Gangnam Dermatology AI Citation Rate Research Report
We cross-measured how major Gangnam dermatology clinics are cited within AI answers across 29 procedure categories and 6 LLMs. Using 1,740 prompts (approximately 500 million prompt tokens), we collected 62,640 responses and analyzed citation rates and variance across models.
Chapter · 01
Summary.
No single clinic monopolizes the Gangnam dermatology AI citation market, but citations are already concentrated among a few. In this measurement, AI mentioned 14,950 clinics in total—yet the top 10 captured 12.9% of all citations, and the top 100 captured 41.3%. Meanwhile, 8,108 clinics were mentioned only once. Some clinics appeared repeatedly no matter which AI was asked. While no single player dominates the market, a clearly identifiable top tier of repeatedly cited clinics does exist.
상위 10 합산
12.9
1위-10위 격차
2.8배
측정 LLM
6개
인용 기업 수
14950
총 호출
62640
Chapter · 02
Methodology.
Engines
ChatGPT · Claude · Gemini · Perplexity · Grok · Google AI Overview (6 engines cross-measured with equal weighting)
Query set
Citation Rate The percentage of all measured responses in which a given clinic was mentioned. Because multiple clinics can be cited within a single response, the sum of all clinics' citation rates exceeds 100%. The citation rate indicates how frequently AI mentions a specific clinic. Share of Mentions The percentage of all clinic mentions generated by AI that a given clinic accounts for. The shares of all clinics sum to 100%, showing each clinic's relative weight within the total citation market. GAP The standard deviation of citation rates across the 6 LLMs. Lower values indicate similar results across models; higher values indicate larger citation differences between models.
Scope
We designed 6 question types—recommendation, ranking, price, reviews, expertise, and safety—across 29 procedures. Applying 5 age groups and 2 gender conditions produced a total of 1,740 prompts, each run 6 times across 6 LLMs. In total, we collected 62,640 responses.
Period
May 2026
Error margin
As a single point-in-time measurement, statistical confidence intervals are not provided. Response variance per model was averaged across 6 repeated measurements per condition; figures may vary over time due to AI model updates and the probabilistic nature of model responses.
The full measurement methodology is published in the GEO Citation Report methodology.
Chapter · 03
Key Findings.
01
AI mentioned many clinics, but citations were concentrated among a few
A total of 14,950 clinics appeared in this measurement, but citations were far from evenly distributed. The top 10 clinics accounted for 12.9% of all citations, the top 100 for 41.3%, while 8,108 clinics were mentioned only once. No single clinic monopolized the market, but a top tier of repeatedly cited clinics was clearly identifiable.
02
Each AI named a different No. 1 clinic for the same question
The 6 AI engines each mentioned different clinics most frequently. ChatGPT and Grok cited 오*클피부과 강남점 most often, Gemini cited 톡*앤필, Perplexity cited 미*미 강남본점, and Claude cited J*성형외과. The rate of directly naming clinics also varied from 34% to 99%, showing that each AI has distinct citation patterns and recommendation structures.
03
Citation gaps between models varied widely by procedure.
While some procedures showed consistent results across models, such as hair transplant (GAP 8.1), the top clinic lineup varied widely by model for scar removal (31.8) and fat-dissolving injection (30.2). For procedures with higher GAP scores, no single clinic held a clear advantage—citation results shifted depending on question intent and model.
Chapter · 04
Top 10 Rankings.
01
오*클피부과 강남점
7.2%
02
톡*앤필
5.8%
03
오*클피부과
5.7%
04
톡*앤필 강남본점
5.5%
05
유*아이의원
3.7%
06
뮤*의원
3.2%
07
미*미 강남본점
3.2%
08
아*디병원
3.1%
09
리*리의원 강남점
2.8%
10
원*성형외과
2.6%
Measured 2026-05 · 62.6K calls combined · Some brand names masked
Chapter · 05
By Engine.
Even for the same question, each LLM cites different clinics. Comparing the top 3 most-cited clinics per engine reveals a clear distinction between clinics strong on only one model and those cited consistently across all engines.
ChatGPT
Claude
Gemini
Perplexity
Grok
Google AI Overview
Chapter · 06
Interpretation.
AI search does not deliver a single definitive answer. Each model mentioned different clinics for the same question, and the No. 1 clinic differed across all 6 AI engines. The rate of directly naming clinics also ranged from 34% to 99% by model. What matters, therefore, is not a single search ranking. How consistently a clinic is cited across multiple AI engines—and how broadly those citations are distributed—becomes the new standard of visibility.
“Search has not disappeared. The first touchpoint where users encounter information is simply shifting from search results to AI answers.”
Citations were concentrated among a handful of clinics, while the vast majority were widely dispersed. The fact that 8,108 of 14,950 clinics were mentioned only once shows that most clinics have yet to secure meaningful visibility within AI. At the same time, no position at the top is locked in yet. No single clinic monopolized the market, and even the top clinics' individual citation shares remained in single digits. Ultimately, the AI citation race varies by model, question intent, and target segment. Clinics mentioned repeatedly across multiple models showed higher visibility, and clinics that steadily accumulated diverse touchpoints—web content, press coverage, and brand activity—were cited relatively more often. Where a single ranking determined exposure in the search era, in the AI citation era, visibility is determined by repeated mentions across multiple models and the accumulation of signals.
Need measurement built around your brand?
We scope the measurement by industry, region, and brand, and deliver it as a custom GEO measurement report.