There has been a lot of coverage of machine learning (ML) for biological research, for radiology, and for other uses where the direct users are academics, researchers, and medical professionals. However, there is an opportunity for some biological information to be useful in the retail industry. One area is in skincare.

In the middle of the last decade, a book simply titles “Gut” described to those of us outside the biological and medical worlds the increased understanding of the importance of the digestive system to the broader working of the body. The understanding of skin has been going through a similar growth in understanding. My favorite quote from an article published in 2013 is, “The skin is an ecosystem composed of 1.8 m2 of diverse habitats with an abundance of folds, invaginations and specialized niches that support a wide range of microorganisms.”

That’s of interest to more than the medical community. According to Statista, the global skincare market totaled more than USD $150 billion in 2021, and growth is expected to continue. There are many skin care product manufacturers, so it is a difficult process for many people to figure out what works for them. ML can have a place in solving that problem.

One company who has decided to address the consumer market is Skin Trust Club. It is an outgrowth of Labskin, a company with labs in UK, Ireland and the US, which had been involved in skin science research, as a B2B solution, for years. The B2C solution provides ML analysis, developed in the B2B market, with an at-home facial microbiome skin test and updates through a skin health tracking app.

“We were already using machine learning to understand the complexity of the skin microbiomes,” said Colin O’Sullivan, CIO, Labskin. “The speed of machine learning was something we decided could be leverage to address a higher volume of analysis required in the consumer market.” The core of their analysis is driven by XGBoost, as the ML work is primarily classification.

While I had questions about the model of the company sending out test kits and customers returning the samples for analysis, there is a clear and interesting response. “Consumers have become accustomed to tracking their health on an app, and one unintended side effect of the last two years of pandemic is that people have become more used to testing,” said Niamh O’Kennedy, Group Marketing Officer, Labskin. In my own opinion, this is even easier since it is only a swab on the surface of your cheek.

The other key B2C question is the initial market space. There are two components prospective customers need to address: the test kit and the cost of the products. Niamh pointed out that the key segment spending the most on skincare products is white collar women old enough to both have the disposable income and the interest. That means that initial partner pitches have been aimed at companies addressing that demographic. As more firms enter the consumer space, there will be more choice and messages to a wider audience.

For those interested in more of the complexity adding to the initially limited focus, the team pointed to a significant difference between male and female skin biomes. Their research is also seeing impacts in local pollution, between rural and urban environments, and for other analysis that could lead to even more product development withing the skincare industry. People also change over time, so another app on the almost ubiquitous smart phones can be added so that data collection can be improved in between physical tests.

With those changes, of course, comes the question about how much can be learned from a ML driven app versus how frequently real samples must be sent in. This ML technology addresses an interesting segment, and that is a question that can be answered over time. It is clear that the growth of the skincare segment provides yet another opportunity for machine learning to be used as part of a market solution.

Follow me on Twitter or LinkedIn. Check out my website.

David A. Teich

Follow

David A. Teich is interested in artificial intelligence (AI), machine learning (ML), robotics, and other advances technologies