Patients no longer Google "dentist near me." They ask ChatGPT which dental practice in Phoenix actually takes their insurance, Gemini whether they need a crown or a filling, and Perplexity which dermatologist has the best reviews for cosmetic work. By the time they reach your website, the recommendation decision has often already been made — by an AI model that did not name your practice.

Healthcare is one of the highest-leverage verticals in AI visibility: patient lifetime values are large, commercial intent is high, and almost no practice has the structured-data, schema, and citation footprint required to be the named recommendation. This guide is the HIPAA-aware playbook for fixing that.

The shift

Why Healthcare Patient Acquisition Broke in 2025

In 2024, the patient journey for healthcare ran almost entirely through Google search, the map pack, and Healthgrades. In 2025, the entry point fragmented. 64% of patients now consult an AI model — ChatGPT, Gemini, Perplexity, or Claude — at some point during provider research, and a growing share use AI as the only research step before booking.

That shift is consequential because of how much each patient is worth. A general dental patient is worth $750 to $3,500 in lifetime value. A cosmetic or implant patient runs $5,000 to $15,000. A med-spa or aesthetic visit averages $400 to $8,000. A specialty MD patient — dermatology, orthopaedics, fertility — can exceed $20,000 over the relationship. Losing the AI recommendation isn't losing a click. It's losing a quarter of a year's mortgage payment.

The mechanic: when a patient asks "best implant dentist in Phoenix," ChatGPT typically names 3 providers. Those three providers capture nearly all the downstream bookings from AI-originated research. The fourth-best practice in the city — clinically excellent, fully booked five years ago — is now invisible to the AI channel.

AI citations are the new "above the fold." Without them, the patient never sees your name. With them, you're the recommendation a friend never made — delivered with the implicit authority of the model itself.

Step 1

Dental & Medical Schema — Use the Right Sub-Type

The single most common technical mistake in healthcare AI visibility is using LocalBusiness as the schema type. It is too generic. AI models use the @type field to decide which category queries you qualify for. A generic LocalBusiness entry is treated as background noise. The specific medical sub-type is treated as an entity.

  1. General Dentist Use schema.org/Dentist. Specialty practices (endodontics, orthodontics) layer the specialty into medicalSpecialty.
  2. Doctor / MD practice Use schema.org/Physician with a populated medicalSpecialty array (Dermatology, OrthopedicTrauma, ReproductiveMedicine, etc.).
  3. Med spa / aesthetic clinic Use schema.org/MedicalBusiness. Add availableService entries for each procedure (Botox, dermal fillers, laser resurfacing).
  4. Veterinarian Use schema.org/VeterinaryCare. Patient-LTV economics are surprisingly similar to dental.
  5. Hospital / multi-specialty clinic Use schema.org/Hospital. Each department can be a child MedicalOrganization entity.

Below is a complete Dentist schema example with the fields that actually move AI citation ranking. Replace placeholder values with your practice data. The same structure adapts directly to Physician or MedicalBusiness by swapping the @type and adjusting medicalSpecialty.

JSON-LD — Dentist Schema 
{
  "@context": "https://schema.org",
  "@type": "Dentist",
  "@id": "https://yourpractice.com/#practice",
  "name": "Practice Name DDS",
  "description": "Cosmetic and implant dentistry in [City], serving patients since [Year].",
  "url": "https://yourpractice.com/",
  "telephone": "+1-555-000-0000",
  "medicalSpecialty": [
    "Endodontics",
    "CosmeticDentistry",
    "DentalImplants"
  ],
  "availableService": [
    {
      "@type": "MedicalProcedure",
      "name": "Dental Implant",
      "procedureType": "https://schema.org/SurgicalProcedure"
    },
    {
      "@type": "MedicalProcedure",
      "name": "Invisalign Clear Aligners"
    }
  ],
  "acceptsReservations": "true",
  "isAcceptingNewPatients": "true",
  "paymentAccepted": "Cash, Credit Card, CareCredit, Insurance",
  "currenciesAccepted": "USD",
  "openingHoursSpecification": [
    {
      "@type": "OpeningHoursSpecification",
      "dayOfWeek": ["Monday","Tuesday","Wednesday","Thursday","Friday"],
      "opens": "08:00",
      "closes": "17:00"
    }
  ]
}

Why isAcceptingNewPatients matters: several AI models now filter local healthcare recommendations by this property. Practices marked as closed to new patients are quietly demoted in citation output. If you're accepting patients, declare it explicitly.

