Reinventing the Beauty Experience with AI

### TL;DR Sephora leverages AI (especially through its partnership with OpenAI) to build a personalized shopping assistant on its website and in ChatGPT, boosting conversion rates by over 5%. It shares three key takeaways: guided experiences, transparent recommendations, and continuous testing and learning. --- ## Sephora’s AI Strategy: Data-Driven Omnichannel Experiences With more than 3,000 stores across 37 countries, 80 million active loyalty members in its customer database, and 2.5 billion annual visits to its website and mobile app, Sephora places AI and technology at the heart of its strategy. The goal is to create more personalized omnichannel experiences: delivering precise recommendations on the website, providing data and insights to brand partners, and equipping employees with AI skills. AI supports every link of Sephora’s value chain: - **Digital Team**: Optimizes the website through personalized recommendations, tailoring the experience based on customer preferences for discounts, free gifts, and more. - **Marketing Team**: Uses generative AI to create content for social media, ad campaigns, and internal communications. - **Supply Chain Team**: Leverages AI to improve sales forecasting and determine inventory quantities to order from brand partners. - **Media Team**: Optimizes media investments through marketing mix modeling, understanding the ROI of each channel, and targeting customers more effectively. ## Two Key Customer Experience Projects with OpenAI Sephora worked with OpenAI on two key projects aimed at bringing the in-store personalized experience online. ### 1. Online Shopping Assistant: Replicating the In-Store Beauty Advisor Experience One of Sephora’s greatest strengths is its in-store beauty advisors, who recommend products based on skin type, skin tone, and explain how to use them. Sephora wanted to recreate this experience online with a conversational shopping assistant that provides personalized recommendations and beauty guidance. #### Demo Flow - A customer searches for “mascara” in the Sephora app and lands on a product page. - Tapping the “AI Beauty” feature at the bottom of the screen, the customer types, “What are the ingredients in this mascara?” - The beauty chat instantly provides a full ingredient list. - The customer then asks, “Is it waterproof?” The assistant answers honestly (this product is not waterproof). - The customer can add the product to their cart and check out with one tap. #### Business Impact This feature increased website conversion rates by over 5%. At Sephora’s scale, a 5% lift is significant, and most personalization initiatives typically only achieve about a 1% conversion improvement. ### 2. Sephora App Integration into ChatGPT The second project integrates the Sephora app directly into ChatGPT, allowing customers to interact with Sephora through a platform they already use daily, further expanding its reach. ## Three Lessons Learned from Building the AI Beauty Assistant ### Guided Experiences Outperform Open Input Guiding customers — for example, providing suggested prompts like “Gifts under €50” or “Summer beauty products” instead of having them type from scratch — significantly boosts user engagement. People can be inherently lazy about typing, so guiding the experience as much as possible is crucial. The user experience must feel seamless and natural, with the assistant embodying Sephora’s tone so that customers feel they are receiving advice from a beauty advisor, not just interacting with AI. ### Explain Recommendations Transparently When the chatbot recommends a product, Sephora proactively explains why — for example, based on the customer’s purchase history, brand preferences, or other known information. Being upfront about why one product is recommended over another is both a responsible use of AI and a way to build trust with customers. ### Embrace a Test-and-Learn Mindset Sephora conducts extensive A/B testing and continuously learns from customer interactions with the beauty assistant, allowing it to quickly release improvements and constantly refine the experience. This iterative approach keeps AI capabilities evolving, rather than being a one-and-done effort. ## Summary: AI Is an Opportunity, Learning Is the Core AI represents an extraordinary opportunity for Sephora. While it is already ahead in some areas, the two use cases shared today are still evolving rapidly. Beyond having advanced algorithms like those from OpenAI, the key to creating sustainable and valuable customer experiences is the ability to learn and evolve together. --- **Source**: Réinventer l'expérience beauté grâce à l'IA - OpenAI (https://www.youtube.com/watch?v=sLYXRA5Ay9g)