The first bilingual AI visibility audit of Dubai's plastic surgery and aesthetic medicine market — measuring how ChatGPT, Claude, Gemini, Grok, Perplexity and Google AI Overviews respond to 100 patient queries across English and Russian.
Dubai's aesthetic medicine market generates millions of AI-assisted patient queries every month. This study measured what AI actually says — and whose name appears when it says it.
For clinic leadership: In Dubai's real estate market, Emaar achieves 79% AI visibility. Jumeirah reaches 89% in hospitality. The best result in aesthetic medicine is 19.6%. No clinic dominates this market in AI yet — the position is still available. The clinic that moves first builds a structural advantage competitors cannot replicate by increasing ad spend.
LET'S TALK →No single brand dominates English-language AI. The market leader appears in fewer than 1 in 5 responses. Most clinics appear in fewer than 1 in 10.
Seline leads Russian AI at 16.8% average — but a 44-point gap between its best platform (Perplexity) and worst (Gemini) reveals a brand with no stable AI identity.
English and Russian AI return different market leaders, different clinic hierarchies, different citation sources. Same patient, same procedure — different world depending on the language they search in.
The most-cited source in Russian AI is EDEN Aesthetics' own website — not a review aggregator. One structured brand site outperforms every third-party in the market.
The DHA licensing registry is cited more than any clinic or surgeon website. AI uses regulatory authority as a trust proxy — brands that appear in it inherit that trust.
Nine brands are completely invisible in English AI. Twelve in Russian. Several have strong Instagram presence and active operations — invisible to patients making booking decisions via AI.
Mention Rate = % of 50 queries where the brand appeared in platform response. AVG = cross-platform average. POS = average line of first mention. CON = platforms where brand appeared out of 5.
| Brand | ChatGPT | Claude | Gemini | Grok | Perplexity | AVG | POS | CON |
|---|---|---|---|---|---|---|---|---|
| CLINICS — TIER 1 & 2 | ||||||||
| CosmeSurge | 19.6% | 10.7 | 5/5 | |||||
| American Hospital Dubai | 17.6% | 10.7 | 5/5 | |||||
| Lucia Clinic | 14.8% | 11.4 | 5/5 | |||||
| Kings College Hospital | 11.2% | 12.0 | 5/5 | |||||
| Cocoona | 10.8% | 6.0 | 5/5 | |||||
| Aesthetics International | 10.0% | 9.1 | 5/5 | |||||
| Novomed | 9.2% | 13.4 | 4/5 | |||||
| Dubai Cosmetic Surgery | 8.4% | 10.3 | 5/5 | |||||
| Perfect Doctors | 7.2% | 10.3 | 3/5 | |||||
| Elite Plastic | 6.8% | 9.3 | 5/5 | |||||
| AACSH | 6.4% | 5.1 | 5/5 | |||||
| Euromed | 5.6% | 14.6 | 5/5 | |||||
| Bizrahmed | 4.0% | 13.3 | 5/5 | |||||
| Enfield Royal | 2.8% | 13.6 | 4/5 | |||||
| EDEN Aesthetics | 2.0% | 8.2 | 3/5 | |||||
| PERSONAL BRANDS — SURGEONS & AESTHETIC DOCTORS | ||||||||
| Dr Rory McGoldrick | 10.4% | 7.8 | 3/5 | |||||
| Dr Matteo Vigo | 9.2% | 10.4 | 5/5 | |||||
| Dr Sanjay Parashar | 6.0% | 4.3 | 5/5 | |||||
| Dr Jaffer Khan | 6.0% | 5.1 | 4/5 | |||||
| Dr Allen Rezai | 4.4% | 10.2 | 5/5 | |||||
| Dr Radmila Lukian | 2.4% | 7.3 | 4/5 | |||||
|
You just found your clinic in this table. If you want to understand which queries you're losing — and what it takes to change that — let's look at your numbers together. Coffee, no deck, no proposal. LET'S TALK → | ||||||||
Mention Rate measures frequency of brand name detection in full AI response text, not sentiment. A mention in a neutral list counts the same as a recommendation. Brands with 0% across all platforms omitted from table above.
Clinics average 7.8% visibility. Surgeons average 4.4%. Dr Sanjay Parashar has 2 million Instagram followers — and 6% average AI visibility.
Google AIO cites instagram.com 18 times in English — yet the most Instagram-native brands in this study are precisely the ones with the lowest LLM visibility. AIO reads Instagram as a source; LLMs do not. A clinic that built its entire presence around Instagram content has a signal that reaches one AI channel and is invisible to the other four. That is the structural risk of single-channel content strategy in the AI era.
Every query where your clinic is absent is a patient conversation that starts with a competitor's name. AI-referred patient sessions convert at multiples of organic search traffic — the patient who arrives from an AI recommendation has already received a personalised endorsement. The commercial consequence is direct and measurable.
