AI guide and the one on AI Overviews and AI Mode.

Which structured data errors does Search Console show?

Search Console shows, under Enhancements and Shopping, one report per rich result type, with valid and invalid items. An invalid item has at least one critical issue, so it cannot appear as a rich result. Non-critical issues only call for improvements.

The reports, though, have limits. They show a sample of items, and the numbers count items, not pages. One item can have several issues. Also, Search Console has no “Fix this issue” button, because the schema and other structured data live on the site. So the fix goes into the code, and only then does validation confirm the result.

Why does valid schema not show up in search?

Even valid schema structured data may not produce a result. Google lists five common reasons:

• Algorithm choice: it may prefer another feature or a text result.
• Different content: the markup does not represent the main content, or it misleads readers.
• Invisible error: the problem exists, but the Rich Results Test does not catch it.
• Hidden content: what the schema describes is not visible to the reader.
• Guidelines: the page does not meet the feature guidelines or the Search policies.

If the markup breaks the guidelines, Google can apply a manual action. In that case, the page loses eligibility for the rich result, but its ranking in web results does not change. The Manual Actions report in Search Console shows whether that happened.

Frequently asked questions

Main sources

• Google Search Central, Intro to How Structured Data Markup Works (2025).
• Google Search Central, Structured Data Markup that Google Search Supports (2026).
• Google Search Central, Latest Google Search Documentation Updates (2026).
• Google Search Central, Changes to HowTo and FAQ rich results (2023).
• Google Search Central, Simplifying the search results page (2025).
• Google Search Central, AI Features and Your Website (2025).
• Google Search Central, Google’s Guide to Optimizing for Generative AI Features on Google Search (2026).
• Google Search Console, Rich result report overview.