From the Journal AI Search: AEO & GEO

Schema Markup for Medspas: The Fields AI Actually Reads

Which structured-data fields are worth checking on a medspa website? We map identity, location, and page relationships while separating documented behavior from AI assumptions.

Schema Markup for Medspas: The Fields AI Actually Reads

At a glance

  • No universal list reveals which fields every AI system reads or how much each field matters.
  • Use structured data to describe verified entities and visible page content consistently.
  • Test syntax, policy eligibility, and factual accuracy separately after every meaningful change.

A schema plugin can produce a green validation result while describing the wrong business. Another site can have valid markup and never receive the rich appearance its owner expected. Structured data is valuable, but its role becomes distorted when it is sold as a secret instruction that forces AI to recommend a clinic.

The answer in plain language

Medspa schema should identify the real business, its verified locations, and the content each page actually presents. Useful fields describe names, canonical URLs, contact details, services, and page relationships. No public universal specification reveals which fields every AI system reads, and Google requires no special AI schema for its search features.

Separate documented use from an AI claim

The title's practical question is which fields deserve attention. We cannot inspect every AI system's internal weighting, and a field's presence does not prove it influenced an answer. Google's documentation describes supported search features and explains that structured data helps classify page information.

That gives us a concrete task: make the markup accurately describe the page and business. It does not justify promising a citation, adding unsupported properties, or claiming that a particular field unlocks an AI ranking. Read the guidance for the feature you are implementing rather than relying on a plugin's label.

Our AI visibility work for medspas treats schema as one part of a readable, accessible website. Useful treatment information and a clear consultation route remain necessary even when every technical field validates.

Prepare a verified entity fact sheet

Before editing code, collect the practice's approved public name, canonical website URL, contact details, real locations, and official profiles. Assign someone to confirm each field. If a location or credential is uncertain, leave the claim out until the appropriate person resolves it.

Distinguish the organization from each physical location and from individual providers. They are related entities, not interchangeable labels. A provider biography should not become the business address record, and a service offered at one location should not silently appear as available everywhere.

This preparation also protects against copy-and-paste errors. Example code often contains a plausible address, rating, or opening schedule. Those details are demonstration data, not defaults. Never publish them simply because the validator accepts the surrounding structure.

Check business identity fields first

  • @id: Use a stable identifier for the entity so related pages can reference the same business rather than create contradictory duplicates.
  • name: Match the real public business name rather than inserting a string of target keywords.
  • url: Reference the canonical public destination, with the site's chosen trailing-slash policy.
  • logo and image: Use real, accessible assets that represent the entity or page appropriately.
  • telephone and sameAs: Include verified contact details and genuine official profiles, not guessed directory links.

These fields make identity explicit. They do not establish medical expertise or commercial results. Keep organization references consistent across the homepage, service pages, and articles so a template change does not generate a new incompatible business record on every URL.

Describe locations only when they are real

For an appropriate local-business implementation, verify address components, public telephone details, and opening information against the actual practice. Google's local-business documentation identifies feature-specific required and recommended properties. Check those requirements rather than assuming any schema type produces the same result.

Use the most appropriate supported type for the real business and representation. A broad category is preferable to an inaccurate clinical designation. Do not invent a Google Business Profile URL, a suite number, coordinates, or hours to complete a checklist. Missing information is a fact-gathering task.

For multiple locations, keep the relationships clear and match visible location pages. A patient's question about where to attend should have a readable answer without inspecting JSON. Structured data supports that answer; it should not be the only place the relevant location exists.

Map the page to its actual purpose

A service page can describe the service it visibly explains and reference the actual provider entity. A breadcrumb describes the navigational hierarchy. An article should identify its headline, publication information, real authoring organization, and canonical page. These roles should not be collapsed into one oversized business object.

Use Article or an appropriate subtype for genuine editorial content, not every sales landing page. Preserve the original publication date and record a modification date when the article changes meaningfully. Do not refresh dates automatically to suggest an editorial review that never happened.

Our broader medspa schema implementation guide provides background on the vocabulary. The GEO foundation guide explains how these page relationships support the wider content system.

