Google says AI Overviews and AI Mode need no special schema.org markup, so you should mark up what each page actually shows rather than hunting for an AI-specific type (1). For schema markup for AI search, Google's Organization documentation says that adding Organization structured data to your home page can help Google better understand your organization's administrative details and disambiguate it in search results (2). Both must match the visible content of the page they describe (1).
Which schema markup should I add so AI search can read my pages?
Google's documentation says that specific optimisation isn't required for AI Overviews and AI Mode, and that there's no special schema.org structured data you need to add (1). Google also states that existing SEO fundamentals still apply, including making sure your structured data matches the visible text on the page (1). In practice, that means marking up the entities and page structure you already show readers (3).
Start with Organization on your home page (2). Google recommends using the most specific schema.org subtype of Organization that matches your organization, and for a local business such as a restaurant or a physical store, the most specific LocalBusiness subtype. An ecommerce site should use OnlineStore rather than OnlineBusiness, and a physical business should use the most specific LocalBusiness subtype and follow that type's required and recommended fields. Google describes the Organization structured data as adding administrative details that can influence visual elements in Search results, such as which logo is shown. Google recommends including a logo property that represents your organization, which can help it choose the logo to show in Search results and knowledge panels. Google suggests adding a sameAs URL that points to another website with additional information about your organization, where applicable. Google describes a VAT code in Organization markup as an important trust signal, because users can look up your business in public VAT registries.
For navigation, add BreadcrumbList (4). Google says a BreadcrumbList must contain at least two ListItems, and each ListItem needs the required properties to be eligible for display with breadcrumbs. Google also recommends breadcrumbs that represent a typical user path to a page rather than mirroring the URL structure. Google also gives the BreadcrumbList position rule: position 1 marks the beginning of the trail.
The limit here is that Google's statement that no special schema.org structured data is needed is made about AI Overviews and AI Mode, and the passage addresses no other AI engine (1).
What implementation rules keep my markup readable and eligible?
Google recommends JSON-LD if your site's setup allows it, because it is the easiest format to implement and maintain at scale (3). JSON-LD is placed in a script tag in the head or body of the HTML page and is not interleaved with user-visible text. Google can also read JSON-LD that is dynamically injected by JavaScript or by an embedded widget in your content management system.
Do not block your structured data pages from Googlebot using robots.txt, noindex, or any other access control methods (5). If items on the same page are more useful when linked, use @id in both items to show how they relate. Google also requires that any image URLs in structured data be crawlable and indexable.
Test during development with the Rich Results Test and check technical compliance with the URL Inspection tool (5). After deployment, monitor the Rich result status reports, because templating or serving issues can break pages that were valid when first published (3). Google's guidelines say that items missing required properties for a rich result type are not eligible for rich results (5). Google asks that the structured data describe content the page actually shows, so the HTML body must describe the same performer or entity the markup describes.
JSON-LD block in a page's script tag shows the recommended format with a single simple item, so you can see where the markup sits before checking each item in the table (3).
Markup to add and how to check each item
| Markup | Rule | How to verify |
|---|---|---|
| Organization | Add it to the home page; there are no required properties, so include the relevant recommended ones. | Rich Results Test in development, Rich result status reports after deployment. |
| BreadcrumbList | Define at least two ListItems and include the required properties. | Rich Results Test in development, Rich result status reports after deployment. |
| LocalBusiness subtype | For a physical business, use the most specific LocalBusiness subtype for administrative details. | Rich Results Test and URL Inspection tool, which catch most technical errors. |
| JSON-LD format | Google recommends JSON-LD in a script tag in the head or body; it can be injected dynamically. | Rich Results Test in development, Rich result status reports after deployment. |
| Linked items on one page | Use @id in both items so Google knows they are linked. | Rich Results Test and URL Inspection tool. |
| Page placement | Put the structured data on the page that it describes. | Rich Results Test and URL Inspection tool. |
| Crawl access | Do not block structured data pages with robots.txt, noindex or other access controls. | URL Inspection tool and Rich Results Test. |
Example: JSON-LD block in a page's script tag
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "Organization",
"name": "Example Organization",
"url": "https://www.example.com",
"logo": "https://www.example.com/logo.png"
}
</script>What should I not expect from schema markup, and which vocabulary should I avoid?
Valid markup does not guarantee display (5). Google states that it does not guarantee that structured data will show up in search results, even when a page is marked up correctly according to the Rich Results Test. Google also says it is more important to supply fewer but complete and accurate recommended properties than to provide every possible property with incomplete or inaccurate data (3).
For Google Search behaviour, treat Google's own documentation as definitive rather than schema.org, because schema.org defines many attributes that Google does not require (3). Data-vocabulary.org markup is no longer eligible for Google rich result features, so migrate any such markup to supported types.
When you choose terms, avoid those still in schema.org's pending section, since they may lack consensus and could change significantly after review (6). Avoid terms in the attic section, which are archived when deprecated or removed. Schema.org states that implementors are cautioned not to use attic terms.
The limit here is that rich result eligibility depends on the required properties of each Google-documented type (5). Google also prohibits using structured data to impersonate a person or organization or to misrepresent your ownership, affiliation or primary purpose.
Next step
Continue with Recala to turn this plan into a concrete next step.
Sources
AI Features and Your Website, Google Search Central, last updated 10 December 2025developers.google.com
Organization Schema Markup / Google Search Central / Documentation / Google for Developers, Google, last updated 8 September 2026developers.google.com
Intro to How Structured Data Markup Works, Google Search Central, last updated 10 December 2025developers.google.com
How To Add Breadcrumb (BreadcrumbList) Markup / Google Search Central / Documentation / Google for Developers, Google, last updated 8 September 2026developers.google.com
General Structured Data Guidelines / Google Search Central / Documentation / Google for Developers, Google, last updated 10 July 2026developers.google.com
Schemas - Schema.org, Schema.org, accessed 10 November 2026schema.org