Step 2

Provider Person Schema — the Credibility Layer

Practice schema places the clinic on the map. Provider schema — an individual Person entity for each dentist, MD, NP, or hygienist — places the clinician in the recommendation. This is what makes a query like "best implant dentist in Phoenix" return a named provider rather than a clinic name.

Every provider on staff should have a dedicated bio page with the schema below in the <head>. The high-signal fields are alumniOf (dental school, medical school), memberOf (ADA, AMA, state board, specialty societies), award (Top Dentist, Castle Connolly, magazine awards), and knowsAbout (the procedures they personally perform).

JSON-LD — Provider Person Schema 
{
  "@context": "https://schema.org",
  "@type": "Person",
  "@id": "https://yourpractice.com/team/dr-smith#person",
  "name": "Dr. Jane Smith, DDS",
  "jobTitle": "Cosmetic & Implant Dentist",
  "worksFor": { "@id": "https://yourpractice.com/#practice" },
  "workLocation": { "@id": "https://yourpractice.com/#practice" },
  "alumniOf": [
    { "@type": "CollegeOrUniversity", "name": "UCLA School of Dentistry" }
  ],
  "memberOf": [
    { "@type": "Organization", "name": "American Dental Association" },
    { "@type": "Organization", "name": "American Academy of Cosmetic Dentistry" }
  ],
  "award": [
    "Top Dentist, Phoenix Magazine 2024",
    "Fellow, International Congress of Oral Implantologists"
  ],
  "knowsAbout": [
    "All-on-4 Implants", "Porcelain Veneers", "Full-Mouth Reconstruction"
  ]
}
Step 3

HIPAA & AI Content — the Dos and Don'ts

Healthcare is the only AIO vertical with a federal compliance layer. The AI optimisation tactics that work for a plumber or restaurant can quietly create HIPAA violations if applied without filter. The rules are not complicated, but they are non-negotiable.

Patient testimonials require HIPAA-compliant written authorization before publication — even when the patient enthusiastically offers one. Before-and-after photos require explicit, specific consent. FAQ content cannot reference identifiable patient cases, even anonymously, if details could identify the individual. Generic procedure FAQs are safe and perform equally well in AI citation ranking.

Use disclaimers in AI-discoverable content: "Results vary. Individual outcomes depend on anatomy, healing response, and adherence to post-procedure care." This is required for FTC compliance on cosmetic and elective content, and AI models trained on medical content respect properly disclaimed claims.

Avoid absolute medical claims. "We cure migraines" is non-compliant; "we treat chronic migraine with Botox under FDA-approved indications" is accurate, safer, and ranks better because AI models recognise medically precise phrasing as a quality signal.

Step 4

The Procedure × Condition FAQ Strategy

This is the single highest-leverage content asset in healthcare AIO. AI models reward content that mirrors how patients actually phrase their searches — procedure name plus condition plus city plus a buying-stage qualifier (cost, recovery, comparison). FAQPage schema lifts citation selection by 2.1× on these queries.

A short list of high-converting FAQ patterns to seed:

  • "How much does Invisalign cost in [city] in 2026?"
  • "Is dental bonding or veneers better for chipped teeth?"
  • "How long does a root canal take and does it hurt?"
  • "How do I know if I need a crown vs a filling?"
  • "What's the recovery time for LASIK?"
  • "When should I see a dermatologist about a mole?"
  • "Botox vs Dysport — which is right for me?"
  • "Can chiropractic care help sciatica?"

Reading level matters: write at a sixth-grade reading level. Patients are not clinicians. AI models trained on patient-facing medical content reward plain phrasing — "your gum tissue" beats "the gingival margin" in citation selection on consumer queries. Save clinical precision for the disclaimer.

Step 5

The Healthcare Directory Triad

Generic local-business directories (Yelp, BBB, Yellow Pages) carry less weight in healthcare entity resolution than vertical-specific medical directories. AI models cross-reference provider claims against the directories below; appearing in a critical mass of them is the entity-trust threshold.

