By Pragati Verma, Contributor

 

Y Combinator-backed startup Proven Skincare is using artificial intelligence (AI) and data analytics to give skincare a makeover. The core idea is to create a skincare routine personalised for each individual.

 

“Each of our product caters to a specific customer’s skin needs,” explained Proven Co-founder, Ming Zhao. “What works for a 20-something living in a high-humidity area might not work as well for a 40- or 50-something living in a dry climate. Similarly, people living in New York or Chicago might not need the same formulations as someone living in rural Oregon.”

 

Potential customers begin by answering an online questionnaire. Proven’s AI engine then compares their answers with data from 8 million consumer reviews and 4,000 scientific journal articles, using natural language processing and fraud-detection algorithms. From here, it formulates proprietary products tailored for each specific skin type and extrinsic condition, taking into account level of pollution, hardness of water in a particular area and UV coverage.

 

This idea of “combining AI and data analytics to understand the correlations and interconnections between people’s skin and the ingredients that work for each person,” she said, “is all set to hack the skincare products market.”


The Best Match


Zhao and her co-founder, Amy Yuan, started this journey with their own skin issues. For Zhao, the struggle to find products for her “difficult skin” proved to be expensive, time-consuming and frustrating. Nothing worked until she found a “skin guru” who created customised skincare products just for her.

 

“[Amy] had atopic dermatitis, [a chronic condition that makes skin red and itchy],” Zhao added, “Being a data scientist and computational physicist, she took a scientific approach to the problem. Amy had done a lot of big scale supercomputing simulations and she wrote an AI engine that analysed reviews for her to find skincare products that were right for her skin.”

 

When they met about four years ago, they combined their epiphanies about data-powered knowledge and personalisation — and Proven was born.

 

A big centrepiece of their offering is their database—the Skin Genome Project, which won them an MIT 2018 AI Idol Award this past year. According to Zhao, it is the most comprehensive analytical database of clinically effective ingredients for skincare ever created.

 

“At this point, our deep learning algorithms pick out useful and relevant information from more than 100,000 products, 8 million testimonials and reviews from over 4,000 scientific journal articles to analyse the effectiveness of over 20,238 skincare ingredients,” Zhao explained.

 

This is huge, she continued, as it eliminates the trial-and-error methods that are traditionally used to match people to products that work best for them. At Proven, their algorithm identifies the ingredients that work best for a person’s skin type, genetics, lifestyle and environment. Proven’s products are then formulated in collaboration with S. Tyler Hollmig, head of aesthetic dermatology at Stanford University.

 

“We have an award-winning cosmetic chemist to help us formulate our products,” Zhao said, “and we have dermatologist advisors to put a human touch on our huge database knowledge.”


Beauty Gets Personal


Despite its revolutionary algorithmic technique, Proven is not alone. Several other startups are taking a tech-heavy approach to personalised skin and hair care.

 

While Function of Beauty uses an algorithm to create customised shampoos and conditioners based on their customers’ hair types and goals, Curology blends a prescription-strength, personalised formula to treat acne, taking into account customers’ skin problems, medical history, photos and skin goals.

 

For Zhao, this influx of data-powered personalised beauty products makes a lot of sense. “Our medication is personalised to what we actually need. Skin is our body’s largest organ and consumers are realising that they need products tailored to their needs. Technology is enabling us to listen to consumers and provide them the products that make the most sense to them.”

 

It’s this personalised product model, said Zhao, that also enables them to reach demographics traditionally underserved by the beauty industry, such as ethnic minorities, people living in rural areas and men. The expanding user base — coupled with its powerful database — is where Proven seems to have cracked the skin care code.

 

“[Our database] is growing stronger all the time, as our users take the quiz and use the products and share that knowledge back with us,” Zhao said. It’s this combination of data, customer experience and personalisation that she and Yuan feel make their brand unique.

 

The duo’s ambitious vision, however, doesn’t stop here. Making Proven’s database a de facto source of truth for the entire beauty industry is no easy feat, but Zhao believes it is possible for two reasons: “Ours is the most comprehensive database in the industry and that [data-powered personalisation] is the direction the industry is moving in general.”

 

Personalised skincare, she noted, is not just necessary, but it’s the future of the industry. If startups like Proven succeed, they will bring deep changes in the skincare industry, projected to exceed $131 billion next year.

 

As Zhao explained with a rhetorical question, “Why would you go to a store to buy one-size-fits-all products when you could get all your beauty products designed just for you?”