Coffret d'Or
vision-AI skin analysis routing to a personalized product recommendation
In 2020, the Japanese brand Coffret d'Or (Kanebo) deployed the COFFmi service on LINE, an AI skin analysis (Perfect Corp) routing to about 7,000 makeup combinations, with time on page multiplied by 2.48; the service was stopped at the end of 2024 along with the entire brand.
Key points
- Coffret d'Or (Kanebo) launched COFFmi on LINE in September 2020: AI skin analysis and makeup recommendations.
- The Perfect Corp AI analyzes hydration, sebum, spots, and skin texture, then recommends from about 7,000 combinations.
- Time on page multiplied by 2.48 and page views by 11.4 after adding the AI analysis (vendor source).
- Deployment stopped at the end of 2024: Kao discontinued the entire Coffret d'Or brand to refocus its portfolio, not because of any AI failure.
Objective
Offer personalized makeup advice around the clock to customers who cannot visit a store, whether for lack of time or proximity, and route each customer to the Coffret d'Or products suited to her skin and features.
The deployment
COFFmi is a digital beauty advice service from the Coffret d'Or brand, accessed by adding an account on the LINE messaging app. The customer takes a photo of her face and answers a short questionnaire; the COFFmi CHECKER function, built on Perfect Corp's AI Face Attribute solution, analyzes hydration, sebum, spots, and skin texture as well as facial features. From the diagnostic, the service recommends suitable products and shades from about 7,000 combinations, and allows a virtual try-on in augmented reality through the camera. Results accumulate on a personal page in the form of a journal. Coffret d'Or presents this setup as the world's first adoption of AI Face Attribute technology.
Results Proof B
Engagement metrics come from Perfect Corp's success story (the technology vendor, so an interested party): capped at B. The Japanese press (Fashionsnap, TechCrunch Japan) corroborates the deployment, the 7,000 combinations, and the skin analysis, without confirming the engagement figures, which remain vendor-sourced.
How it works
Documented architectureThe stack in detail
- plateforme AI Face Attribute Analysis of facial features and skin indicators, supplied by Perfect Corp
- outil YouCam Makeup Makeup try-on in augmented reality
- plateforme LINE Conversational channel for accessing the service in Japan
How it runs, concretely
For ops teams-
1Enter the service customer
The customer adds the COFFmi account as a friend on LINE and starts COFFmi CHECKER.
-
2Capture and analysis AI
A photo of the face and a short questionnaire; the AI measures hydration, sebum, spots, skin texture, and facial features.
-
3Product recommendation AI
The system selects products and shades from about 7,000 combinations based on the diagnostic.
-
4Virtual try-on customer
The customer tests the recommended products in augmented reality through the camera.
-
5Follow-up AI
Results accumulate on a personal page in journal format, for re-engagement over time.
The face photo sent by the customer and her answers to the questionnaire. Without a usable image (framing, lighting), the diagnostic cannot run and the recommendation falls flat.
How your customers perceive this type of use
Sourced studiesLe paradoxe est documente des deux cotes : 71% des consommateurs attendent des interactions personnalisees et 76% sont frustres quand elles manquent (McKinsey, 2021), mais 75% declarent ne pas acheter aupres d'organisations auxquelles ils ne confient pas leurs donnees (Cisco, 2024). La « creepy line » est localisee : messages recus quelques secondes apres une recherche et suivi de localisation sont les pratiques qui mettent le plus mal a l'aise (Periscope by McKinsey, 2019).
Acceptance conditions
- La confiance dans le traitement des donnees precede l'achat : 75% ne achetent pas sans elle (Cisco 2024)
- Un cadre legal protecteur rassure : 59% des consommateurs disent que des lois fortes sur la vie privee les rendent plus a l'aise pour partager des informations dans des applications IA (Cisco 2024)
- La personnalisation elle-meme est attendue quand elle est consentie : environ la moitie des consommateurs (US 55%, UK 52%) disent s'inscrire souvent ou parfois a des services personnalises (Periscope by McKinsey 2019)
Red lines
- Le message declenche quelques secondes apres une recherche ou un achat : deuxieme ou troisieme cause de malaise selon les pays (Periscope by McKinsey 2019)
- Le suivi de localisation percu comme de la surveillance : 40% de malaise en Allemagne et au Royaume-Uni (Periscope by McKinsey 2019)
- Le mesusage des donnees personnelles par l'IA, devenu la premiere inquietude des consommateurs, a 53% et en hausse (Qualtrics 2025)
Sources: McKinsey & Company 2021 · Periscope by McKinsey 2019 · Cisco 2024 · Qualtrics 2025
How to replicate
Inference, not sourcedData prerequisites
- A structured product catalog with attributes (shades, finishes, skin targets) usable by a recommendation engine
- Explicit consent to process the face photo and skin indicators (sensitive biometric data)
Org prerequisites
- A digital/CRM team able to run a conversational channel and read what comes back from it
- Legal/DPO sign-off on the processing of facial images on the EU side
Possible stack
- An off-the-shelf skin analysis and AR try-on component (such as Perfect Corp, Revieve, ModiFace)
- A conversational channel suited to the market (WhatsApp, web app, mini-app)
- A recommendation engine connected to the product catalog
First step: Map the product catalog attributes and define the skin concerns to detect, before connecting an off-the-shelf analysis component to a test channel.
The same pattern, proven in other industries
Taobao
Real-time product recommendation with deep learning, served across hundreds of scenarios in the app
Amazon
Item-to-item collaborative filtering product recommendation engine, served across every touchpoint
ASOS
catalog-scale deep learning product recommendation, plus a conversational style assistant
Sources
- S1 COFFRET D'OR - Perfect Corp Business Success Story Interested party archive pending
- S2 コフレドールが世界初、スマホの顔写真と診断結果から最適なメイクを提案するサービス開始 Established press archive pending
- S3 花王、化粧品「コフレドール」販売終了 構造改革で Established press archive pending
An error, newer info, a source?
This page lives on its accuracy. If a figure has moved, if the deployment has changed, or if you have a higher-quality source, tell us. Every sourced correction is verified before publication.