Highlights
- The dataset: This study is an analysis of over 110,000 AI source citations using prompts specific to the cosmetic surgery industry. It analyzes AI responses from four AI models run in 12 countries over the course of Q2 2026.
- Reddit makes up 13% of ChatGPT's citations: Its single most-cited source by a wide margin.
- Instagram makes up 8% of Google AI Overview's citations: More than its next nine domains combined.
- 10% average overlap between any two models: The most divergent pair, ChatGPT and Google AI Overview, shares a single top-20 domain.
- 75% of Microsoft Copilot's citations are from commercial domains: Clinic and directory sites, versus 37% for ChatGPT.
Which sources should you target to get cited as a cosmetic surgery brand?
The four models each lean on a distinctly different set of top-cited domains. Each model's five most-cited sources, with their share of that model's total citations:
| Model | Top cited sources | Citation character |
|---|---|---|
| Grok | RealSelf (4.8%), Reddit (3.8%), Yelp (3.5%), WhatClinic (2.4%), us-uk.bookimed.com (1.8%) | Reviews + community |
| ChatGPT | Reddit (12.8%), skinsage.uk (2.6%), top-rated.online (2.5%), cirugiasplasticas.info (1.9%), elpais.com (1.3%) | Community + niche + press |
| Microsoft Copilot | esteticas.com.ar (1.5%), ak.se (1.4%), Doctoralia Mexico (1.4%), whatclinic.com (1.0%), facelandclinic.com (0.9%) | Local directories + clinics |
| Google AI Overview | Instagram (8.1%), topdoctors.com.ar (1.7%), topdoctors.mx (1.5%), tiktok.com (1.4%), docteurchirurgieplastique.com (1.3%) | Social + local directories |
ChatGPT and Google AI Overview are the most concentrated at the top, one domain (Reddit, Instagram respectively) carries an outsized share. Grok and Microsoft Copilot spread their citations more thinly across many sources.
How have AI source rankings changed over time in the cosmetic surgery industry?
We split the roughly sixteen-week window at its midpoint (early May 2026) and ranked the top domains by their share of citations within each half.
Reddit and Instagram held or strengthened their top positions (Reddit #2 → #1; Instagram #5 → #2), but the broad global aggregators, Yelp, Facebook, WhatClinic (#4 → #11), gave ground to country-specific directories like Doctoralia, TopDoctors, and esteticas.com.ar. We observe this as a shift in the citation mix over the window; with a single split it is suggestive of a localizing trend rather than proof of one.
What type of content do AI models cite for cosmetic & plastic surgery clinics?
We classified every cited domain by category. Commercial domains, clinic homepages and for-profit directories, dominate the category as a whole, which is unsurprising for a service industry. But the mix differs sharply by model.
ChatGPT is the clear outlier. It is the only model that doesn't lean primarily commercial, barely over a third of its citations are commercial, and it cites far more editorial (10.2%) and reference (15.8%) content than any peer. That diversity tracks with its heavy Reddit reliance and press citations (El País, The Times). Microsoft Copilot is the mirror image: three-quarters of its citations are commercial, with almost no editorial or reference material. Grok and Google AI Overview both carry substantial user-generated content (26.9% and 13.5%), reflecting Grok's review-site habit and Google AI Overview's social-media habit.
Do AI models cite local-language content for cosmetic & plastic surgery clinics?
The country drives local sourcing almost entirely, and not in the way you might expect. Treating a citation as local when its domain ends in the market's own country code, we measured the share of citations pointing to a local source in each non-English market. (Spanish-language markets were scored against their own national code: .ar for Argentina, .mx for Mexico, .es for Spain.)
Across all non-English markets, 48.9% of citations point to a local source. But the average hides a sharp split: Germanic and Italian markets cluster around 73%, while Spanish-speaking markets and France sit near 30%. The likeliest explanation is structural, not behavioral, Germany (.de), the Netherlands (.nl), and Italy (.it) have dense, mature national directories of clinics and doctors, whereas Spanish-language and French clinics frequently publish on .com domains. This measure therefore undercounts genuinely local content in markets that favor generic domains, so treat the Spanish and French figures as a floor, not a true localization rate.
Which AI model relies most on local sources for cosmetic & plastic surgery clinics?
Holding the market constant, the models differ in how local they go. Cells marked below the n≈30 sample threshold should be read as directional only; Google AI Overview produced too few non-Spanish citations to report reliably.
Microsoft Copilot is the most aggressively local in the markets with strong national directories, 82-83% local in German, Dutch, and Swedish, consistent with its overall reliance on clinic-directory content. Grok is the most globalized, sitting at or below the pack in every language and dipping to 53% in Swedish. ChatGPT lands in between, but with the same Spanish/French dip every model shows. The pattern is consistent: the country drives localization far more than the model does, but within a country Microsoft Copilot reaches for local directories while Grok reaches for global review platforms.
Should cosmetic & plastic surgery clinics optimize for each AI model separately?
Very little overlap, so yes. We built each model's ranked list of its top-20 most-cited domains and measured how much every pair shares across all six model pairs. The average overlap is 9.8%, ranging from 2.6% to 17.6%.
