AI Overviews shift user attention from blue links to the generated answer. Visibility now means being cited inside that answer, not just ranking first underneath it.
For SEO teams, the challenge moves from “how do we rank” to “how do we become a trusted source the model quotes.” Structured data is part of that story, but not a magic switch. Google says there is no special schema for AI Overviews, and no extra requirements beyond normal Search best practices. Yet observational studies find that pages with schema are cited more often, which raises a harder question: what is structured data really doing for AI search visibility?
- 8% vs 15%Visits that clicked a normal result, with and without an AI summary
- −34.5%Average CTR of the top result when an AI Overview appears
- ~3×How much more likely AI-cited pages were to use JSON-LD
- ≈ 0Measured effect of adding JSON-LD in a controlled test
From Rankings to Citations
AI Overviews reduce clicks on traditional results, so ranking alone no longer guarantees traffic. Pew Research found that when an AI summary appeared, users clicked a traditional result in 8 percent of visits, compared with 15 percent when no summary appeared, and clicked a link inside the summary in only 1 percent of visits. Sessions ended on 26 percent of pages with an AI summary versus 16 percent without, a clear sign of zero-click behaviour.
Ahrefs, looking at 300,000 keywords, reported that the presence of an AI Overview correlated with a 34.5 percent lower average click-through rate for the top-ranking page. AI is absorbing attention. To stay visible, you now need both strong rankings and a credible chance of being cited inside the AI answer. Measuring that shift is covered in more depth in the AI SEO hub and the dedicated AI Overviews click quality study.
What Research Says About Structured Data
The studies point in two directions, and the difference is in how they were run.
Observational studies
~3×In Ahrefs’ 2026 analysis of AI citations, pages cited by AI systems were almost three times more likely to use JSON-LD than non-cited pages. GEO-16 found the same pattern across 1,702 citations in Google AI Overviews, Brave and Perplexity, where structured data showed one of the strongest associations with being cited.
Controlled test
≈ 0When Ahrefs tested 1,885 pages that actually added JSON-LD, citations barely moved over 30 days, and none of the changes was statistically distinguishable from zero.
Two statistics now circulate that claim pages in AI Overviews had 23% more structured data and that FAQ markup made pages 31% more likely to be cited. Since the original study URL returns 404 and no archive exists, these figures cannot be treated as evidence.
The gap between strong correlation and near-zero causal effect is not surprising. Sites that implement schema tend to be better maintained, with clearer content and stronger information architecture, which are exactly the kinds of signals AI systems and classic ranking algorithms already reward.
Why Structured Data Matters
Structured data tells machines, in a consistent vocabulary, what a page is and who stands behind it. Three types do most of the work on a content site:
What the page is
Schema.org’s Article type describes an article such as a news story or investigative report, and BlogPosting is a subtype for blog posts. Marking content up this way clarifies that it is a substantive piece with a headline, body, date and author.
Who stands behind it
Organization schema describes the entity behind the site, such as a company or NGO, and ties together your name, logo, URLs and social profiles.
Questions and answers
FAQPage describes a page that presents one or more frequently asked questions and their answers.
Google’s AI features documentation states there are no extra requirements and no special schema needed for AI Overviews, and stresses that structured data must match visible content. The value of schema is clarity and consistency across your site, not a shortcut to rankings or citations.
Does FAQ Schema Improve Citations?
FAQ schema has been promoted as a way to stand out, but its visible benefits have largely disappeared:
- Aug 2023
Limited
Google limited FAQ rich results to well-known government and health sites.
- 7 May 2026
Retired
The FAQ rich result stopped appearing in search.
- Jun 2026
Undocumented
The FAQ documentation was removed. FAQ markup no longer produces any special search feature.
The widely cited 31 percent figure in industry coverage, attributed to an iPullRank study that is no longer online, should be treated as correlation at best. Today, the practical value lies in the underlying format.
Clear question-and-answer sections, written in plain language and visible on the page, give AI systems extractable snippets they can quote, whether or not FAQPage schema is present.
How to Improve AI Search Visibility
- Mark up what is visibly on the page, and validate it.
Use Article or BlogPosting, Organization and, where appropriate, FAQPage, but only for content that actually appears. Validate with tools or a workflow such as the Schema Gap Analyzer.
- Keep entity and author details consistent.
Use Organization schema to represent your brand the same way across pages, and pair it with clear, visible bylines so AI systems can connect content to a stable source.
- Write direct, extractable answers.
Structure sections around specific questions, then answer them in the first sentence or two. This gives ranking systems and AI models concise statements they can quote.
- Publish original evidence and data.
Studies, benchmarks, pricing breakdowns and worked examples give AI systems something distinctive to reference, which observational work like GEO-16 suggests is more likely to be cited.
- Measure AI citations and referrals explicitly.
Track when your brand appears in AI Overviews, AI Mode and chat tools, and monitor referral traffic from those surfaces. For broader tactics, see the guide on how to get cited by ChatGPT.
Conclusion
AI Overviews have shifted search from a list of links to a synthesized answer, so visibility now depends on being a trusted citation source as much as on ranking. Research so far shows strong associations between structured data and AI citations, but controlled tests do not support schema as a direct lever.
What consistently helps is structured, well-maintained content that machines can interpret confidently. Article and Organization schema, clear question-and-answer sections, and original evidence all contribute to that clarity. Treat structured data as infrastructure that makes your expertise legible to AI systems, not as a growth hack.
Frequently Asked Questions
Does structured data help with AI Overviews?
Research like GEO-16 and Ahrefs shows correlations between schema usage and AI citations, but no causal proof. Structured data mainly improves clarity and consistency, which indirectly supports visibility.
A 31 percent uplift figure circulates from an unverified, now-unavailable study, and FAQ rich results are deprecated, so any AI citation benefit is uncertain. Clear on-page Q&A content matters more than FAQPage markup.
What schema types are useful for AI SEO?
Article or BlogPosting for content, Organization for your brand entity, and FAQPage where you genuinely present FAQs. All must reflect visible content, since Google rejects hidden or misleading markup.