AI Visibility for Plastic Surgery Clinics

Cosmetic surgery patients research for weeks and ask long, specific questions. Engines often turn those questions into their own searches: OpenAI says ChatGPT search typically rewrites a prompt into one or more targeted queries. You can't see those rewritten queries, but you can test the prompts patients are likely to type.

These are hypothetical examples of the kinds of prompts worth testing:

• "Who are the best board-certified plastic surgeons for rhinoplasty in Denver?"
• "How much does a tummy tuck cost in Miami, and what does the price include?"
• "Is Dr. [name] board certified, and what do patients say about her breast augmentation results?"
• "Which cosmetic surgery clinics near me offer financing for liposuction?"
• "Which surgeons in Chicago do deep plane facelifts, and how long is recovery?"
• "Compare [Clinic A] and [Clinic B] for BBL safety and reviews."
• "Does [clinic] operate in an accredited surgical facility?"

They fall into three groups: discovery (who should I see), cost (what will it cost and what's included) and verification (is this surgeon qualified and safe). Verification prompts matter most for a clinic, because a wrong answer there costs you a patient who already knew your name.

Next move: write eight to ten prompts in your patients' own words, covering all three groups and your top procedures.

Each engine draws on its own index and its own crawler, so one healthy channel tells you nothing about the others. As of October 2026:

• Google AI Overviews and AI Mode. Google says a page must be indexed and eligible to show in Search with a snippet to appear as a supporting link, with no additional technical requirements. Local procedure searches often start here.
• ChatGPT. Search answers depend on OAI-SearchBot access. ChatGPT also answers some prompts without searching, from training data frozen at a knowledge cutoff. A surgeon who joined after that cutoff, or a price you changed, won't appear in those answers whatever you do this quarter.
• Gemini. The Gemini app's grounding is governed by the Google-Extended control, separate from Google Search.
• Perplexity and Claude. Both search the web with their own crawlers (PerplexityBot and Claude-SearchBot) and cite sources in their answers.

What you test through an API can differ from what a patient sees in the app, which is worth understanding before you read any report. See how API answers differ from consumer app answers.

Next move: run your prompt list on each engine and record the results per engine, never as one blended number.

AI platforms get specific facts wrong about specific businesses. For a plastic surgery clinic, these are the facts worth checking first:

• Board certification. Which board certified each surgeon, and whether "board certified" refers to plastic surgery or another specialty.
• The surgeon roster. Who currently operates at the clinic, and who performs which procedure. Departed surgeons linger on old directory pages.
• Procedure prices. Starting prices, what they include (anaesthesia, facility fee, follow-up visits) and consultation fees. Old forum posts and cost-comparison pages keep outdated numbers alive.
• Facility details. Locations, whether surgery happens in an accredited facility, and hospital privileges.
• Financing and policies. Payment plans, deposits and revision policies.

Errors usually come from three places: your own site contradicting itself (the pricing page says one thing, a blog post another), third-party pages that went stale, and training data older than your last change. The first one is entirely yours to fix.

The pattern is close to what medical practices and dental clinics face, with higher stakes on credentials.

Next move: pick one canonical page per fact, then make every other page on your site match it word for word.

A before-and-after gallery is a set of images. Vercel found that ChatGPT and Claude crawlers fetch JavaScript files but don't execute them, so they can't read content that only appears after scripts run. Googlebot does render JavaScript. That means a "Meet the surgeon" tab, a price calculator or a slider caption loaded by script may be invisible to most AI crawlers.

State the facts as plain text in the HTML:

• Each surgeon's full name, certifying board, hospital privileges and the procedures they perform.
• Your pricing approach. If you publish prices, give starting prices and what they include. If you quote only after consultation, say that in one sentence and explain what drives the quote.
• Facility accreditation and locations.
• Procedure names and the surgeon in a text caption beside each gallery, not only inside the image.

A specific fact on a page that can be fetched gives an engine a sentence it can quote. That is the only positive experimental signal for on-page content, and it promises nothing about selection.

Next move: load each key page with JavaScript turned off, or view the raw source, and confirm the credentials and pricing text are there.

