A plastic surgery practice gets cited by AI assistants when its site states, in machine-readable form, who the surgeon is, what the practice treats, and where those facts are confirmed elsewhere. The work has three parts: an llms.txt file, Physician, MedicalClinic, MedicalProcedure, and FAQPage schema, and consistent sameAs and directory profiles.
AEO, GEO, AI visibility, and AI SEO are four names for the same work: getting a practice named and cited inside AI answers. This post uses AEO and covers only the hands-on technical layer. The strategy-level question of why AEO matters and what it changes is covered in the ClinicAds post on [AI search visibility for plastic surgeons](/blog/ai-search-plastic-surgery-visibility), and this post does not repeat it.
What does the technical layer of AEO actually include?
The technical layer of AEO has four artifacts: an llms.txt file at the site root, JSON-LD schema on practice, surgeon, procedure, and FAQ pages, a set of sameAs links that tie the surgeon to outside profiles, and a directory footprint whose name, address, and credential text matches the site. Each artifact reduces ambiguity about one entity.
None of the four is a ranking trick. Each one removes a guess an AI system would otherwise make, such as whether Dr. Rivera on the practice site is the same Dr. Rivera on the board certification directory. Content quality still decides whether a page is worth citing. The technical layer decides whether the right entity gets the credit.
- llms.txt: a plain-language map of the practice, placed at the domain root
- Schema: JSON-LD that names the practice, surgeon, procedures, and questions
- sameAs: links from the site entity to the same entity on outside profiles
- Directory consistency: identical name, address, phone, and credential wording everywhere
What is llms.txt, and does it help a surgeon get cited?
llms.txt is a markdown file proposed in September 2024 by Jeremy Howard that gives language models a short, curated summary of a site. It sits at the domain root, for example example.com/llms.txt. No major AI provider has published a commitment to read it, so ClinicAds treats it as low-cost insurance and a clarity exercise, not a proven ranking lever.
The file is worth an hour of work for a practice anyway. Writing it forces the practice to state its identity, procedures, and service area in plain sentences, and those same sentences become the source for the schema and directory text. A practice that cannot write the file in one page usually has an entity problem to fix first.
ClinicAds publishes its own file at clinicads.com/llms.txt, which opens with a one-sentence definition, a disambiguation paragraph, and a list of key pages. The same shape works for a practice.
What should a plastic surgery llms.txt contain?
A plastic surgery llms.txt should contain the practice name, a one-sentence definition, the surgeon names with credentials, the city and service area, the procedures performed, the key pages with one-line descriptions, and a disambiguation sentence. Keep it under 80 lines, written as facts, with no promotional adjectives and no claims the practice cannot document.
The disambiguation sentence matters more than it looks. Dr. Michael Chen is a common name, and an AI answer that merges two surgeons with the same name is a real failure mode. State the full name, the city, the board, and the practice domain in one sentence so the file separates the surgeon from namesakes.
- Line 1: the practice name as an H1, matching the signage and the board listing exactly
- Line 2: a one-sentence definition, such as a board-certified plastic surgery practice in a named city performing facial, breast, and body procedures
- Section: surgeons, each with full name, certifying board, and years in practice if verifiable
- Section: procedures performed, one line each, linking to the procedure page
- Section: service area and consult process, including the consult fee if the practice publishes one
- Section: disambiguation, one sentence naming the domain and city
Which schema types does a plastic surgery site need?
A plastic surgery site needs four schema.org types in JSON-LD: MedicalClinic (or its parent MedicalBusiness) for the practice, Physician for each surgeon, MedicalProcedure for each procedure page, and FAQPage for pages with visible questions. Schema.org defines all four, and all four can be validated with the Schema Markup Validator at validator.schema.org.
Each type goes on the page that is actually about that entity. The MedicalClinic block belongs on the home and contact pages, the Physician block on each surgeon biography, the MedicalProcedure block on each procedure page, and FAQPage only where the questions and answers appear in the page text. Schema that describes content a visitor cannot see is a violation of Google's structured data guidelines.
| Schema type | Page it belongs on | Properties to fill | What it settles |
|---|---|---|---|
| MedicalClinic | Home, contact | name, url, address, telephone, geo, medicalSpecialty, sameAs | Which practice this is and where it operates |
| Physician | Each surgeon biography | name, medicalSpecialty, memberOf, alumniOf, worksFor, sameAs | Which surgeon this is, and which practice employs them |
| MedicalProcedure | Each procedure page | name, procedureType, bodyLocation, preparation, followup | What the practice performs, in standard terms |
| FAQPage | Pages with visible Q&A | mainEntity of Question and Answer pairs | Which questions the page answers, and the exact answer text |
How do you write Physician schema for a surgeon?
