Artificial intelligence to test lipstick, shampoo and cream – 07/15/2023 – Market

Checking the color of a lipstick or foundation on the screen is often not enough to make a good choice. In the same way, buying a shampoo or conditioner without knowing for sure what your hair really needs can seem like a long shot. From a manufacturer’s point of view, it can be extremely risky to launch a beauty product on the market without being sure how effective it is.

The cosmetics and beauty industry has been using artificial intelligence to increase the level of security in these situations involving the purchase or launch of products. Companies such as multinationals L’Oréal, Natura & Co. and the Brazilian Boticário have been investing more and more in technology in order to guarantee assertiveness in the choices, whether by the consumer or their own bets on the market – which, by the way, is only growing.

Brazil is the second largest perfumery market in the world, only behind the United States. Sales in 2022 moved US$ 6.8 billion (R$ 32.6 billion), according to the Euromonitor consultancy. In makeup, the country is among the ten largest global consumers, with revenues of US$ 2 billion (R$ 9.5 billion) last year.

L’Oréal, the world’s largest conglomerate in the beauty sector, with revenues of US$ 38.26 billion in 2022 (R$ 183.1 billion), has been using the augmented reality tool ModiFace to increase its sales. “It’s a powerful recommendation tool for the consumer”, says Guilherme Eler, new business and innovation director at L’Oréal Brasil. The company bought ModiFace in 2018 and, since then, has offered more than 1,400 services through augmented reality, involving face and hair.

Among them is the Virtual Beauty Club, a partnership between the Maybelline brand, from L’Oréal, and Shopee: in the online retailer’s application, it is possible to “try on” the colors and effects of lipsticks, bases and eyelash masks.

Another service was Effaclar Spotscan, by La Roche-Posay. The skin analysis tool uses a database of 6,000 scientific images of men and women with different skin types and different levels of acne severity to pinpoint the best treatment.

“The client takes three selfies and receives a personalized recommendation of the care routine he should adopt”, says Eler. According to the executive, after the tool was launched, the sales conversion rate doubled and the average ticket was 25% higher.

In June, L’Oréal attended Viva Technology 2023 in Paris, where it showcased the latest in data-powered diagnostic devices. Among them is the Kérastase K-Scan: an online hair diagnostic tool. By answering a series of questions, which lasts about two minutes, the consumer receives an indication of the beauty routine her hair needs.

Something similar is already offered by the Brazilian multinational Natura & Co., owner of the Natura, Avon and The Body Shop brands. The company launched the personalized recommendation of Lumina products via capillary diagnosis. Through an application, the smartphone camera and a portable lens attached to the device, it is possible to increase the image resolution of the hair strand by up to 20 times to observe the damage present.

To standardize the images and obtain assertiveness in the diagnosis, a database was built containing more than 18 thousand reference images of hairs, contemplating hair of different curvatures, colors and levels of damage, including the diversity of the population.

In this way, it was possible to recognize patterns in the images referring to the damage to the surface of the wires, which allowed the generation of the AI ​​algorithm for the personalized diagnosis with a precision that is improving as new images are included in the database. That is, the system itself relearns and updates itself through the machine learning process.

The artificial intelligence of hair diagnosis is now capable of recommending more than 15,000 product combinations from the Natura Lumina line. Innovation is present in nine Natura physical stores.

At Boticário, AI is starting to be used in product quality control. Using proprietary software, Lyra, the company has been working on crossing data on formulas and packaging with product stability tests, even before launch. The test serves to check the effectiveness of items such as creams and makeup.

“We have gathered data from the last 14 years, involving 3,000 stability studies, with 7,500 tests carried out a year, and we are going to start crossing this information”, says Juliana Canellas, quality director of the Boticário group —owner of brands such as Boticário, Eudora and Who Said, Berenice?.

The objective, according to her, is to reduce the time of arrival of the product to the final consumer, in addition to reducing testing costs. “Our expectation is to advance the launch of a product between 60 and 90 days, in addition to generating savings of R$ 6 million a year with stability tests.”

The company was the first to use artificial intelligence to produce a perfume – Egeo, launched in 2019. Since then, investments in technology have accelerated. In the last four years, the company multiplied by ten the investments in the area (not disclosed). In personnel, the team increased from 200 to 2,500 people, in an effort to bring more innovation and quality to the products.