When engines cite other sites instead of yours, being named accurately on those sites is the route into the answer. That's an association in the evidence, not a guarantee. Check these places for stale or wrong facts:

• Review platforms and physician rating sites.
• Medical and business directories, including local listings.
• The certifying board's public verification lookup.
• Hospital affiliation pages listing your surgeons.
• Cost-comparison and medical tourism sites that quote prices.
• Forums where patients discuss your clinic.

Look for departed surgeons, old prices and the wrong board. Update the profiles you control and request corrections on the rest. Our guides on whether reviews affect AI recommendations and keeping review platform facts accurate cover the details, and this page on directories covers listings.

Some pages try to steer AI answers with hidden text or injected instructions. That is illegitimate, and it is also a risk to you: a third-party page can misrepresent your clinic this way. Never use it, and watch for it.

Next move: list the ten third-party pages that show up most for your clinic name and audit each one.

Access is where the evidence is strongest. Run these checks before touching content.

robots.txt, per bot

OpenAI says each of its crawler settings is independent. GPTBot governs training only. OAI-SearchBot controls whether you appear in ChatGPT search. A blanket AI block can remove you from ChatGPT search without anyone noticing. A minimal allow block looks like this:

User-agent: OAI-SearchBot
Allow: /

User-agent: PerplexityBot
Allow: /

User-agent: Claude-SearchBot
Allow: /

Firewall and CDN

OpenAI's help centre says eligibility means allowing OAI-SearchBot and confirming your host or CDN allows traffic from its published IP addresses. A firewall can return 403 to a crawler that robots.txt allows. Cloudflare changed AI crawler defaults for some customers on 15 September 2026, so check even if you changed nothing. 99Visibility's free AI Crawler Checker requests a page as each AI crawler and shows which ones get turned away.

Snippet controls

noindex, nosnippet, max-snippet:0 and data-nosnippet keep content out of AI answers. Clinics sometimes add them to hide patient photos and accidentally hide surgeon bios too. Confirm every one is intentional, and check the generative AI setting in Google Search Console if your property has it.

1. Days 1 to 7: baseline. Write your prompt list and run it on each engine. The free checker on this site shows whether engines name you for three buyer questions. Record every wrong fact by engine.
2. Days 8 to 14: access. Fix robots.txt rules per bot, test firewall responses for each AI crawler, and remove any accidental noindex or nosnippet.
3. Days 15 to 21: facts in HTML. Publish surgeon credentials, procedures and your pricing approach as server-rendered text. Remove contradictions across your own pages.
4. Days 22 to 30: third parties. Correct the directory, review and affiliation profiles you found, then rerun the same prompts.

Results won't be instant, and nobody can promise a citation. Retrieval-on answers can change once access and facts are fixed. Training-data answers change only when a model is retrained. Put the prompt set on a quarterly AI visibility review so you catch drift.

Does schema markup help plastic surgeons appear in ChatGPT? There's no credible evidence that it does. Google says structured data isn't required for generative AI search, and Ahrefs' matched test found AI Overview citations fell 4.6% on pages after adding schema. Existing markup does no harm, so there's no need to remove it. Spend the effort on access checks and on stating credentials and prices as plain text instead.

Should we publish exact prices if we only quote after a consultation? You don't have to publish exact prices, but you should publish your approach. One plain sentence explaining that quotes follow consultation, plus what affects the price, gives engines something accurate to repeat. If you do publish starting prices, say what they include. Without either, engines fill the gap from cost-comparison sites and forums you don't control.

Why does ChatGPT still list a surgeon who left our practice? Two likely causes. The answer may come from training data frozen before the surgeon left, which you can't change until the model is retrained. Or ChatGPT may be reading a stale third-party page, such as an old directory profile. Remove the surgeon everywhere on your own site, then correct the third-party pages that still list them.

Can we ask an AI company to correct a wrong fact about our clinic? No engine documents a direct correction process for this. What you can do is fix the sources it reads: make your own pages accurate and fetchable, and correct third-party profiles. Retrieval-based answers draw on those pages. Then retest the same prompts on each engine to see whether the wrong fact has gone.

Do before-and-after photos help AI recommend our clinic? Photos help patients, but they carry no text facts for crawlers. An engine can't read a surgeon's name or a procedure from an image alone. Add a short HTML caption next to each gallery naming the procedure and the surgeon who performed it. That caption is text an engine can quote, unlike the image.