A June 2026 Criteo Shopper Survey found that 38% of shoppers already use AI assistants when shopping for beauty and personal care products, and 57% of those shoppers said the products recommended in AI responses influence what they ultimately buy. For K-beauty brands competing for U.S. shelf space, that raises the question: which brands are the LLMs actually recommending to American consumers, and why? Meltwater used GenAI Lens, its AI Visibility Tracking capability, to find out.
Table of Contents
Methodology
Overall AI Visibility Across K-Beauty Brands
Each AI Engine Tells a Different Brand Story
Citation Sources Across K-Beauty AI Responses
AI Visibility Shifts by Consumer Priority
From Insight to Intelligence: How GenAI Lens Helps
Conclusion
FAQ: AI Visibility in the K-Beauty Category
Methodology
| Capability | Meltwater GenAI Lens |
| Analysis window | 1 July – 14 July 2026 (2 weeks) |
| Market | United States |
| Data sources | Claude, DeepSeek, ChatGPT, xAI Grok, Meta Llama, Google AI Mode, Google AI Overviews, Perplexity |
| Prompts | General: "What are the best Korean skincare brands and products available in the U.S.?" Accessibility: "Can you recommend a Korean skincare routine using products that are easy to find in the U.S.?" Anti-aging: "What are the best Korean skincare products for anti-aging available in the U.S.?" Sensitive skin: "What are the best Korean skincare products for sensitive skin available in the U.S.?" |
| Key term definitions | Visibility Score: a combined 0–99 measure of an entity's overall AI presence, blending Prevalence, Frequency and Prominence. Prevalence: the % of mentions that also reference related products, people, or entities. Prominence: how central an entity is within the response it appears in. Frequency: average citations per response. Mention Rate: the share of responses in which the entity appears. Sentiment: a 0–100 score, primarily entity-level. |
Overall AI Visibility Across K-Beauty Brands
COSRX leads most performance metrics tracked across the four core prompts. It ranked highest in AI visibility with a visibility score of 84, sitting 17 points aboveLaneige (67) and 31 points clear of Beauty of Joseon (53).
COSRX’s visibility was primarily driven by volume, recording the highest mention rate and frequency among all brands at 69 and 2.32, respectively. Its mention rate was 21 points higher than its closest competitor, Laneige (48), while its mention frequency exceeded Laneige’s by 1.05 points. This strong presence across AI-generated answers ultimately reinforced COSRX’s overall visibility.
Prominence provided a different way of boosting visibility from mentions and frequency. Sulwhasoo ranked seventh in mentions (26), but second on prominence (73), just behind COSRX, which boosted its visibility ranking. This strong prominence lifted its overall visibility ranking, showing that while Sulwhasoo appeared less frequently in AI answers, it tended to be positioned more prominently when mentioned.
Unlike mentions, frequency, and prominence, sentiment did not seem to be an AI visibility differentiator for the k-beauty brands, with every top-10 brand scoring within a tight 50–54 range.
Why it matters: There is more than one route to AI visibility. Brands should identify which levers - volume, frequency, prominence or sentiment - actually create separation in their category.
Each AI Engine Tells a Different Brand Story
Across the 8 generative AI engines surveyed, brand visibility swings from 0 to 91 for the same brand depending on which AI engine is asked. Consensus across the eight models holds only for COSRX, the category leader, while every other brand's ranking reshuffles significantly from model to model.
DeepSeek surfaced the broadest range of brands with consistently strong visibility. It scored all 10 brands 45 or higher, ranking as the top-scoring platform for eight of them. While other AI engines sharply differentiated between brands, DeepSeek gave both category leaders and smaller challengers meaningful visibility, suggesting a broader set of k-beauty brands can enter the consideration set on DeepSeek.
Google AI Mode and Google AI Overviews were the hardest platforms for brands to appear on. Six of the ten brands scored low on both, including Dr.Jart+, Missha, Dear Klairs and Purito, which scored 0. Even category leader COSRX posted its lowest scores on Google AI Mode (42) and AI Overviews (41), suggesting that visibility on Google cannot be assumed from strong performance on other AI platforms.
Claude consistently gave lower visibility scores to all k-beauty brands. COSRX was the only brand to score above 50, while every other brand remained below that threshold. Brands that performed strongly on other platforms therefore appeared much less visible on Claude.
Why it matters: A strong position on one engine does not translate to visibility on another, meaning brands cannot treat “AI visibility” as a single metric. Brands can lead on one platform and not be visible on another, making platform-by-platform visibility critical to understanding their true position in AI discovery.
Citation Sources Across K-Beauty AI Responses
Across the top 10 cited sources, social platforms account for the highest individual citation counts, with beauty publications and retailers also featuring prominently.
Social and community platforms are among the most-cited sources. Instagram leads with 217 citations, followed by YouTube at 189, while Reddit also ranks highly at 149. Their prominence shows that the content and conversations happening across these platforms form a significant part of the information AI draws on for K-beauty.
Beauty publications remain among the most frequently cited sources. Glamour (168 citations), Allure (155), and Byrdie (90) all place in the top 10, highlighting the importance established editorial sources continue to play in generative AI engine answers.
Beauty retail also ranks alongside social and editorial sources. Ulta (148 citations) and Sephora (107 citations) both rank in the top 10, showing the role multi-brand beauty retailers play in the information AI platforms draw on. For K-beauty brands, this reinforces the value of strong retail partnerships—not only for distribution and consumer discovery, but as part of the wider information ecosystem shaping AI responses.
