Interest in schema markup for AI search has grown quickly, and so have the myths. Some guides present structured data as a switch that puts a site inside AI answers. It is not. What schema does is remove ambiguity: it tells a machine, in a fixed vocabulary, that this page is about this organization, this service, this author and this address. Below we separate what is documented from what is assumed, and list the types we actually implement.
What schema markup is
Schema markup is code that uses the vocabulary published at schema.org, a project started by Google, Microsoft, Yahoo and Yandex, to label entities and their properties. Google recommends the JSON-LD format, a script block that sits in the page without changing what visitors see. A LocalBusiness block, for example, states the business name, address, phone, opening hours and the profiles that belong to it. The same facts are on the page in plain text; the markup simply makes them unambiguous.
What Google says about schema markup for AI search
Google's guidance on AI features in Search says there are no additional requirements to appear in AI Overviews or AI Mode beyond the usual eligibility for Search, and that no special schema is needed. Its structured data policies also require that markup matches content visible on the page. In other words, schema markup for AI search works through the same channel as schema for regular search: it helps systems understand the page, and the content still has to earn its place.
Two changes in 2023 are worth knowing because many older guides still recommend otherwise. Google limited FAQ rich results to well known government and health websites, and it removed HowTo rich results. FAQPage markup is still valid vocabulary and can still describe a page, but it no longer produces the expanded result most sites used to get.
Why structured data still matters for AI answers
AI systems that write answers need to decide whether a page is about the entity the user asked about. Consistent structured data helps with that decision in three ways:
- Entity identity: Organization markup with a sameAs list connects your site to your official profiles, which reduces confusion with similarly named businesses.
- Fact consistency: the address, services and hours in the markup match those on the page and on your Google Business Profile.
- Authorship: Article markup with a Person author links content to a real expert, which supports the trust signals discussed in E-E-A-T.
OpenAI, Perplexity and other assistant providers have not published detailed documentation on how they use schema. Any crawler that reads the raw HTML can see JSON-LD, but we do not claim more than that.
Schema types worth implementing
| Type | Where it goes | What it clarifies |
|---|---|---|
| Organization | Home page or about page | Legal name, logo, official profiles through sameAs |
| LocalBusiness | Pages of businesses with a physical location or service area | Address, phone, hours, area served |
| Service | Each service page | Service name, provider, area served |
| Article with Person | Blog posts and guides | Author, publication and update dates |
| BreadcrumbList | All inner pages | Where the page sits in the site structure |
| Product | Product pages | Name, brand, offers, reviews from real customers |
Mistakes that undo the benefit
- Marking up content that is not on the page. This breaks Google's structured data policies and can lead to a manual action.
- Self-serving reviews. Google does not show review stars for reviews a business publishes about itself on LocalBusiness or Organization markup.
- Conflicting facts. A phone number in schema that differs from the one in the footer weakens the very signal you are trying to send.
- Plugin defaults left unchecked. Many CMS plugins generate generic markup with wrong business types or missing fields.
How we implement and validate it
We start by listing the entities that matter for the business, then map one schema type to each page template, write the JSON-LD from the visible content and test it in Google's Rich Results Test and the Schema Markup Validator. After launch, Search Console's enhancement reports show errors as Google recrawls. The work is part of how our generative engine optimization services prepare a site for AI answers, alongside the content and authority work explained in what generative engine optimization is. For pages built around questions, the writing guidance in what answer engine optimization is matters more than any markup.
Treat schema markup for AI search as plumbing: invisible when it works, costly when it is wrong. If you want a second look at yours, book a call with our team and we will review your structured data against your visible content.