We caught up with Sephora Global Chief Digital Officer Anca Marola from the sidelines at the recent World Retail Congress in Berlin. She outlined how the retailer is attempting to bridge the gap between its stores, ecommerce operations and emerging AI-powered shopping experiences, including a new pilot partnership with ChatGPT in the U.S.
Marola stressed that Sephora remains on track to open approximately 150 new locations annually, adding to it 3,400-store global footprint. Sephora is also increasingly using AI and digital tools within the in-store experience. One of the retailer’s most significant initiatives is its skin scan technology, which has now been rolled out across 24 countries.
The system combines hardware, AI and Sephora’s proprietary software to create personalized beauty recommendations. Customers can receive skin diagnostics, shade matching and skincare analysis directly in-store; the results are then stored in the Sephora app if users opt in.
“People still love the stores, the excitement remains, and I think that’s pretty irreplaceable,” she said. “Our goal is to capture the consumer in-store, where they can try on make-up and smell the fragrances. There is no way you can replace that, being guided by a real person with genuine expertise, but, given that around 30 percent of sales are online, we wanted to look at how we blend the physical and online experiences.”
Can high tech enable the high touch of beauty products? And the answer is: Sephora pioneers conversational AI to bridge the store and online app to accelerate personalization and customized experiences.
Sephora In-Store Scan
The scan is free but needs to be booked ahead of time and provides skin diagnostics and shade analysis using hardware and AI to control the lighting and to make it extra precise. Marola emphasized that the physical hardware is a crucial differentiator because it enables far greater accuracy than smartphone-only solutions. “What is important is that our approach includes hardware, AI and the app; we rolled out our scanner to 24 countries,” she said. “This brings real value to the customer and is exclusive to Sephora, because the hardware provides a very accurate scan of the client’s skin and tone.”
The system can adjust lighting conditions, zoom into the skin and even measure hydration levels before the data is analyzed by Sephora’s software and matched against the retailer’s proprietary color swatches developed in-house. “The skin tones are compared with the proprietary color swatches that were developed at our labs and, if the customer gives permission, all this can be stored on the app, and that information can be used to guide what products the customer sees, to make sure they are aligned to their skin type and tone,” Marola added.
Sephora’s Omnichannel Approach
The skin tone and type initiative forms part of Sephora’s broader effort to create a more integrated omnichannel retail ecosystem and unified communication strategy, especially as consumers themselves change their interaction preferences.
“To me, that’s exciting because it’s the first time, we can bridge that gap between the stores and online,” she said. “We are evolving our ecommerce into more conversational advice, just like in-store, when we can tailor the experience. We do have a lot of information, including 14 million product reviews, loyalty programs and preferences, the beauty scan, connecting this to make it more personal,” Marola said.
Sephora’s AI ambitions are particularly at the forefront in the U.S., where the retailer has begun piloting a dedicated app within ChatGPT in partnership with OpenAI. Not only is North America a major market, but data privacy laws are far more relaxed than in regulation-heavy Europe, hence the U.S. is typically a test market currently for AI pilots.
The initiative allows consumers to ask natural language beauty questions directly within ChatGPT and to receive personalized recommendations linked to their Sephora Beauty Insider accounts. The pilot also lays the groundwork for future in-chat purchasing and checkout without ever leaving ChatGPT.
“We are innovating our own app within ChatGPT in the U.S. because we can see the consumer evolving and enjoying this more conversational approach with agentic AI,” Marola said. “The AI also has access to our bank of 14 million product reviews by consumers, giving those reviews real weight and authenticity because it is not just about what we are saying.”
It’s a move that positions Sephora among a growing number of major retailers experimenting with AI-driven commerce. Retailers including Walmart, Amazon, JD Sports, Nike and Ulta Beauty are all investing heavily in conversational shopping tools and AI-powered recommendation systems and in launching their own inside-AI purchasing platform at the start of the year, JD Sports boss Regis Schutlz said that many younger shoppers actively prefer talking to a chatbot.
However, Sephora believes beauty retail is particularly suited to conversational AI because customers often seek guidance, reassurance and expertise when choosing products and increasingly prefer for that dialogue to be more natural. The retailer also sees AI as a tool for localization and marketing adaptation across its growing international footprint, which now spans 36 markets. “We have local teams everywhere, which means we can localize the messaging and communication, and we try to understand and adapt to the local culture, which we are also using AI to help us with creatively,” Marola said.
Adopt and Adapt
For Sephora, the future appears to lie in using AI to streamline discovery, personalization and convenience, while preserving the emotional and experiential aspects of beauty shopping that still drive consumers into stores. Marola added that she wants to double down on omnichannel convenience, giving customers the fastest delivery possible, which is down to two hours for in-store click and collect.
But the real debate for retail is about speed versus control. Customers are also progressing, seeking new channels, and getting more comfortable seeking advice on ChatGPT and the like. But broader adoption of technology in parallel with the huge amount of user-driven content means that brands are loosening the strict guardrails within which their IP sits. The question for any retailer is whether they want to protect that control or feel that it is worthwhile to move outside their comfort, and control, zones.
“Customer behavior is definitely changing in response to their experiences with AI, and we are using more and more chatbots to help them,” Marola said. “Our view is that it’s going to happen, and we want to be an early adopter, test and learn. That means we get to help shape it. You can’t stop the momentum, so you have to decide whether you want to be part of it or not.”