Same 40 brands. Same 5 platforms. 50 Russian-language queries. The results reveal a market that AI describes differently in almost every dimension.
| Brand | ChatGPT | Claude | Gemini | Grok | Perplexity | AVG | POS | CON |
|---|---|---|---|---|---|---|---|---|
| CLINICS | ||||||||
| Seline | 16.8% | 7.5 | 4/5 | |||||
| Bizrahmed | 12.4% | 13.3 | 4/5 | |||||
| CosmeSurge | 12.0% | 7.3 | 4/5 | |||||
| EDEN Aesthetics | 12.0% | 8.4 | 4/5 | |||||
| Dubai Cosmetic Surgery | 12.0% | 8.5 | 5/5 | |||||
| Lucia Clinic | 10.8% | 7.4 | 4/5 | |||||
| Cocoona | 8.0% | 8.0 | 3/5 | |||||
| American Hospital Dubai | 7.2% | 8.1 | 4/5 | |||||
| Euromed | 5.2% | 11.2 | 1/5 | |||||
| PERSONAL BRANDS (RU) | ||||||||
| Dr Rory McGoldrick | 5.6% | 7.9 | 2/5 | |||||
| Dr Sanjay Parashar | 2.4% | 4.8 | 2/5 | |||||
RU clinics average 6.6%. RU surgeons average 1.6%. Russian-language AI almost never recommends surgeons by name — it describes clinics.
A surgeon without a strong clinic association is effectively invisible to the Russian-speaking patient searching via AI. Your existing patients find you through word of mouth. The patients you haven't met yet are asking AI — and if your name isn't in the response, they book someone who is.
An English-speaking patient and a Russian-speaking patient asking the same question about plastic surgery in Dubai receive fundamentally different answers — different brand hierarchies, different clinic names, different authority signals.
Seline is #1 in Russian AI. It ranks 19th in English AI at 4.8%.
A clinic reading only its English-language data would conclude it has a modest AI presence. Its Russian-speaking patients receive it as the top recommendation. The inverse holds: American Hospital Dubai ranks 2nd in English and 8th in Russian. Two clinics, four very different AI realities — none visible without a bilingual measurement approach.
Bizrahmed scores 4.0% in English LLM. Claude alone gives it 36% in Russian.
Same brand, same platform — a 28-point gap based solely on the language of the question. The patient asking in Russian and the patient asking in English receive different answers from Claude on the same day. This is the most extreme language-platform intersection in the study.
EDEN's website is the single most-cited source in all Russian-language AI.
edenderma.com appears 107 times across Russian LLM platforms — more than Bookimed, Medigence, and every aggregator in that tier combined. A clinic that has, apparently without strategic intent, built near-perfect Russian-language AI visibility through web content structure. This is what GEO looks like in the wild — and it happened without a GEO strategy. Imagine what intentional optimisation produces.
No brand in this study has a stable AI identity across platforms. The divergence gap is the most actionable number in this report — it reveals where visibility exists, where it doesn't, and why.
The #1 Russian AI brand exists almost entirely on two platforms. A Russian patient using Gemini or Claude will never encounter Seline.
Google AIO cites Dr McGoldrick 26%. That signal doesn't transfer to ChatGPT or Gemini. Strong SEO architecture does not equal AI visibility.
Russian Claude cites Bizrahmed more than any other brand. Russian Gemini never mentions it. This is a content structure problem, not a brand problem.
Euromed exists in Russian AI only on Gemini. Consensus score 1/5 — maximally fragile to any algorithm update.
AI platforms don't form opinions — they cite sources. The sources they cite are the brands that appear in their answers. These are the domains controlling the AI narrative in Dubai's aesthetic medicine market right now.
✓ brand-owned · grey = aggregator · blue = regulatory
dha.gov.ae is cited 105× in English and 84× in Russian — more than any clinic website in either language.
This is unique to the medical segment. AI systems use regulatory authority as a credibility anchor in healthcare. A clinic's DHA licence page, consistently linked from external sources, becomes an indirect AI citation asset that no amount of blog content can replicate.
edenderma.com is cited 107× in Russian LLMs — more than Bookimed and Medigence combined in that tier.
EDEN Aesthetics' site is structured in a way that makes it exceptionally readable by AI — detailed procedure descriptions, structured FAQ content, clear entity signals connecting clinic to procedures to location. AI cites them as a primary source. This is what effective GEO looks like — and it happened without a GEO strategy. Imagine what intentional optimisation produces.
Instagram tops AIO in English — yet the most Instagram-native brands are invisible in LLMs.
Bookimed controls the Russian patient's first AI encounter with this market.
Yandex favours medical tourism aggregators — a different ecosystem from Google entirely.
Yandex Maps local signals matter — reads its own ecosystem preferentially for Russian queries.
AIO Delta = brand's Google AI Overview presence minus its LLM average. Positive: over-performs on Google relative to conversational AI. Negative: in LLM conversations but absent from Google's AI layer.
Kings College Hospital: the largest AIO Delta in the study at +32.8% in English.
Google AIO cites them in 44% of relevant queries while LLMs average 11.2%. Brand authority exists. Content exists. The citation architecture that transfers that authority into conversational AI hasn't been built yet. This is a precise, addressable GEO optimisation gap — and it applies to several brands in this study.