Why this matters: AI engines draw on the full beauty ecosystem - from social conversations and editorial coverage to retailer content. Building AI visibility requires brands to show up consistently across the places consumers discover, discuss, and shop for beauty.
AI Visibility Shifts by Consumer Priority
The brands leading AI visibility change depending on the consumer priority and prompt they ask the generative AI Engines. General K-beauty queries establish COSRX, Laneige and Beauty of Joseon as the leading brands, but that hierarchy changes as consumers introduce more specific priorities.
Anti-aging produces a different category leader, with Sulwhasoo overtaking COSRX on both Visibility (73 vs. 70) and Prominence (81 vs. 68), despite not leading general K-beauty queries. This shows that broad category visibility does not necessarily translate into leadership around a specific skin concern.
Sensitive-skin queries bring specialist brands into view. COSRX remains firmly in first place at 75 Visibility, but brands like Pyunkang Yul and Purito all enter the top 10 despite not appearing in the k-beauty rankings. More specific skincare needs can therefore surface brands that are less prominent in broader K-beauty discovery.
Accessibility more closely mirrors the general K-beauty hierarchy. COSRX (84) and Laneige (75) retain the top positions, suggesting their broader visibility carries through when ease of access becomes part of the consideration.
Why it matters: AI visibility changes with what consumers are looking for. Broad category strength matters, but specific priorities can elevate different brands and reshape the competitive set.
From Insight to Intelligence: How GenAI Lens Helps
Use Case 1: Cross-Model Visibility Auditing
Track how your brand and named competitors score across every major LLM, not just one, so you're never blind to a platform where your visibility has collapsed to zero.
Use it to:
- Monitor Visibility Score, Prominence, and Mention Rate for your brand across Claude, DeepSeek, ChatGPT, Grok, Llama, Google AI Mode, Google AI Overviews, and Perplexity in one place
- Flag any platform where your brand scores near zero before a competitor or analyst notices it first
- Benchmark your brand against named category leaders and challengers on the exact same prompts
The result: Brands stop mistaking one AI engine's results for the full picture and start managing AI visibility as a multi-platform discipline.
Use Case 2: Source Ecosystem Mapping
Identify exactly which sources — social, retail, or earned media — are driving citations for your brand and your category, model by model.
Use it to:
- Pinpoint which platforms (Instagram, Reddit, Ulta, Sephora, Glamour) are feeding which LLMs in your category
- Prioritise content and partnership investment toward the specific source-model pairings driving the most citations
- Track shifts in source mix over time as retail and social content continues to displace traditional press coverage
The result: Marketing and PR spend gets redirected toward the channels demonstrably feeding AI recommendations, not just the ones that feel familiar from search-era SEO.
Use Case 3: Prompt-Level and Category Tracking
See how visibility shifts across specific use-case prompts — general, accessibility, anti-aging, sensitive skin — rather than relying on a single, blended brand score.
Use it to:
- Reveal which specific prompts your brand under-performs on, even while leading overall
- Identify open categories where no brand has established dominance yet
- Monitor product-level Visibility and Mention Rate to see whether any single SKU is carrying disproportionate weight
The result: Brands can target investment at the exact use-case queries where a challenger is winning, rather than defending an overall score that may be masking a real gap.
Conclusion
K-beauty’s AI visibility landscape is far from settled. COSRX has emerged as the strongest overall brand, but leadership shifts depending on both the AI platform and what consumers are looking for. Sulwhasoo overtakes COSRX for anti-aging, while individual models surface markedly different competitive sets.
Beyond the overall leader, there is significant room for movement. Specialist brands emerge around specific consumer priorities, visibility varies widely between AI platforms, and no single product or channel defines a brand’s presence. For K-beauty brands, building AI visibility means establishing relevance across the different platforms, consumer needs, and sources shaping AI discovery.
FAQ: AI Visibility in the K-Beauty Category
Which K-beauty brand has the highest AI visibility in the U.S.?
COSRX leads with a Visibility Score of 84, according to Meltwater's GenAI Lens analysis — 17 points ahead of Laneige (67) and 31 ahead of Beauty of Joseon (53), based on two weeks of data across eight LLMs in July 2026.
Does one AI model represent the whole category?
No. Meltwater's GenAI Lens analysis found the same brand's Visibility Score can swing from 0 to 91 depending on the model, with DeepSeek scoring most brands highest and Google AI Mode/Overviews scoring several at zero.
What drives AI citations for K-beauty brands?
Social platforms account for the highest individual citation counts among the top 10 sources tracked, according to Meltwater's GenAI Lens data, with Instagram (217) and YouTube (189) ranking at the top two sources. Beauty publications (Glamour, Allure, Byrdie) and multi-brand retailers (Ulta, Sephora) also feature prominently, but rank behind the top social sources.
Is COSRX’s lead the same across every type of skincare query?
No. Meltwater's GenAI Lens analysis shows COSRX loses the anti-aging prompt outright to Sulwhasoo (73 vs. 70 Visibility), showing that category leadership is prompt-specific rather than absolute.
Does sentiment explain why some K-beauty brands outperform others in AI?
According to Meltwater's GenAI Lens analysis, every top-10 k-beauty brand scores within a tight 50–54 sentiment band regardless of visibility, meaning frequency and prominence of mention are the real differentiators, not favourability.
Do individual products drive K-beauty AI visibility, or is it brand-level?
Meltwater's GenAI Lens analysis found even the top-scoring product (COSRX’s Advanced Snail 96 Mucin Power Essence) reaches only a 20% Mention Rate, showing LLMs favour brand recognition over any single "viral" product.
