From Zero to 37 Weekly ChatGPT Mentions: How a Korean Beauty Brand Built US AI Visibility
Background: What Zero AI Visibility Costs
A Korean skincare brand with a strong domestic following had been selling in the US for over two years. But in AI search, they were invisible. When US consumers asked ChatGPT or Perplexity "best Korean moisturizer for dry skin," this brand never appeared — not once.
The team initially dismissed this as a minor gap. Then, in late 2025, they confirmed through data that competitor brands were acquiring US customers directly from AI recommendations. The gap was real and growing.
AI visibility at the start
- • Weekly ChatGPT brand mentions: 0
- • Perplexity product recommendations: 0
- • Top competitor's weekly mentions on same queries: 23
Diagnosis: Why AI Engines Didn't Know This Brand
Analysis through AEKO found the issue wasn't product quality — it was a lack of English-language content infrastructure that AI engines could reference. AI engines prioritize brands they encounter repeatedly across independent, authoritative sources.
Spec-only product descriptions
English product pages listed ingredients and claims but weren't structured to answer the questions buyers actually ask AI.
No US media citations
The brand had no mentions in Allure, Byrdie, Into The Gloss, or other English-language beauty publications that AI engines weight heavily.
Missing structured data
No JSON-LD Product, Review, or Brand schema — so AI crawlers had no structured signal about what the products were or how they performed.
The AEO Strategy
Three parallel workstreams ran over three months.
1. Conversational product description rewrites
All English product descriptions were restructured to directly answer real buyer questions: "What makes this good for dry skin?", "Is it safe for sensitive skin?", "How does it compare to a hyaluronic acid serum?" These formats are exactly what AI engines pull from when constructing product recommendations.
2. Structured data implementation
JSON-LD Product, AggregateRating, and Brand schema were added to every product page. A new llms.txt file was written to guide AI agents through the site's structure and key product categories.
3. English-language citation building
PR and influencer efforts were focused on earning mentions in Byrdie, Refinery29, and mid-tier beauty newsletters. Repeated, independent third-party mentions are a core signal AI engines use to establish brand credibility.
Results: 3-Month AI Visibility Metrics
Weekly ChatGPT mentions
37
+37
Perplexity recommendations
14/week
New channel
AI share vs. competitors
18%
+18pp
By month four, traffic arriving from AI recommendations converted at 1.6x the organic average — because those users had already received a recommendation before clicking. The intent was already formed.
Key takeaways
AI visibility is a function of content infrastructure, not product quality.
Independent English-language media mentions are a primary trust signal for AI recommendation engines.
Structured data and conversational content directly increase the probability of being cited in AI answers.
Meaningful results typically emerge two to four months after implementation begins.
This case study is a composite based on anonymized AEKO client engagements. Supporting industry research:
See how your brand shows up in AI
Monitor how AI engines recommend your products across global markets.
Start for FreeFrequently Asked Questions
Yes. The brand in this case went from 0 weekly ChatGPT mentions to 37 in three months through structured data, conversational content, and media citation building.
In this case, results were achieved with zero ad spend. The key investments were content restructuring, JSON-LD implementation, and PR efforts. Existing SEO investments became the AEO foundation.
Optimization ran over 3 months. By month 4, AI recommendation traffic converted at 1.6x the organic average.