Beauty may only be skin deep, but that doesn’t mean you have to be confused when trying to care for that skin. Olay is helping consumers learn which products are best for them with a new AI-powered tool called Olay Skin Advisor. The first deep learning application in the beauty industry, Olay Skin Advisor is a web-based experience that uses selfies and customer concerns to recommend a personalized skin care regimen.

The beauty industry is an early adopter of consumer technology, utilizing augmented and virtual reality to help consumers visualize the latest trends—but Olay has always been ahead of the curve and it revealed the AI tool during Mobile World Congress 2017.

Olay Skin Advisor was inspired by decades of research into skin measurements and image analysis,” Dr. Frauke Neuser, Olay principal scientist told [a]lisdaily. “Many people don’t realize that it was Olay scientists who developed the technology behind the portable facial imaging systems most commonly found today in department stores and dermatologist offices. These systems are great tools to provide a personalized skin analysis—but the technology behind it is linked to feature-based image analysis and is therefore hardware based. In order to make this type of consultation mobile and available to many more women, we knew we had to come up with a different, ‘smarter’ approach. AI was the perfect fit to provide an extremely accurate, yet educational and engaging user experience.”

Users who visit the Olay Skin Advisor website from any smartphone or tablet are prompted to answer a few questions about their existing skin regimen, age and areas of concern such as wrinkles, dark spots, acne, etc. After snapping a selfie, the tool identifies the user’s “skin age,” the best skin areas and the areas that need the most improvement. The program learns from “hundreds of thousands of selfies” in its database—but don’t worry, the company won’t hold your no makeup selfie for ransom.

“The first selfies originated from research specifically aimed to collect enough selfies to train Olay Skin Advisor’s VizIDTM deep learning algorithm,” explained Dr. Neuser. “Additional selfies are now coming in from the hundreds of thousands of women who have used Olay Skin Advisor to date. The images collected from each upload are used to make the tool ‘smarter with every use’ by strengthening the algorithm to analyze and identify each woman’s skin age even more accurately. These images will also be used in research over the next 10 years, to ultimately improve skin understanding and deep learning model development. No other personally identifiable information is stored with the images.”

Naturally, users will receive recommendations from Olay’s line of skin care products based on the AI’s analysis, along with instructions on how to apply, when and in which order. Consumers are shown pictures of these products along with descriptions, reviews and a link to buy. Club Olay members can save their recommendations and newcomers quickly sign up for free when prompted.

I can say from experience that the beauty isle is downright saturated with skin care products, so having any idea where to start is extremely helpful.

“There are several advantages of using AI over a live consultant,” Dr. Neuser said. “One is accuracy, as Olay Skin Advisor uses AI to more precisely calculate the user’s skin age and pinpoint their best and improvement areas. Like a real-life beauty advisor, the more faces it sees, the more it will understand about skin, and the more it will be able to make the most personalized recommendations for skincare regimens. However, unlike a live consultant, AI is unbiased. In addition, Olay Skin Advisor can be used anywhere, anytime.”