How L’Oréal is using AI and Machine Learning to disrupt the beauty industry
Personalization is the future of the beauty industry
Anyone who purchased a makeup one day knows it is not an easy task: the market is taken by an infinity number of brands, people have very different skin colors and face types, and products are usually expensive to allow for error.
In order to solve those issues, beauty brands have recently launched different apps and gadgets to promote their portfolio and connect to the customer on a more personalized way. These technologies help to directly address customers concerns, solving their beauty pain points without wasting time or money with products that won’t actually do a great job. Companies can have access to infinite behavior consumption data, using it as input to develop products and more targeted and efficient market campaigns. Personalization builds loyalty by offering a more intimate relationship and is a crucial tool to differentiate a company’s product and customer experience in the competitive beauty market.
L’Oreal, the world’s largest cosmetics company, is taking the lead in the beauty industry and betting on augmented reality and machine learning. Since 2012, the company has its own Technology Incubator division, a start-up like team focused on matching beauty and technology  and in March of this year, L’Oréal announced the acquisition of Modiface, a major Augmented Reality beauty company that provided solutions for L’Oréal and its competitors. The acquisition gives L’Oréal a leg up and highlights the importance of securing a technological march over competitors already in the short-term. 
The first product launch of the Technology Incubator Division was the Makeup Genius in 2015, a mobile app that uses Augmented Reality to let customers try makeup using their cellphones and uses a collection of data points around facial characteristics, including face shape, skin tone and wrinkles. Most recently, in 2018, the company launched a similar app in partnership with Modiface called Style my Hair 3D, this time focused on simulating different hair colors in live mode. These technologies help retailers determine how people with specific facial characteristics may be more likely to purchase certain types of products, and help predicting inventory with greater accuracy. 
Another innovative technology focused on hair was put in the market in 2017 with the Kérastase Hair Coach, the world’s first smart hairbrush.  The brush has sensors equipped with modern microphone technology that captures the hair’s soundwaves and submits them through analytical algorithms to diagnose the hair health. The customer can use the accompanying mobile app to check a detailed report that takes into account factors such as humidity, temperature and wind, and to receive personalized treatment recommendations. The company suggests a continuous use of the product to allow time for the algorithm to compare the data and learn more about your behavior to give a more accurate recommendation.
Over the longer term, L’Oréal is also investing on 3D printing and taking the lead on addressing the next wave of disruption in the industry: the product personalization. The company has already many prototypes and has recently applied for a patent of a 3D printing of fake nails that envisions molding a custom-shaped false nail for its customers.  The 3D printing can also play an important role on product development and trial phase. On May 2015, L’Oréal announced a partnership with the bio-engineering start-up Organovo to work on 3D printing of human skin. Currently, L’Oréal grows skin samples in a lab using incubating technique performed by hand by their 60+ scientists. The adoption of an automated process using 3D technology could increase efficiency of production significantly and for some companies replace animal testing. 
L’Oréal is on the right track leading the beauty industry technology disruption and made a smart move in acquiring Modiface this year. The company should keep eyes open to new companies and technologies in the market and work to increase awareness and adoption of its current technologies. It is very important that L’Oréal also takes the lead on the ethics of the market and builds trustful relationships with its customers regarding to Data Security and reliable recommendations.
For the longer term I see opportunities in partnerships with healthcare players to apply these technologies to a more profound diagnosis of customers health. For instance, data collected from the hair/skin could diagnose high UV exposure or lack of a specific vitamin and a better understanding of the impact of human behaviors (i.e. hours of sleep) on hair/skin health. These advancements could lead to development of powerful personalized products. However, some questions remain open on the customer side: can we trust L’Oréal personalized recommendation or it serves the company’s best interest for selling specific products? Will the smart detection technology be extended to other dimensions of our lives? (i.e. glasses that monitor our vision or teeth brushes that monitor oral hygiene)
Student comments on How L’Oréal is using AI and Machine Learning to disrupt the beauty industry
Your article made a good point on the importance of personalization in the beauty industry, and how technology plays a vital role in facilitating such services. However, it is not clear which megatrend ( AI, Augmented Reality or Additive manufacturing ) your want to focus. For example, I would be very curious to learn how exactly Modiface uses data to predict customers’ preference over how he/she would look like: what type of foundation, what type of brushes, and what lipstick color. These are the fundamental question we need to address when talking about AI, because it is the process of how analyzing data helps an organization serve its customer more effectively and thus make a profit that matters the most. What I do appreciate from your article is that you brought the perspective of healthcare in the scene as it’s an important factor in personalizing one’s make up product purchases.
Looking at Modiface’s augmented reality product offering, I’m finding difficulty understanding how the platform leverages machine learning specifically to generate its recommendations so would have loved to see more detail on that front (e.g. are they leveraging a customer’s previous purchasing behavior to refine their model? Or is it simply based on superficial features like face shape, etc). I agree with you that there might be some distrust from customers in any of these technological offerings since they may not fully understand the logic behind predictions (e.g. are the recommendation engines providing predictions based on higher margin SKUs or truly based on what is best for the customer) so it might be in L’Oréal’s interest to provide more transparency to the consumer and also give consumers a chance to provide direct feedback to the model.
I enjoyed reading your essay on L’Oreal and the commentary on the make up industry, as I know first hand the difficulty of finding the right products to use. I like thinking about the extension this will have on the health industry as well. I agree that trust will be a very important guide here, as most people will say they don’t just wear one brand of make up but vary make up brands based on product. Thus, will the recommendations given to the consumer be truly a representation of what is best for them, or merely what they want to sell to make the most profit.
Really exciting innovations happening at L’Oreal, thanks for sharing! With the increasing move to e-commerce, I think that the augmented reality technology that allows you to “try on” products has a lot of potential to drive online demand and reduce costs of holding inventory in brick and mortar stores. The smart hairbrush and other measurement- type technology could also be an interesting way to track progress and encourage regular use of a product, to help users recognize gradual results (especially in skincare). I do think that a big risk that L’Oreal will need to overcome is the trust aspect of ML and data collection. I assume that people trust a human much more to recommend products than a program (though I may be wrong!) and the idea of scanning something as personal as your face, can be scary, with implications for facial tracking, identity theft, etc.