GEO for medical aesthetics in Tijuana: we audited ten clinics, only one has the schema that gets cited.

The vertical, in numbers.

Tijuana is the largest cross-border medical aesthetics hub in the world. The city has more than 200 dental clinics serving international patients (Heva) and dozens of plastic surgery practices catering to American patients crossing for procedures that cost 40-70% less than the equivalent in the US. Bariatric, dental, plastic surgery, and dermatology are the four core sub-verticals. The driver hasn't changed for fifteen years: cost. The discovery path has.

An American patient considering rhinoplasty in Tijuana in 2026 increasingly opens ChatGPT or Perplexity before they call a clinic. The model produces a synthesized answer naming 3-5 surgeons, summarizing accreditation, recovery times, average cost ranges, and patient experiences. The patient then narrows from there.

The clinic named in that synthesis isn't always the best one. It's the one whose website the model could read and trust enough to attribute. We wanted to know — concretely, with real data — which Tijuana aesthetic clinics that is right now, and which it isn't.

The audit: ten clinics, May 18, 2026.

We ran our public 5-Foundation scanner against ten of the most-recognized Tijuana medical aesthetic clinic websites, sampled across plastic surgery, dental, and dermatology. The sites:

• evoclinic.mx — plastic surgery
• cosmeticsurgeryinmexico.com — plastic surgery (Dr. Manuel Gutierrez Romero)
• viveplasticsurgery.com — plastic surgery
• drdiegoampudia.com — plastic surgery
• cerplasticsurgery.com — plastic surgery (Hospital CER)
• mexicocosmeticcenter.com — plastic surgery
• tijuanadentistcenter.com — dental (Dr. Mexico)
• advancedsmilesdentistry.com — dental
• thedentaldistrict.com — dental
• trustdentalcare.com — dental

Selection criterion: clinics with public-facing English-language sites identifiable through standard cross-border medical tourism search. We're not ranking these clinics on quality of care — these are operators investing real money in their public presence, and the patterns below are characteristic of the vertical at this moment. They're not flaws of any individual practice.

The headline finding: one clinic passes.

Only drdiegoampudia.com scored 4 of 5 Foundations in our audit. The site has:

• PlasticSurgery schema — the Schema.org vertical-specific type. None of the other plastic surgery sites in our sample had this.
• MedicalOrganization schema for the clinic itself.
• FAQPage schema with patient-facing questions in structured form.
• Five different schema types in total. Clean metadata. Clean structure (one H1, multiple H2s, FAQ).

The only Foundation Dr. Ampudia's site misses is llms.txt — a fifteen-minute fix, not a content-strategy problem. If the rest of the Tijuana aesthetic vertical caught up to this site's technical floor, the cross-border AI search landscape would look completely different than it does now.

By contrast, the next-best score in our sample — three clinics tied at 3 of 5 — got there with looser schema and less structure. The median was 2 of 5. The vertical is uneven.

The aggregate, by Foundation.

• Foundation 1 (Crawler access): 9 of 10 pass. One fails. We name the one below.
• Foundation 2 (Schema.org): 3 of 10 pass at the scanner's full-pass threshold (3+ relevant types). Seven score warn or fail. The most schema-rich site in the sample — trustdentalcare.com, with fifteen schema types including Dentist, MedicalClinic, and MedicalOrganization — still didn't reach a full pass because it lacked FAQPage. Schema richness alone doesn't earn citations; the specific types matching patient queries do.
• Foundation 3 (llms.txt): 2 of 10 pass. Adoption in aesthetics is meaningfully behind medical tourism, where 3 of 5 clinics we sampled have it. Aesthetics hasn't caught the file yet.
• Foundation 4 (Page metadata): 10 of 10 pass. Every site has clean basic metadata at the homepage level. Hreflang and bilingual coverage is uneven on closer inspection, but the homepage scan passes universally.
• Foundation 5 (Content structure): 1 of 10 pass — drdiegoampudia.com. Every other site fails or warns because none have FAQPage schema. Same pattern we saw in the medical tourism audit.

Median: 2 of 5. Top: 4 of 5 (one clinic). Bottom: 2 of 5 (six clinics tied).

The bot-blocking finding.

The Dental District (thedentaldistrict.com) is blocking the critical AI crawlers in robots.txt as of May 18, 2026. ChatGPT, Claude, Perplexity, and Google's AI-Overview crawler all return "blocked" on our scan. Whatever content, schema, surgeon credentials, and patient outcomes the clinic has built — invisible to these models. They could be the best dental clinic in Tijuana and a patient asking ChatGPT for recommendations would never see them named.

We don't know whether the blocking is intentional or inherited from a managed hosting template. We've audited individual clinics where this exact pattern — CCBot or GPTBot blocked by default on a Cloudflare or Wix template, without anyone realizing — turned out to be the single broken thing. The fix is a one-line robots.txt change. If the blocking is intentional, that's a strategic position worth examining; if it's accidental, that's the highest-leverage fix in the entire audit.

The 5 Foundations, applied to a Tijuana aesthetic clinic.

The same five technical foundations apply to aesthetics as to medical tourism, but the schema specifics shift. Here's what each looks like for plastic surgery and dental practices specifically.