These are not obscure brands. Several have active operations, professional websites, and consistent social media presence. Zero current AI visibility reflects a content and citation architecture that AI systems cannot yet read — not a lack of reputation or clinical quality. That gap is closable.
Surgeons average 4.4% AI visibility in English and 1.6% in Russian. 9 of 18 surgeons tracked have zero visibility in at least one language.
A clinic has a business entity, an address, a DHA licence, a website with multiple service pages, and citation history across directories. A surgeon has a name — often inconsistently spelled across sources — and a website that focuses on personality rather than structured procedure content. AI cannot disambiguate "Dr Kayle" from "Dr Dany Kayle" from "Dr D. Kayle" — that fragmentation destroys citation count.
The mechanism is concrete. drrorymcgoldrick.com is cited 44 times across LLM platforms because each procedure page carries schema markup, a linked DHA reference, and consistent entity signals connecting "Dr Rory McGoldrick" to "rhinoplasty" to "Dubai" across every citation point. That is why Grok mentions him in 38% of relevant queries. It is not reputation. It is architecture. A surgeon with equivalent reputation but without that architecture scores 0%.
If you're a surgeon and want to see where your name appears — or doesn't — across AI platforms, we'd be glad to show you over a coffee. No commitment. Just your data explained in plain terms. info@barsagency.ae
Only one query across all 100 prompts returned zero brand mentions across all five platforms simultaneously: "медицинский туризм пластика Дубай" (medical tourism plastic surgery Dubai). In English, every query triggered at least one brand mention somewhere. This market has significantly higher baseline AI brand density than real estate or hospitality — but the distribution is extremely uneven.
Your clinic or your name may be in one of those lists. Zero is a starting position, not a verdict — and the path from zero to visible is shorter than most clinics expect. Let's look at your specific situation over coffee.
LET'S TALK →Brand density per query = how many distinct brands appear across all five platforms for one patient question. High-density queries are where competition for AI presence is most direct — and where one well-optimised content piece can displace multiple competitors simultaneously.
"Best cosmetic surgeon Dubai" — 15 brands competing for one patient. These are the highest-volume queries and the ones where AI has the most room to include or exclude any given clinic.
A brand appearing consistently at position 3 across all 5 platforms on these queries receives more patient exposure than a brand ranking first on a low-density procedural query. GEO strategy for this market starts here. The conversation that shows you exactly where your clinic stands on each of these queries starts with a coffee.
Every finding in this report points to the same conclusion: AI visibility in Dubai's aesthetic medicine market is not correlated with clinic size, marketing budget, or Instagram following. It is the result of how a brand's content is structured, distributed, and cited across the web.
AI systems recognise brands as entities, not keywords. Consistent name formats, linked data, and schema markup across all web presences build the entity signal that determines citation frequency.
English and Russian AI are separate ecosystems. A clinic optimising only in English is invisible to a major patient segment. EDEN's Russian-language performance is the clearest proof of concept in this dataset.
dha.gov.ae is the most-cited source in English-language AI. Clinics that ensure their DHA profiles are complete and consistently linked across external sources gain trust-authority that content alone cannot create.
A brand visible on 1–2 platforms is structurally fragile. Cross-platform GEO builds consensus scores above 4/5 — the threshold that correlates with stable, predictable AI presence across algorithm updates.
Bookimed, Whatclinic, Doctify, and RealSelf are cited more than most clinic websites. Consistent, complete brand presence on these platforms is the primary AI citation channel — not optional.
High-density queries reward clinics with structured, AI-readable procedure pages. Generic service listings rank for nothing. Specific, entity-rich procedure content becomes a primary AI source.
A patient asking "best rhinoplasty surgeon Dubai" in English and a patient asking "лучший хирург ринопластика Дубай" in Russian receive different clinic names from the same AI platform on the same day. The AI is reflecting the content architecture of indexed brands — not clinical quality or reputation. A patient's AI-mediated perception of this market depends on the language they speak. This is a structural characteristic of the current AI ecosystem that clinics, regulators, and patients should be aware of.
Whether you're a clinic CMO who wants to understand exactly which queries you're losing, a CEO asking whether this matters for your growth plan, or a surgeon wondering if your name even appears — we can show you your specific numbers over coffee.
No pitch. No deck. No proposal. Just data and an honest conversation about what it means for you specifically.
50 English + 50 Russian queries run through 5 LLM platforms and Google AIO + Yandex. Full response text and citation URLs captured. Date: July 16, 2026.
Case-insensitive substring matching across all aliases per brand. Name variant fragmentation may cause marginal undercounting for brands with inconsistent naming conventions.
ChatGPT, Claude, Gemini, Grok, Perplexity, Google AI Overviews, Yandex AI. Grok citations: 45/50 EN, 44/50 RU — minor gap, documented.
Mention Rate, Platform Consensus (0–5), Position Score (line of first mention), AIO Delta (AIO% minus LLM avg%), Citation Domain Count.