L'Oreal (Vichy, La Roche-Posay, L'Oreal Paris)
AI diagnosis from a selfie plus routine recommendation
L'Oreal's AI skin diagnosis (SkinConsult AI, ModiFace engine) is deployed in stores in 56 countries across 7 brands, with up to 70% of consumers buying after the experience.
Key points
- AI skin diagnosis from a selfie, then product routine recommendation.
- ModiFace / SkinConsult AI engine, trained on the 6,000-image Skin Ageing Atlases dataset.
- Up to 70% purchase after the experience, deployed in stores across 56 countries via 7 brands.
- Evidence B, confirmed status: documented in L'Oreal's 2024 annual report.
Objective
Turn a selfie into a personalized skin diagnosis and product routine, to steer online and in-store purchase and lift conversion.
The deployment
SkinConsult AI analyzes a selfie to assess signs of skin ageing (wrinkles, spots, pores, firmness) through a comparative algorithm trained on L'Oreal's Skin Ageing Atlases image database, then proposes a suitable product routine. The technology, powered by ModiFace, feeds the skin diagnosis of the group's brands (Vichy, La Roche-Posay, L'Oreal Paris) on their sites, in stores, and on Tmall in China. In 2024, L'Oreal reports providing skin diagnosis and routine in stores in 56 countries across seven of its brands.
Results Proof B
Conversion figure and market coverage published in L'Oreal's official annual report (primary source), technology described by established press and by the brand. Concordant sources, one of them primary.
How it works
Documented architectureThe stack in detail
- outil SkinConsult AI diagnosis of skin ageing signs (wrinkles, spots, pores, firmness) from a selfie, with routine recommendation
- plateforme ModiFace computer vision engine (L'Oreal subsidiary) that powers the diagnosis for the group's brands
- infra Skin Ageing Atlases (base d'entrainement) proprietary L'Oreal database of 6,000 graded images used as a comparative reference for the algorithm
- infra Tmall distribution channel for the diagnosis in China, in addition to brand sites and in-store kiosks
How it runs, concretely
For ops teams-
1Selfie capture customer
The user uploads a selfie on the brand site or in store.
-
2Comparative analysis AI (ModiFace / SkinConsult AI)
The algorithm compares the face against the graded image database and scores signs of ageing.
-
3Diagnosis and routine AI
The system returns a diagnosis and a personalized product routine.
-
4Advice and purchase advisor / customer
The advisor (store) or the page (web) steers the customer toward buying the recommended products.
-
5Model maintenance data / brand team
The teams update the image database and the diagnosis-to-product mapping.
The quality and representativeness of the reference image database (Skin Ageing Atlases). A biased or too narrow training set degrades the diagnosis for certain skin phototypes.
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
- graded and diverse reference image database
- diagnosis-to-product mapping
- image consent and non-retention policy
Org prerequisites
- dermatology/data expertise for training and validation
- site integration plus training of in-store advisors
Possible stack
- ModiFace
- Perfect Corp (skin analysis)
- Haut.AI
- internal vision model
The plan, step by step
- Step 1Build or license a graded and diverse reference image database, and frame compliance (consent, non-retention of the image, non-medical nature of the diagnosis)Deliverable: Validated reference dataset + GDPR impact assessment
- Step 2Define the diagnosis-to-product and routine mapping with domain expertsDeliverable: Validated diagnosis/routine matrix
- Step 3Train the classification model or integrate a skin-analysis SaaS, then validate accuracy across varied phototypesDeliverable: Prototype with error rate measured by phototype
- Step 4Integrate the selfie, diagnosis, then routine journey on the site and in a store pilot, with advisor trainingDeliverable: Journey in production on one market + trained advisors
- Step 5Measure the post-diagnosis purchase rate and extend to other brands and marketsDeliverable: Post-diagnosis conversion dashboard + expansion plan
First step: Build or license a validated reference image database and define the diagnosis-to-product mapping before any deployment.
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 L'Oreal, the Beauty Tech champion (Annual Report 2024) Primary archive pending
- S2 L'Oreal to Offer AI-Powered Skin Diagnosis Using Selfies Established press archive pending
- S3 SkinConsult AI by Vichy Primary 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.