How schema markup fits into search AI – without the hype

Does schema markup really benefit search AI development? Some suggest that it can triple your quotes or greatly increase AI visibility. But if you dig into the evidence, the picture is very different.

Let’s separate the known from the assumed, and look at how schema fits into an AI search strategy.

How schema fits into search AI now

Search goes from SERP exposure with green links to AI overviews, generating answers, and conversation-style summaries that include content in addition to links.

To make your content visible in this model, your site should be understood as entities – unique objects, unique things or concepts, such as a person, place, or event – and the relationship between them, not just strings of text.

Schema markup is one of the few tools SEOs have to make those associations and relationships clear and understandable to AI: This is a person, he works for this organization, this product is offered at this price, this article is endorsed by that person, etc.

In AI, three things are very important:

• Business description: What brands, authors, services, or SKUs are on the page.
• Attribute clarity: Which properties belong to which business (eg prices, availability, estimates, job titles).​
• Business relationship: How businesses communicate (eg. offeredBy, worksForauthoredBy, and sameAs schema tags).

When the schema is created with stable values ​​(@id) and structure (@graph), it starts to behave like a small graph of internal information.

AI systems won’t have to guess who you are and how your content fits together, and will be able to follow the obvious connections between your brand, your authors, and your topics.

Dig deep: Why business authority is the foundation of AI search visibility

How AI search platforms use schema

Two major platforms have confirmed that schema tagging helps their AIs understand content. In these platforms, infrastructure is verified, not speculation.

What about ChatGPT, Perplexity, and other AI search platforms?

We don’t know how these platforms use the schema yet. They haven’t publicly confirmed whether they store the schema during web crawling or use it to extract it. The technical capabilities exist for LLMs to process structured data, but that doesn’t mean their search systems do.

Dig deeper: When and how to use information graphs and organizations for SEO

Research on schema and AI

Here are a few case studies that show how schema can benefit AI search.

Quotation rates

A December 2024 study from Search/Atlas found no correlation between schema markup and citation rates. Sites with extensive schema did not perform as well as sites with little or no schema markup.

This doesn’t mean that schema is useless, it just means that schema alone doesn’t drive quotes. LLM systems seem to prioritize relevance, subject authority, and semantic clarity over whether content has systematic remedies.

A February 2024 Nature Communications study found that LLMs extracted information more accurately when given structured instructions with defined fields versus unstructured “giveaway” instructions.

Put differently, LLMs do best when you give them a structured form to fill out, not a blank canvas. When models are asked to extract from predefined fields, they make fewer errors than when they are told to “extract what is important.”

Schema markup on a page is the web equivalent of that form: a set of obvious business, type, product, price, author, and title fields that a program can map to, instead of saying everything in unstructured prose.

What the research tells us

This tells us that LLMs have the technical ability to process structured data more accurately than unstructured text.

However, this does not tell us whether AI search engines preserve schema markups during web crawling, or whether they use them to guide the rendering of web pages, or whether this results in better visibility.

Jumping from “LLMs can process structured data” to “web schema markup improves AI search visibility” requires assumptions that we cannot validate in most platforms.

In Microsoft Bing and Google AI Overviews, the schema probably improves the accuracy of extraction, as they have confirmed that they use it. For other platforms, we have no guarantee of actual usage.

Dive deep: Business SEO first: How to align content with Google’s Knowledge Graph

What we don’t know about schema and AI search

AI search is so new – for example, ChatGPT search was only launched in October 2024 – that companies have yet to disclose their targeting methods. Measurement is difficult with indeterminate AI responses. There are significant gaps in what we can confirm.

To date, there are no peer-reviewed studies on the effect of schema on AI search visibility, or controlled experiments on LLM citation behavior and schema markup.

OpenAI, Anthropic, Perplexity, and other platforms other than Microsoft or Google have not yet published their targeting methods.

This gap exists because AI search is really new (ChatGPT search was launched in October 2024), companies do not disclose targeting methods, and measurement is difficult with undefined AI responses.

How a schema forms a business graph

In traditional SEO, most implementations stop at adding Article or Organization he marks himself. In AI search, the most useful pattern is to connect nodes in a parallel graph using @id. Example:

• An Organization a node with a stub @id that represents your product.
• A Person author node that works for your organization.
• An Article place authoredBy that person again publishedBy that organization, about the properties that declare the main topics.

{  
  "@context": "  
  "@graph": [  
    {  
      "@id": "  
      "@type": "Organization",  
      "name": "Example Digital"  
    },  
    {  
      "@id": "  
      "@type": "Person",  
      "name": "Jane Doe",  
      "worksFor": { "@id": " }  
    },  
    {  
      "@type": "Article",  
      "@id": "  
      "headline": "Schema Markup for AI Search",  
      "author": { "@id": " },  
      "publisher": { "@id": " }  
    }  
  ]  
} 

That connected pattern transforms your schema from a set of disconnected schemas into a reusable business graph. In any AI system that supports JSON-LD, it becomes clearer which type owns the content, who is responsible for it, and what the top-level topics are about, regardless of how the page layout or copy changes over time.​​

Dig deep: How structured data supports spatial visibility across Google and AI

Recommendations for using AI search schema

In search AI, the best way to set up a schema is currently to:

• Make businesses and relationships machine-readable on platforms that store and use structured data (proven in Bing Copilot and Google AI Overview).
• Reduce ambiguity about brand, author, and product ownership so that the release, if it happens, is clean and consistent.
• Fill in the subject’s depth, authority, and clear brand signals, don’t change them.

Use the schema tag:

• Improving visibility in Bing Copilot.
• Support for inclusion in Google AI Overview.
• Improving traditional SEO.
• Making content easier to parse (good practice without AI).
• Maintaining a low-cost implementation is likely to result as platforms evolve.

However, don’t expect:

• Verified quotes on ChatGPT or Perplexity.
• A dramatic elevation of visibility in the schema only.
• Schema to compensate for weak content or low authority.

Important schema types (based on the field guide) include:

• Organization (brand corporate identity).
• Article or BlogPosting (attribute to content and composition)
• Person (author’s credentials and business connections).
• Product or Service (commercial transparency).
• FAQPage (Q&A content formats).

Dive deep: Business home: The page that shapes how search, AI, and users see your product

Use AI search schema today

Schema markup is infrastructure, not a magic bullet. It won’t necessarily get you more citations, but it’s one of the few things you can control that platforms like Bing and Google AI Overviews use clearly.

Real opportunity is not a solitary schema. It is a combination of structured data with relevant business relationships, high quality, subject-authorized content, clear business identity and brand signals, and strategic implementation. @graph again @id building business links.