Keep FAQ and review markup within their limits

FAQ markup should reflect questions and answers visitors can access. A closed-by-default accordion can still provide the content in the page. A separate hidden set written only for search systems creates a mismatch. Generate visible answers and schema from the same data where practical.

Google restricts FAQ rich results to well-known, authoritative government or health websites. Publishing a health-related page does not itself establish eligibility. A matching FAQ implementation can organize content correctly without producing an expanded search appearance.

Reviews need equal care. Do not copy ratings from another platform into a practice's structured data to pursue stars. Google restricts self-serving review features for businesses, including certain embedded third-party reviews. A truthful testimonial and eligibility for a search feature are different questions.

Use a field-review handoff instead of guessing

Send this template to the person responsible for the website: “For [canonical URL], please compare the rendered structured data with our approved public facts. Confirm the business identifier, name, URL, location references, and page type. Flag duplicate entities, unsupported claims, and any property that is not represented accurately in the visible page.”

Ask for a short change log: old value, new value, reason, approving person, and test result. If the developer finds conflicting plugin output, resolve the source of duplication rather than adding a third script. The goal is one coherent description, not the greatest possible amount of JSON.

Make ownership explicit after the initial setup. Operations should tell the website team when public hours or service availability changes. The person maintaining the template should know which source supplies each field. Otherwise, a correct launch can slowly become inaccurate as a clinic changes while its markup stays frozen.

For a small site, an approved spreadsheet may be enough to coordinate those facts. For a larger site, structured content fields can reduce repeated entry. Either way, updates need a review step and a public-page check. Automation helps distribute confirmed information; it should not guess which facts are true or publish unapproved clinical statements.

Test syntax, visibility, and truth separately

First inspect the public HTML and confirm that the expected script is present. Use an appropriate schema validator for vocabulary and a Google testing tool for supported rich-result checks. Read warnings in context; an optional missing field is not permission to invent its value.

Then compare the markup with the rendered page and approved fact sheet. Finally, confirm that the canonical URL is accessible and eligible for indexing. A valid script on a blocked or duplicate page does not repair the page's access problem.

The thirty-minute AI visibility audit checklist places this review alongside indexing, internal links, patient answers, and booking checks. That sequence helps the team prioritize real obstacles instead of treating validation as the end of the project.

The payoff is a maintainable description of the practice and its content. When hours, locations, or services change, the same approved facts can update both the page and its markup. We gain clearer implementation and fewer contradictions without claiming access to a hidden AI recommendation formula.

Set the next review around a real change rather than an arbitrary schema score. A new location, a retired service, or a revised booking destination all justify checking the affected entities and pages. Small factual checks at those moments are easier to maintain than discovering months later that a template has been publishing outdated information.

Frequently Asked Questions

Which schema fields does every AI system read?

There is no public universal field list or weighting formula for every AI system. Google documents how it uses supported structured data and says its AI features do not require special schema. We prioritize accurate business identity, page relationships, and visible facts rather than claim knowledge of hidden model behavior.

Does a medspa need special AI schema?

Google says no special schema is required for its AI features. Use appropriate existing vocabulary to describe the real business and page. Structured data should support the visible information, not introduce an invented service, location, review, credential, or outcome.

Can FAQ schema include answers hidden from visitors?

Do not add full questions and answers that visitors cannot access on the page. An accordion can contain accessible answers while remaining closed initially. Its structured data should match those same answers. Google also limits FAQ rich results, so matching markup does not establish eligibility for a special display.

Does passing the Rich Results Test confirm our information is true?

No. A tool can identify supported markup and some implementation problems, but it cannot verify the practice's address, service availability, credentials, or medical statements. Compare the public page with an approved fact sheet and review search-platform policies separately from syntax.

Sources & citations

  1. Understand how structured data works. Google Search Central.
  2. AI features and your website. Google Search Central.
  3. General structured data guidelines. Google Search Central.
  4. Local business structured data. Google Search Central.
  5. Review snippet structured data. Google Search Central.

Let’s talk about your search.

Tell us what your practice wants patients to find. We’ll look at your treatments, locations, and where search can help.

Book a Consultation