Why AI search engines struggle to understand health businesses
When someone asks ChatGPT or Perplexity "find me a physiotherapist in Manchester that accepts private patients," the AI has to work out who provides that service, where they are, what conditions they treat, and whether they are credible. That is a lot of inference to make from unstructured text alone.
Most clinic websites give AI crawlers very little to work with. There is a homepage with a warm headline, a services page with some bullet points, and a contact page with a Google Maps embed. None of that tells a language model with any certainty that your business is a registered physiotherapy clinic, what your opening hours are on a bank holiday, or whether you treat sports injuries specifically.
This is exactly the problem that MedicalBusiness schema solves. It gives AI crawlers a structured, machine-readable description of your health practice so they do not have to guess. The result is that your clinic becomes far more likely to be cited, recommended, or surfaced in AI-generated health answers.
What MedicalBusiness schema actually is
MedicalBusiness is a type defined in the Schema.org vocabulary. It sits under the LocalBusiness hierarchy, which means it inherits all the standard local business properties like address, opening hours, and telephone number, but it also unlocks health-specific properties that generic business schema simply does not support.
The full inheritance chain looks like this: Thing > Organization > LocalBusiness > MedicalBusiness. Because it extends LocalBusiness, you do not need to choose between them. MedicalBusiness gives you everything LocalBusiness does, plus more.
There are also a number of subtypes you can use depending on your specific practice type:
- Dentist - for dental practices
- Optician - for optometry businesses
- Pharmacy - for dispensing pharmacies
- Physician - for GP practices or private consultants
- MedicalClinic - for multi-specialty clinics
- IndividualPhysician - for sole-practitioner doctors
Using the most specific subtype available is always better than using the generic MedicalBusiness type. If you run a dental practice, use Dentist. AI systems are trained to understand these distinctions and they will categorise your business more accurately as a result.
The core properties you need to include
Getting MedicalBusiness schema right is not about stuffing in every possible property. It is about covering the properties that AI systems actually rely on when formulating health-related answers. Here are the ones that matter most.
@type and @context
Always declare both. Your @context should be https://schema.org and your @type should be the most specific subtype relevant to your practice.
name, url, and description
The name should be your official registered business name. The url should point to your homepage or the specific page being marked up. The description field is particularly valuable for AI visibility because it gives the model a concise, authoritative summary of what you do. Write it in plain English, be specific about your specialities, and keep it between 100 and 200 characters.
address using PostalAddress
Structured address data is non-negotiable. Break your address into its component parts using the PostalAddress type: streetAddress, addressLocality, addressRegion, postalCode, and addressCountry. AI systems use this data to answer location-based health queries like "find a GP near Leeds LS1."
telephone and openingHoursSpecification
Include your main contact number. For opening hours, use openingHoursSpecification rather than the simpler openingHours string format. The specification format allows you to define different hours for different days, which is important for practices that have reduced Saturday hours or are closed on Wednesdays, for example.
medicalSpecialty
This is a property unique to medical types and it is one of the most powerful for AI visibility. Schema.org defines a fixed list of MedicalSpecialty values including Cardiology, Dermatology, Physiotherapy, Psychiatry, and many others. Use the relevant ones for your practice. You can supply multiple values as an array.
availableService
This links your MedicalBusiness to one or more MedicalTherapy or MedicalProcedure entities. For a physiotherapy clinic, this might include ManualTherapy, SportsRehabilitation, or AcupunctureTherapy. For AI systems trying to match a user's query to the right provider, this level of specificity is very helpful.
priceRange and paymentAccepted
Include priceRange using a simple notation like "££" or a string like "from £80 per session". Use paymentAccepted to specify whether you accept private insurance, NHS referrals, or self-pay. These details matter when AI answers include practical guidance like "here are private physio clinics in Bristol that accept self-referrals."
A working JSON-LD example for a physiotherapy clinic
Here is a complete, ready-to-use example you can adapt for your own practice. This would be added inside a <script type="application/ld+json"> tag in the <head> or <body> of your relevant page.
{
"@context": "https://schema.org",
"@type": "MedicalClinic",
"name": "Northgate Physiotherapy Clinic",
"url": "https://www.northgatephysio.co.uk",
"description": "Private physiotherapy clinic in Leeds specialising in sports injury rehabilitation, back pain, and post-surgical recovery.",
"telephone": "+441132345678",
"medicalSpecialty": [
"https://schema.org/Physiotherapy",
"https://schema.org/SportsRehabilitation"
],
"availableService": [
{
"@type": "MedicalTherapy",
"name": "Manual Therapy"
},
{
"@type": "MedicalTherapy",
"name": "Sports Massage"
},
{
"@type": "MedicalProcedure",
"name": "Biomechanical Assessment"
}
],
"address": {
"@type": "PostalAddress",
"streetAddress": "42 Albion Street",
"addressLocality": "Leeds",
"addressRegion": "West Yorkshire",
"postalCode": "LS1 5AA",
"addressCountry": "GB"
},
"openingHoursSpecification": [
{
"@type": "OpeningHoursSpecification",
"dayOfWeek": ["Monday", "Tuesday", "Wednesday", "Thursday", "Friday"],
"opens": "08:00",
"closes": "19:00"
},
{
"@type": "OpeningHoursSpecification",
"dayOfWeek": "Saturday",
"opens": "09:00",
"closes": "13:00"
}
],
"priceRange": "££",
"paymentAccepted": "Cash, Card, Private Health Insurance",
"currenciesAccepted": "GBP",
"image": "https://www.northgatephysio.co.uk/images/clinic-front.jpg",
"sameAs": [
"https://www.facebook.com/northgatephysio",
"https://www.instagram.com/northgatephysio"
]
}Notice the use of sameAs to link out to your social profiles. This helps AI systems cross-reference your entity and build confidence that your business is genuine and established.
