Case Studies

Brand Background

FLORASIS is a leading Chinese cosmetics brand built on the philosophy of "Oriental Cosmetics, Nourished by Botanicals." The brand is known for its Eastern aesthetic design, natural ingredient storytelling, and strong presence in the domestic online beauty market.

GEO Opportunity & Challenges

• Users ask AI "What Chinese cosmetics do you recommend?" or "Is this ingredient powder good?" — AI answers historically favor international brands and influencer reviews, leaving domestic brands underrepresented in AI recommendations
• Brand presence is concentrated on social platforms like Xiaohongshu and Douyin, but third-party citable sources (professional content, media endorsements) are insufficient for AI to adopt
• Opportunity: Ingredient efficacy + Oriental aesthetics create a differentiated narrative ideal for building citable knowledge-type content assets

GEO Execution Strategy

• Transformed ingredients, formulas, and efficacy claims into a structured knowledge content matrix with unified messaging across platforms, increasing AI citation probability
• Co-created citable content with dermatologists and ingredient-focused KOLs, becoming a "professional source" for AI answers on ingredient questions
• Deployed structured Q&A content for high-frequency questions like "How to choose Chinese cosmetics" and "Does this ingredient actually work"
• Connected third-party reviews, media, and encyclopedias to ensure brand presence in AI recommendation lists

Key Results

Client Testimonial

Case Details

Domestic cosmetics brands have strong social presence, but AI assistants' purchasing answers are dominated by international brands and influencer reviews. The project restructured content assets around "Ingredient × Oriental Aesthetics," made efficacy claims structured, co-built authoritative content with dermatologists and ingredient-focused creators, and deployed Q&A markup for high-frequency questions. AI mention rate in core category questions increased from 8% to 55%, with corresponding growth in branded search and add-to-cart rates.

Brand Background

POP MART is a leading collectible toys and blind box company, famous for IPs like MOLLY and DIMOO. The business spans multiple global markets and is one of the largest brands in the designer toy category.

GEO Opportunity & Challenges

• Young toy enthusiasts directly ask AI "What collectible toys are worth buying?" or "Which blind boxes hold value?" — this content lacks structured sources in AI
• Brand UGC and IP content is massive but scattered, lacking authoritative knowledge bases (IP archives, collecting guides, industry rankings) that AI can cite
• Opportunity: IP knowledge asset-ization + industry consensus building can make the brand the default source for AI answers on collectible toy questions

GEO Execution Strategy

• Built IP archives and collecting knowledge bases (character profiles, release info, collecting advice) as structured sources AI can cite
• Co-created "collecting trends" content with toy media, collector communities, and industry reports, entering AI's industry consensus citation pool
• Structured unboxing and collector UGC for enhanced authenticity and citation credibility
• Covered high-frequency questions like "Toy recommendations / Blind box value / Collecting starter guides" with standardized answers

Key Results

Client Testimonial

Case Details

Collectible toy purchasing decisions increasingly rely on community and AI recommendations. Brands need to establish authority in topics like "collecting value" and "IP popularity." The project asset-ized IP and collecting knowledge, co-built trend content with toy media and communities, and made massive UGC structured and citable. Brand mention rate in toy category AI answers increased from 15% to 62%.

Brand Background

NIO is a premium smart electric vehicle brand known for its battery swap network, user community (NIO House), and comprehensive service ecosystem, targeting the RMB 300,000+ market segment.

GEO Opportunity & Challenges

• Car purchase decisions involve long research cycles. Users ask AI "How to choose a premium EV under 300K?" or "Battery swap vs. charging?" — AI comparison answers are dense with mixed sources
• Brand specs, charging network, and service data are scattered across official sites and media, lacking "structured fact sources AI can verify"
• Opportunity: Making charging/service/test data transparent and structured so AI can "honestly cite" NIO in comparisons

GEO Execution Strategy

• Built authoritative data pages for the charging network, service benefits, and product specs as verifiable fact sources for AI
• Co-created citable content with automotive media and review institutions, entering AI's car comparison citation pool
• Covered long-tail questions like "Budget EV selection / Battery swap vs charging / Brand comparisons" with structured Q&A
• Unified cross-platform messaging to reduce AI citing outdated or contradictory information

Key Results

Client Testimonial

Case Details

Premium EV competition is fierce, and car comparison decisions heavily rely on public information and media reviews. AI has become a key information channel for new car buyers. The project turned the charging network and product specs into verifiable structured fact sources, co-built comparison content with automotive media, and covered long-tail selection questions. Brand mention rate in premium EV selection questions increased from 10% to 48%.