Physician schema names the surgeon, links the surgeon to the practice with worksFor, lists the specialty as PlasticSurgery, records the certifying body in memberOf, and carries sameAs links to the board directory, state license lookup, and professional society profile. Each property should repeat text that already appears on the biography page word for word.
Two properties do most of the work. The worksFor property ties the surgeon to the MedicalClinic entity by its @id, so a model reads one connected graph rather than two unrelated blocks. The sameAs array points to profiles that independently confirm the surgeon exists, which is the schema-level version of the consistency check a patient runs in ChatGPT after seeing a name.
- name: the surgeon's name exactly as the board directory spells it, including credentials such as MD
- medicalSpecialty: PlasticSurgery, using the schema.org enumeration value
- worksFor: the practice, referenced by the same @id used on the home page
- memberOf: the certifying board and any society membership the surgeon can document
- sameAs: the board directory profile, society profile, and any state license lookup page that carries a stable URL
Should procedure pages carry MedicalProcedure and FAQ schema?
Procedure pages should carry MedicalProcedure schema and, where the page shows real questions, FAQPage schema. MedicalProcedure describes what the page is about in standard terms, and FAQPage hands a model exact question-and-answer pairs. Since August 2023 Google limits FAQ rich results to well-known government and health sites, so the FAQ markup is for AI parsing, not for SERP decoration.
That limit changes the expectation, not the task. A practice should not add FAQ schema hoping for a SERP feature. It should add the markup because the visible questions on a rhinoplasty or facelift page are the same questions patients type into an AI assistant, and the markup labels the answer text cleanly. Keep each answer between 40 and 75 words, state the procedure by name, and match the markup to the visible text exactly.
How do sameAs links and directory profiles build the entity?
The sameAs property and directory profiles build the entity by letting an AI system confirm the surgeon on several independent sources. When the site, the board directory, the society profile, and the review platforms all give the same name, address, phone number, and credential, a model has no conflict to resolve. When they disagree, the model hedges or omits the practice.
The audit is mechanical. List every place the practice and each surgeon appear, copy the name, address, phone, and credential text into one spreadsheet, and mark every difference. Fix the practice's own pages first, then the profiles the practice controls, then request edits where it does not. Typical differences are a suite number, a retired phone line, and an abbreviation in a degree.
- Certifying board directory: confirm name spelling, certification status, and practice address
- Professional society profile: confirm membership wording and practice address
- Google Business Profile: match the practice name, categories, hours, and phone to the site
- Review and healthcare directories: match name, address, phone, and specialty
- Hospital or surgery center affiliation pages: match the surgeon's name and the practice name
How do you check that the setup works?
A practice checks the setup in three steps: validate every schema block, confirm llms.txt loads at the root with a 200 status, and run a fixed set of prompts in ChatGPT, Perplexity, and Gemini each month to log whether the practice is named. Validation proves the markup is well formed. Only the prompt log shows whether the engines changed their answers.
Expect a lag. ClinicAds' post on [how long AEO takes for a plastic surgery practice](/blog/how-long-does-aeo-take-to-work-for-a-plastic-surgery-practice) lays out the month-by-month timeline, and the post on [measuring AI visibility for a plastic surgery practice](/blog/measure-ai-visibility-plastic-surgery-practice) owns the prompt-log protocol. For a practice paying $80 to $150 per booked consult in paid media, each consult an AI answer produces without ad spend is worth that avoided cost. These are agency averages, not guarantees.
This setup is one part of the search visibility system ClinicAds builds for [plastic surgery practices](/surgery), alongside paid media and speed-to-lead. It is not sold as a separate service.
- Validate every JSON-LD block at validator.schema.org and fix every error before launch
- Fetch /llms.txt in a browser and confirm a 200 response and plain-text rendering
- Search the practice and surgeon names in three AI assistants and record the answers
- Re-run the same prompts every 30 days and log changes against the schema and profile edits
Does a plastic surgery practice need llms.txt to appear in ChatGPT?
No. No major AI provider has committed to reading llms.txt, and practices are named in ChatGPT answers without one. ClinicAds recommends the file because it is cheap to write and forces the practice to state its identity clearly, which supports the schema and directory work.
Which schema type is most important for a plastic surgeon?
Physician schema on each surgeon biography, linked to the MedicalClinic entity with worksFor and to outside profiles with sameAs. It settles which surgeon the practice means, which is the fact a patient verifies after an AI answer names the surgeon.
Will FAQ schema produce rich results on Google for a surgery site?
Usually not. Since August 2023 Google shows FAQ rich results only for well-known government and health authority sites. The markup still labels question-and-answer text cleanly for AI systems, so it remains worth adding when the questions are visible on the page.
Can the practice's web developer handle this setup?
Yes for the mechanics. A developer can publish the file and the JSON-LD in a day. The harder work is the content: confirming credentials, matching directory text, and writing procedure answers a surgeon is willing to sign. That review belongs to the practice, not the developer.