Does Schema Markup Help You Get Cited in AI Overviews?

What Google, Microsoft and OpenAI say about structured data and AI answers, which schema types still produce results, and where schema causes problems.

Gunakesh Parmar Published 4 min read

Schema markup does not get you into AI Overviews by itself. Google’s documentation says there is no special structured data needed to appear in AI Overviews or AI Mode; a page only has to be indexed and eligible to show with a snippet. What schema does is describe your page, your company and your authors in a form machines read without guessing, which helps search engines and AI systems understand and identify you correctly. That makes it useful, just not a shortcut.

What Google says

Google’s page on AI features and your website states that SEO best practices remain relevant for AI Overviews and AI Mode, that there are no additional requirements or special optimisations to appear in them, and that no special schema.org structured data is needed. Google’s general documentation on structured data says it helps Google understand the content of a page and can make a page eligible for rich results.

Those two statements fit together. Structured data is part of normal SEO. It is not an extra key for AI features.

What other search companies have said

Microsoft has been more positive. Fabrice Canel of Microsoft Bing said at SMX Munich in March 2025 that schema markup helps Microsoft’s language models understand content. Bing’s index feeds Microsoft Copilot, so for Copilot the benefit is stated more directly. OpenAI’s guidance for publishers focuses on crawler access and does not mention structured data as a requirement for ChatGPT search.

Which schema types still produce visible results in Google

Schema type Status in Google Search
Product, Offer, Review (for products) Supported for product rich results and merchant listings
Article, BlogPosting Used to understand articles, dates and authors
Organization, LocalBusiness Used for company details, logo and knowledge panel information
BreadcrumbList Supported for breadcrumb display
FAQPage Rich results limited to well-known government and health websites since August 2023
HowTo Rich results no longer shown

For most businesses, FAQ schema no longer creates an FAQ box in Google. It can still describe a page accurately, but adding it for rich results is wasted effort.

Where schema helps AI visibility

Identifying your company and people

Organization and Person schema with sameAs links to LinkedIn, Wikidata and other official profiles tell machines that these profiles belong to one entity. This matters most when your brand name is shared with another business. See our guide to entity resolution.

Connecting authors to articles

An Article whose author points to a Person node, which points to your Organization, connects the content to the people and company behind it. This supports the experience and expertise signals described in E-E-A-T. This site does the same: each post’s author is linked to a co-founder’s Person node.

Exact product facts

Prices, availability, shipping and return policies in Product schema give systems precise values that are easy to extract correctly.

Where schema causes problems

A common issue is two tools each writing their own schema. A theme outputs an Organization node, an SEO plugin outputs another, and the two disagree:

# Example: two Organization nodes on one page (illustrative)
{ "@type": "Organization", "name": "Example Consulting", "logo": ".../logo-old.png" }
{ "@type": "Organization", "name": "Example Consulting Pvt Ltd", "logo": ".../logo-2026.png" }
# Two names and two logos for the same company

Other common problems:

  • Schema describing content that is not visible on the page, which goes against Google’s guidelines
  • Review ratings on service pages that do not come from real, independent customer reviews
  • Authors given as a plain name instead of a Person node
  • dateModified changing on every save, even for small edits

The fix is one source of truth. The Brilad theme outputs its own graph only when no SEO plugin is active, and switches it off automatically when Yoast, Rank Math or a similar plugin is installed, so the site never carries two versions.

How to check your schema

  1. Run key pages through Google’s Rich Results Test for Google-supported types.
  2. Run them through the Schema Markup Validator at validator.schema.org for all types.
  3. View the page source and search for “application/ld+json”. Count the Organization nodes. More than one is a warning sign.
  4. Compare the names, logos and addresses in the schema with your visible page and your profiles.

Frequently asked questions

Should I remove FAQ schema?

Not necessarily. If the questions and answers are visible on the page, the markup is accurate and harmless. Just do not expect an FAQ rich result unless you are a well-known government or health site.

Does schema help in ChatGPT?

OpenAI has not said it uses structured data to choose sources. Schema that helps Google and Bing identify your company can still help indirectly, since ChatGPT search draws on search results.

Is JSON-LD better than Microdata?

Google recommends JSON-LD because it is easier to maintain. Both work if they are correct. More in our JSON-LD guide.

Sources

  1. Google Search Central, “AI features and your website”
  2. Google Search Central, “Introduction to structured data markup in Google Search”
  3. Google Search Central blog, “Changes to HowTo and FAQ rich results”, August 2023
  4. OpenAI Help Center, “Publishers and Developers FAQ”

Related reading