Where to place the schema on your site
For most health practices, the right page to place MedicalBusiness schema is your homepage or your main "About" or "Clinic" page. If you have multiple locations, create a separate page for each location and add location-specific schema to each one. Do not apply the same address details across all pages if those pages represent different sites.
If you are on WordPress, you can add JSON-LD manually using a plugin like WPCode, or by editing your theme's functions.php file to inject the script into the page head. If you are on Shopify and have a health or wellness store, the schema can be added via a custom snippet in your theme's theme.liquid file.
Avoid relying on generic SEO plugins to generate your MedicalBusiness schema automatically. Most of them do not support the health-specific properties like medicalSpecialty or availableService, which means you end up with watered-down LocalBusiness markup that does not differentiate your practice at all. At FlinnSchema, this is exactly the kind of gap we focus on, building schema that goes beyond what plugins can do automatically.
Common mistakes health businesses make with schema
Using LocalBusiness instead of a medical subtype
Hundreds of clinics have LocalBusiness schema applied by their web developers. It is better than nothing, but it tells AI systems nothing specific about the medical nature of the business. Always use the most specific type available.
Missing the medicalSpecialty property
This is the single most important differentiator in health schema. Without it, your clinic looks identical to a generic service business. With it, AI systems can accurately match your practice to queries about specific conditions and treatment types.
Inconsistent NAP data
Your Name, Address, and Phone number must match exactly across your schema, your Google Business Profile, and your website footer. Even small inconsistencies, like "St." versus "Street" in the address, can reduce an AI system's confidence in your business data.
No image or sameAs links
Entity disambiguation is a big part of how AI systems decide who to cite. Adding an image property and linking to your verified social profiles via sameAs strengthens your entity footprint considerably.
Testing and validating your implementation
Once you have added your schema, test it using Google's Rich Results Test at search.google.com/test/rich-results. This will confirm whether your JSON-LD is valid and flag any missing recommended properties. For a deeper check, paste your schema into Schema.org's validator at validator.schema.org to see the full property tree and spot anything unexpected.
You should also check whether AI crawlers are actually visiting your site and indexing your structured content. This guide on checking which AI crawlers are visiting your site walks through how to inspect your server logs and identify GPTBot, PerplexityBot, and other AI agents.
If you want a full picture of how your site looks to AI search engines right now, FlinnSchema's free AI visibility audit is a good place to start. It covers schema coverage, entity clarity, and crawlability in one report.
How MedicalBusiness schema fits into a broader AI visibility strategy
Schema markup is one part of the puzzle. AI systems also look at your content quality, your entity mentions across the web, and how clearly your pages answer specific health questions. A clinic that has solid MedicalBusiness schema AND well-structured FAQ content AND a strong entity presence on third-party directories will outperform a clinic that only has one of those things.
Think of schema as the foundation. It tells AI systems what you are. Your content then tells them what you know and who you help. Together, they give AI a complete enough picture to confidently cite your practice in a health search answer.
For health practices that want to go further with structured data, it is worth exploring FAQPage schema to trigger AI answer boxes on your treatment and condition pages. That combination of MedicalBusiness schema on your clinic page and FAQPage schema on your service pages is one of the most effective setups we have seen for AI health search visibility.
Frequently Asked Questions
Is MedicalBusiness schema only for large hospitals and NHS trusts?
Not at all. MedicalBusiness schema is well-suited to sole-practitioner clinics, small private practices, physiotherapy studios, dental surgeries, opticians, and wellness centres. The schema types like IndividualPhysician and Dentist are specifically designed for smaller, independent health businesses.
Will adding MedicalBusiness schema affect my Google rankings?
Google does not guarantee ranking boosts from schema alone, but structured data does improve how your business is understood and categorised. More practically, better AI visibility means more referrals from ChatGPT, Perplexity, and Gemini, which is increasingly where health searches are starting. It also supports rich results in Google Search for knowledge panels and local packs.
Do I need to register with any authority to use MedicalBusiness schema?
No. Schema.org is an open vocabulary and anyone can use it. That said, accuracy matters enormously in health contexts. Only declare medical specialties and services you genuinely offer, and make sure your regulatory body registration details are correct if you choose to include them using properties like hasCredential.
How long does it take for AI search engines to pick up new schema markup?
It varies. GPTBot and PerplexityBot crawl on their own schedules, which are not publicly documented in the way Googlebot's is. In practice, most sites see AI crawlers revisiting within a few weeks of a significant content or schema change. Submitting your sitemap and ensuring your robots.txt does not block AI crawlers will help speed things along.