WebMDDA 92
BBBDA 92
HealthgradesDA 88
ZocdocDA 85
VitalsDA 79
RealselfDA 78
ADA Find-a-DentistDental
1-800-DentistDental

Weight by specialty: Zocdoc is the single highest-leverage directory for any practice that can integrate online booking — the booking button is itself an AI trust signal. Realself is a near-mandatory listing for aesthetic and cosmetic practices and largely irrelevant for general medicine. ADA and AAFP directories carry disproportionate weight for the specialties they cover because AI models recognise them as authoritative professional bodies.

For every directory you appear on: claim the listing, complete every field, add a full provider bio, list every procedure offered, and upload at least eight photos. Half-finished listings underperform no listing at all because they create a conflict signal in entity resolution.

Step 6

Google Business Profile — the Booking-Button Play

Google Business Profile remains the single biggest healthcare visibility lever because Gemini, Google's AI Overviews, and Google Maps all pull from it directly. For healthcare, four GBP optimisations matter more than the other twenty combined.

Enable the Book button. Integrate with NexHealth, Zocdoc, or Weave so patients can book from the GBP card without leaving Google. The presence of a working booking button is itself an AI signal — it tells Gemini you are actively accepting patients.

Add every service as a separate item. Not "general dentistry," but "dental crowns," "porcelain veneers," "single-tooth implants," "All-on-4 implants," "Invisalign," and so on, each with its own short description and price range when possible.

Seed the Q&A. GBP allows the owner to ask and answer questions. Pre-populate the top ten patient questions yourself. Unanswered questions get answered by strangers — sometimes inaccurately.

Photos weekly. Reviews answered within 24 hours. Photo upload cadence is a freshness signal. Review response speed is a quality signal. Both are observable to AI models scraping GBP.

Step 7

The 150-Review Threshold (and How to Hit It Ethically)

Across the AI citation studies we've run on healthcare queries, the consistent finding is a threshold effect at 150+ reviews. Below that, you are rarely cited even with excellent schema and complete directory presence. Above it, citation rates rise sharply. Hitting and maintaining the threshold is non-negotiable for AI-channel patient acquisition.

Ethical tactics that work: automated post-visit SMS via your practice management system — Dentrix, Eaglesoft, Open Dental, Athena, NextGen — sent 2 to 4 hours after the appointment. A QR code at the checkout counter linking directly to your Google review form. Zocdoc's built-in post-visit review prompt. NexHealth's automated review request workflow.

Critical compliance note: never offer incentives in exchange for reviews. This violates HIPAA (when the incentive is tied to PHI use), FTC guidelines (undisclosed material connection), and the terms of service of every major review platform. Request neutral language: "How was your experience today?" — not "Would you mind leaving us a 5-star review?"

Timeline

Timeline & Expected Outcomes

Healthcare AIO timelines are predictable. A practice that begins implementation on day one and works through the steps below in sequence typically sees first measurable patient acquisition lift between months three and four.

  1. Week 1–2 — Schema deployment + Bing submission Deploy practice schema (Dentist / Physician / MedicalBusiness) and a Person schema for every provider. Submit sitemap to Bing Webmaster Tools. This starts the AI index clock.
  2. Week 3–4 — GBP to 100% + ten procedure-FAQ pages Complete every GBP field, enable the booking button, add all services. Publish ten procedure FAQ pages with FAQPage schema targeting your top procedures by patient LTV.
  3. Month 2 — 150+ reviews + healthcare directory triad Activate the review-request workflow. Claim and fully complete Healthgrades, Vitals, Zocdoc, WebMD, and any specialty directories relevant to your practice.
  4. Month 3 — First AI citations on procedure queries Begin appearing in ChatGPT, Gemini, and Perplexity answers for procedure + city queries. Niche, specific procedure queries cite first. Generic "best dentist in [city]" takes longer.
  5. Month 4–6 — 6–15 new patients per month from AI Citation footprint compounds. Most practices reach a steady-state of 6–15 net-new patients per month attributable to AI-originated research, with higher numbers for elective and aesthetic specialties.
Common questions

Frequently Asked Questions

Healthcare AIO

See where your practice stands in ChatGPT, Gemini & Perplexity

Run the free AI Visibility Diagnostic — we audit your schema, provider entities, GBP, review velocity, and simulate patient queries across all 7 AI platforms. Or jump straight to the Dominant tier built for healthcare economics.

Run the free diagnostic Get the Dominant tier

Dominant: $1,497 / mo  ·  Built for healthcare LTV  ·  No long-term contract