Foundation 1 — Crawler access. The robots.txt entries that explicitly allow AI bots: GPTBot, ChatGPT-User, OAI-SearchBot, ClaudeBot, Claude-Web, PerplexityBot, Perplexity-User, Google-Extended, Bytespider, CCBot, Applebot-Extended, Meta-ExternalAgent. One block in robots.txt. Universal pass except where managed templates block by default.

Foundation 2 — Schema for aesthetics specifically.

• Plastic surgery practice: PlasticSurgery (the dedicated Schema.org type), Physician per surgeon with sameAs linking to CMCPER, MedicalProcedure per procedure (rhinoplasty, breast augmentation, liposuction, mommy makeover, brazilian butt lift, etc.), MedicalOrganization for the clinic.
• Dental practice: Dentist, MedicalClinic, Person per practicing dentist, MedicalProcedure per procedure (all-on-4 implants, veneers, crowns, root canals, full-mouth reconstruction).
• Dermatology practice: Dermatologist, MedicalSpecialty, MedicalProcedure per procedure (botox, fillers, laser, mole removal, melasma treatment).

The most schema-rich site in our sample (trustdentalcare.com) had Dentist, MedicalClinic, Person, MedicalOrganization, and a dozen other types — and still didn't pass because it didn't have FAQPage. Specificity to patient queries beats raw type count.

Foundation 3 — llms.txt. A short Markdown file at the root listing the homepage, each procedure landing page, each surgeon page, the accreditation page, the patient-experience or testimonial page, and the FAQ. Fifteen minutes to create. The cost of not having it is no confirmed harm and a small opportunity cost as the standard solidifies.

Foundation 4 — Page metadata. Title, description, og:image, og:title, og:description, canonical, hreflang. For bilingual aesthetic clinics — most of them are — the hreflang piece is uneven. American patient lands on Spanish page. Spanish-only patient lands on English page. Model gives up on both.

Foundation 5 — Content structure with FAQPage schema. One H1 per page, two or more H2s, and FAQPage schema covering the 15-20 questions every cross-border aesthetic patient asks: board certification (CMCPER for plastic surgery, specific dental boards for dentistry), hospital accreditation, recovery timeline, English-speaking staff in surgery and recovery, total all-in cost, complication handling, post-op follow-up by telemedicine, payment methods. Every question answered in structured form is a citation hook the model can attribute to the clinic by name.

What we'd ship this quarter for the average clinic in this sample.

For the seven clinics in the sample scoring 2 of 5, in order:

One: ship FAQPage schema. Same play as in medical tourism. 12-18 patient-facing questions with structured-data Q&A markup on the homepage and each procedure landing page. Half a day of structured-data work. Largest single move.

Two: add vertical-specific schema types. PlasticSurgery for plastic surgery practices, Dentist for dental, Dermatologist for dermatology. Then MedicalProcedure blocks for each procedure actually performed. Most of the clinics in our sample have Organization or LocalBusiness — generic. The vertical-specific types are what tells the model "this is a plastic surgery practice, not a generic local business." Another half day.

Three: add llms.txt. Minutes. Should point to homepage, each surgeon page, each procedure landing page, the team page, the accreditation/credentials page, and the patient-experience or testimonial pages.

Four: audit and fix hreflang on bilingual pages. Half a day if the site is small, two days if the architecture is sprawling. For aesthetic clinics that serve both Mexican and cross-border patients, this is load-bearing — the American searches in English, the Mexican in Spanish, and they should land on the right language version with proper canonical/alternate signals.

And the one clinic — The Dental District — that's blocking the AI crawlers should change one line in robots.txt before any of the above. It's the single highest-leverage fix in the audit.

Why Dr. Ampudia's site matters as a case study.

A clinic site that scores 4 of 5 in May 2026, in a vertical where the median is 2 of 5, has a real first-mover advantage on AI citations. Perplexity already weighs structured-data-rich sites more heavily; ChatGPT and Claude will follow as their indexes catch up over the next 30-90 days. By the time the rest of the vertical catches up — and they will, this is solvable engineering, not strategic moat — Dr. Ampudia's site will have months of accumulated citation weight.

This is the asymmetric window. 2026 to 2027 is the period where Tijuana aesthetic clinics that ship the technical floor enter the cross-border patient research path already cited. After 2027, the floor is universal and competition moves back to where it always has been: content depth, surgeon authority, and patient-experience signal. The technical lead is a 12-to-18-month head start, not a permanent moat.

The 5 Foundations are not strategy. They are the floor.

None of what we've described is a marketing strategy. It's a technical floor — what we call The 5 Foundations of AI visibility, the same five our free scanner measures, the same five we build into every client engagement. Without them, content depth and surgeon authority don't compound into citations. With them, the next levers (procedure-specific content, surgeon profile pages with credentials, patient-experience markup, external mentions on directory listings and review platforms) start producing measurable lift in AI Share of Voice 30-90 days after they ship.

The aesthetic medicine clinics in Tijuana that get serious about this between 2026 and 2027 enter the cross-border patient research path while the gap is still open. Anyone who waits until it's obvious arrives when the rest of the vertical has also figured it out — the same curve as 2010 SEO with long-tail keywords, except this time the lever is the citation, not the rank.

The American patient asking Perplexity tonight about a rhinoplasty surgeon in Tijuana doesn't see the best surgeon. The patient sees the surgeon whose site the model could read. One Tijuana clinic figured this out already. Nine haven't.