Question

The Role of Structured Data in Getting Cited by AI Overviews

Google AI Overviews 5 min read

Google has answered this question in writing. As of October 2026, its documentation on AI features says: “There’s also no special schema.org structured data that you need to add.” Its AI optimization guide repeats the point in a section that addresses common myths: structured data is not required for generative AI search, although it is still useful for rich results. Structured data, in other words, is not a documented route to citation in Google’s AI Overviews. It keeps a real but narrower role: helping Google understand a page and an organization, and qualifying pages for the rich results that still exist.

If you have been told that schema markup, and FAQ markup in particular, is the first thing to deploy for AI search, check when that advice was written. Several of the features it relied on have since been retired.

What Google has said, and when

Several changes since mid-2025 reshaped what structured data does in Google Search.

  • June 2025. Google announced it was simplifying search results by phasing out support for seven structured data types: Book Actions, Course Info, Claim Review, Estimated Salary, Learning Video, Special Announcement and Vehicle Listing. Google stated that “this update won’t affect how pages are ranked.”
  • May 2026. According to Google’s Search Central documentation changelog, FAQ rich results stopped appearing in Google Search starting May 7, 2026, and the FAQ structured data documentation was removed on June 15, 2026.
  • May 2026. The same changelog records the addition of the AI optimization guide on May 15, 2026. The guide, last updated in July 2026, is where Google’s statements about structured data and AI features now live.

The pattern is consistent. Google has been reducing the number of markup types that produce visible features, and it has said plainly that no markup type is a condition for appearing in AI Overviews or AI Mode.

Why FAQ schema in particular is a weak bet

FAQ markup was recommended for AI search for an understandable reason. AI Overviews answer questions, so a page marked up as questions and answers seemed like a natural fit. Three facts now cut against it.

First, the rich result it produced no longer exists in Google Search. Second, Google’s AI documentation does not ask for it. Third, the Q&A format itself carries a quality risk when it is used mechanically. Google’s Search Quality Rater Guidelines describe collecting questions from sources like “People also ask” and answering them “with unoriginal, paraphrased content” as a common technique for producing scaled content.

None of this makes a well-written question-and-answer section harmful to readers. It means the markup around it has lost the Google benefit it was added for, and the section itself has to earn its place on the page.

Where structured data still helps

Structured data has not become pointless. Its legitimate jobs are understanding and eligibility.

Understanding an organization. Google’s Organization structured data documentation, updated in September 2026, says that adding it to a home page “can help Google better understand your organization’s administrative details and disambiguate your organization in search results.” Some properties work behind the scenes to tell your organization apart from other organizations. Others can influence visible elements, such as the logo shown in results. For a brand whose name is shared with other companies, that disambiguation is directly relevant.

Rich result eligibility. Google’s AI guide notes that structured data remains useful for rich results. Markup types that still produce them can change how a page appears in classic results, and that appearance can earn clicks whether or not an AI Overview is shown. Check Google’s current documentation before relying on any single type, since the supported list has shrunk more than once.

Consistency. Markup that describes what is visibly on the page gives Google a second, structured statement of the same facts. That consistency is a reasonable aim on its own terms. It is not documented as a factor in AI Overview selection, and it should not be sold as one.

How other engines describe structured data

Google is not the only AI surface, and here the platforms genuinely differ.

Microsoft has been more open to markup. Microsoft Advertising’s guidance on appearing in AI search answers, published in October 2025, explains that assistants like Copilot break content into smaller pieces, a process it calls parsing. It recommends schema markup alongside clear headings, lists, tables and question-and-answer formatting. It also advises against hiding important content in tabs or accordions and against leaving key information only in PDFs.

The practical consequence is that one markup decision can have different value on different surfaces. For Google, schema is an understanding and rich result tool. For Microsoft’s AI experiences, the company itself describes it as helpful for parsing. If Copilot visibility matters to your audience, that difference is a reason to keep good markup, while being clear about which engine the reason applies to.

Some teams prefer to hand AI search work to an outside provider, through programs such as Southern Digital Consulting’s AI SEO services. Whoever does the work, the same test applies: each markup type should be justified by a documented benefit on a named surface, not by a general promise of AI visibility.

Apply a simple rule to every markup type on the site: keep it if you can name the surface it serves and the benefit that surface documents.

  1. Organization markup on the home page. Keep it and complete it. Google documents its role in understanding and disambiguating your organization.
  2. Markup that still produces rich results. Keep it where the rich result is relevant to the page, and validate it after template changes.
  3. FAQ and other retired types. Stop treating them as a Google tactic. If the markup is already in place and accurate, removing it is not urgent. Do not add new FAQ markup expecting a Google AI benefit.
  4. Markup with no visible counterpart. Remove or fix markup that describes content readers cannot see. It has no documented benefit on any surface discussed here.
  5. Non-Google surfaces. If Microsoft Copilot matters to your audience, keep clean, accurate markup and well-structured HTML, since Microsoft’s own guidance recommends both.

Structured data plays a supporting role in AI search. It helps Google understand who you are and qualifies pages for the rich results that remain. Google has said directly that no special markup is needed for AI features, so markup should not be expected to make a page more likely to be cited in an AI Overview. The work that earns citations is the work that earns rankings: content worth retrieving, on pages Google can crawl, index and show with a snippet.