Even the most-aligned pair, Grok and ChatGPT, shares only six of twenty domains, and those are the generic anchors of the category (Reddit, WhatClinic, and a handful of TopDoctors localizations). The most divergent pair, Google AI Overview and ChatGPT, shares exactly one domain (topdoctors.mx). A cosmetic-surgery clinic optimizing for one AI model is absent from roughly 90% of the domains the other models prefer.
How many sources does each AI model cite per answer?
The models differ in how many sources they cite per answer. Grok is by far the most source-hungry; ChatGPT the most economical.
Grok cites roughly 5x as many sources per response as ChatGPT. For a clinic, that cuts two ways: Grok offers far more citation slots to compete for, but each individual citation carries proportionally less weight; ChatGPT cites few sources, so earning one of its five is both harder and more valuable. (These averages count only responses that cited at least one source. Many ChatGPT responses cited none, which is why its cited-response count is well below its total response count.)
Context
This analysis draws from Temso's AI visibility monitoring platform, which tracks how brands appear in AI model responses across ChatGPT, Microsoft Copilot, xAI's Grok, and Google AI Overview. The dataset covers cosmetic and plastic surgery prompts across 12 markets (Argentina, Australia, Canada, France, Germany, Italy, Mexico, the Netherlands, Spain, Sweden, the United Kingdom, and the United States) and seven languages, between March 10 and June 29, 2026.
The findings describe which sources AI models cite, not which clinics or treatments are best. A high citation share means a domain is frequently surfaced by a model, not that it is authoritative or accurate. Citation behavior is a moving target: models update their retrieval and ranking continuously, so a snapshot like this captures a period, not a permanent state.
Methodology
How we measured this
We analyzed every AI response and its cited sources for the cosmetic and plastic surgery category over the observation window: 10,724 responses and over 110,000 cited source citations. A "citation" is a web source a model explicitly surfaced in support of an answer; we count only sources flagged as cited. For each model we ranked its most-cited domains, measured how much any two models' top-20 lists overlap, classified each cited domain by content type from a global registry, tracked how domain rankings moved between the first and second halves of the window, and counted how many sources each model cited per answer.
Localization is measured by a country-code-domain proxy: we treat a citation as local when its domain ends in the market's own country code (e.g. .de for Germany; .ar/.mx/.es for the three Spanish markets), so those figures are directional and exclude English-language prompts, which sit overwhelmingly on generic domains. This proxy undercounts local content published on generic .com domains, so Spanish and French rates are conservative floors. Several model × language cells fall below a adequate sample size (n < 30) and are treated as directional only. Model coverage varied across the observation window, so temporal movements are reported as within-half citation shares and the movement findings are suggestive rather than conclusive.
Frequently asked questions
Do different AI models cite the same sources for cosmetic surgery questions?
No, barely. Any two models share only about 9.8% of their top-20 most-cited domains on average, so roughly 90% of the sources one model relies on are absent from another's top tier. There is no single set of "AI-trusted" sources in this industry; there are four distinct ones.
Which sources does each model favor for cosmetic surgery?
Grok favors review and community platforms (RealSelf, Reddit, Yelp). ChatGPT leans on Reddit plus a wide mix of niche clinic sites, reference material, and press. Microsoft Copilot funnels into local clinic directories (esteticas.com.ar, Doctoralia, TopDoctors). Google AI Overview is the most social, with Instagram its single most-cited domain at 8.1%.
Does local language matter for getting a clinic cited?
Yes, but the market matters more than the model. In Germany, the Netherlands, and Italy, roughly 73% of citations go to local-domain sources; in Spain, France, Mexico, and Argentina, only about 27-34% do. Part of that gap is that Spanish and French clinics often publish on .com domains, which our method counts as non-local, so those figures are conservative.
Which AI model is most likely to cite a local clinic's website?
Microsoft Copilot. Three-quarters of its citations are commercial (clinic and directory) content, and it reaches 82-83% local-domain rates in German, Dutch, and Swedish markets. If your priority is being cited as a local clinic, Microsoft Copilot is the most receptive system.
Which model cites the most sources, and does that help a clinic?
Grok, by a wide margin, about 28.6 cited sources per response, roughly five times ChatGPT's 5.8. More sources mean more slots to compete for in Grok, but each carries less weight. ChatGPT's economy of citations makes each of its few slots harder to win and more valuable.
Is being on Reddit or Instagram worth it for a clinic?
For some models, materially so. Reddit is ChatGPT's single most-cited domain (12.8% of its citations) and ranks highly for Grok; Instagram alone is 8.1% of Google AI Overview's citations. User-generated and social platforms are a real path to AI visibility, but only for the models that lean on them.
Did the sources AI models cite change over the period?
They shifted. Between the first and second halves of the window, large global aggregators like Yelp and Facebook lost citation share while country-specific directories, Doctoralia Mexico, esteticas.com.ar, TopDoctors Mexico, gained. We read this as a suggestive trend toward localization, not a proven one, given the single time split.
What should a cosmetic surgery clinic do with this?
Optimize per model, not in general. There is no universal AI-visibility strategy here: presence on local national directories helps with Microsoft Copilot, review and community platforms help with Grok, social media helps with Google AI Overview, and a diverse footprint spanning community discussion, reference, and press helps with ChatGPT. Treat each model as a separate channel with its own preferred sources.
