AIOInsights / AI Discoverability / Does structured data help AI find your business?
AI Discoverability

Does structured data help AI find your business?

Sometimes, a little, and only if it tells the same story as the page it sits on.

Picture a dental practice in Ventura that drops Saturday mornings, updates the footer of its website and forgets the block of JSON-LD a developer pasted into the page header two years earlier. From that day on, every machine that reads the page is handed two answers to the same question. That is the real risk in schema markup.

So here is the honest version: what the companies that build these systems actually say, what markup can plausibly do for AI discoverability, and what it cannot. If you want the definition first, our lexicon covers structured data and how JSON-LD works.

What the companies actually document

Google

Google's guidance on AI features is direct: "There are no additional requirements to appear in AI Overviews or AI Mode," and "There's also no special schema.org structured data that you need to add." Separately, its introduction to structured data says Google "uses structured data that it finds on the web to understand the content of the page," and recommends JSON-LD. So markup is a way to be understood by Google, not a ticket into its AI answers.

OpenAI, Anthropic and Perplexity

We read the crawler documentation from OpenAI, Anthropic and Perplexity. Each explains which bots fetch pages and for what purpose. None of them says whether its assistant reads schema markup or gives it any weight. So we make no claim either way, and neither should anyone who cannot point to where it is written.

What markup can do, stated carefully

Structured data states facts in a fixed vocabulary instead of leaving them to be inferred from layout. For a system that does read it, "the phone number in the footer belongs to this business" stops being a guess. It can also point at your identity elsewhere: Schema.org defines sameAs as the "URL of a reference Web page that unambiguously indicates the item's identity," which is how a page can say that this practice and that Google Business Profile or that directory listing are the same entity. None of this replaces the visible sentence a person reads. It repeats it for machines, which is a real but modest contribution to AI discoverability.

The types that matter for a local business

  • LocalBusiness, in its most specific form. Google's local business documentation says to "use the most specific LocalBusiness sub-type possible." A dental practice has one: Schema.org defines a Dentist type that sits under both LocalBusiness and MedicalBusiness. Google lists name and address as required, and telephone, url, opening hours and geo coordinates among the recommended properties.
  • sameAs links to profiles that are genuinely yours, so the entity on your site connects to the entity everywhere else.
  • FAQPage, only where the page really shows those questions and answers to a reader.

One type to handle with care: review stars. Google's review snippet guidelines say that if the entity being reviewed controls the reviews about itself, its pages using LocalBusiness or Organization markup are "ineligible for star review feature."

Why markup that disagrees with the page is worse than none

Go back to the Ventura practice. With no markup, a machine reads the footer and gets the right hours. With stale markup, it gets Saturday in one place and no Saturday in the other, and nothing on the page says which to believe. Google's structured data policies are clear on the principle: "Don't mark up content that is not visible to readers of the page," and a structured data issue "can result in a manual action," under which "a page loses eligibility for appearance as a rich result." Markup that contradicts your page is a second, wrong answer carrying your name, and our page on what happens when your details disagree explains why conflicting facts hurt.

A twenty minute check

  • Paste your homepage into Google's Rich Results Test and the Schema Markup Validator.
  • Read every value they report: name, address, phone, hours, type. Compare each one against the visible page, not against memory.
  • Delete anything that is wrong or unverifiable before adding anything new. A missing field is a gap; a wrong one is a contradiction.
  • Put a note in your calendar: whenever hours, address or phone change, the markup changes the same day.

Markup is a small piece of the job, and it only works when the page underneath it is already clear. To see whether your page states its facts in a form a machine can read, run the free AIOInsights check, which reads the JSON-LD on the page and compares it with what the page says.

Questions

Does schema markup help a business appear in AI answers?

Schema markup helps a business appear in AI answers only indirectly and without any guarantee. Google says there is no special schema.org structured data needed to appear in AI Overviews or AI Mode, though Google does use structured data to understand page content. OpenAI, Anthropic and Perplexity do not state in their crawler documentation whether their assistants read schema, so claims that markup gets a business into AI answers go beyond what those companies have published.

Which structured data should a local business add first?

A local business should add one LocalBusiness block first, using the most specific subtype that fits, such as Dentist, with the name and address Google lists as required and the telephone, url and opening hours it recommends. Every value must match what the visible page says, because Google tells site owners not to mark up content that readers of the page cannot see.

More on AI discoverability

This page is part of AI Discoverability, the AIOInsights guide to whether AI systems can find, read and name a business.

Run the free AIOInsights check