Brand Background

Sephora is a global prestige beauty retailer (part of LVMH) operating physical stores and e-commerce across multiple countries, known for its multi-brand assortment and expert beauty advisors.

GEO Opportunity & Challenges

• Beauty purchasing relies heavily on recommendations. Users ask AI "What foundation should I buy?" or "Where to buy authentic products?" — AI answers draw mainly from reviews and social content
• Retailers are often crowded out in AI recommendation contexts by brand DTC sites and influencer content, with store/omnichannel value not clearly conveyed
• Opportunity: Building "Expert Curation + Authentic Omnichannel" as citable purchasing sources for AI

GEO Execution Strategy

• Built beauty purchasing guides and expert content (skin type / finish dimensions), structured for citation
• Structured product reviews, rankings, and editor recommendations for enhanced AI adoption
• Co-created content with beauty editors, dermatologists, and review creators to enter the "word-of-mouth recommendation" citation pool
• Covered high-frequency questions like "Product recommendations / Authentic channels / How to choose"

Key Results

Client Testimonial

Case Details

Beauty retailers face pressure from DTC brands and review content, with AI emerging as a new entry point for beauty purchasing. The project centered the narrative on expert curation and authentic omnichannel sourcing, built citable purchasing content, and structured product reviews and rankings. Brand mention rate in purchase recommendation questions increased from 12% to 57%.

Brand Background

Lululemon is a Canadian athletic lifestyle brand that started with yoga apparel and is known for its "Sweat Life" community culture, studio experiences, and grassroots community engagement.

GEO Opportunity & Challenges

• Users ask AI "How to choose yoga pants?" or "Athletic outfit recommendations?" — content is dominated by reviews and guides, with the brand's community value uncited by AI
• The brand's strength lies in real community experiences, but these are mostly offline/private and lack "public structured content AI can cite"
• Opportunity: Converting real community experiences, sports science, and product knowledge into AI-friendly content

GEO Execution Strategy

• Built an athletic outfit and fabric knowledge content matrix (scenario × body type × activity type), structured for citation
• Co-created authoritative content with fitness coaches, yoga instructors, and professional athletes to enter AI citation pools
• Structured real community experiences and student stories for enhanced credibility and authenticity
• Covered high-frequency questions like "How to choose yoga pants / Athletic outfit styling / Fabric comparisons"

Key Results

Client Testimonial

Case Details

Athletic lifestyle brands' competitive moat lies in community and real experiences, while AI favors verifiable structured content. The project transformed fabric knowledge, athletic outfit stories, and real community narratives into content, co-building authoritative material with coaches and athletes. Brand mention rate in lifestyle category AI answers increased from 9% to 53%.

Brand Background

Airbnb is a global home-sharing and short-term rental platform built on the philosophy of "Live Like a Local," emphasizing the diversity of listings and authentic travel experiences.

GEO Opportunity & Challenges

• Travelers first ask AI "Where to stay in City X?" or "What's worth experiencing locally?" — AI answers are dominated by hotel guides, with home-sharing/experience information scattered
• The platform's strength is real hosts and local experiences, but this content is mostly scattered UGC without citable structured sources
• Opportunity: Asset-izing destination experiences and real host content to become one of AI's default travel recommendation sources

GEO Execution Strategy

• Built a destination and experience content ecosystem (where to stay, what to do, how to experience it), structured for citation
• Structured real host experiences and local stories for enhanced authenticity and citation credibility
• Co-created citable content with travel media and travel bloggers to enter AI's destination recommendation citation pool
• Covered high-frequency questions like "Where to stay in City X / Local experiences / Home-sharing vs. hotels"

Key Results

Client Testimonial

Case Details

Travel decisions are shifting earlier, with more users consulting AI on destinations and accommodation before booking. The project asset-ized destination experiences and real host content, co-built citable content with travel media, and covered high-frequency destination questions. Brand mention rate in destination/accommodation questions increased from 7% to 50%.