L’Oreal Expands Virtual Try-On Service

The company’s U.S. subsidiary rolled out a service this week that lets shoppers use their phone’s camera to simulate various hair-color shades from L’Oréal products Garnier Nutrisse and Garnier Olia using augmented reality technology.

Using the Google Lens image-recognition mobile app, made by Google Inc., shoppers can point their phone cameras at the hair-color boxes to activate the virtual try-on service. The application identifies the product and color and then activates the virtual try-on service, powered by technology from ModiFace, a company that L’Oréal acquired in 2018 for an undisclosed amount.

Customers can use the feature anywhere the two Garnier products are sold, but 500 participating

Walmart Inc.

stores will have signs near the products demonstrating how the virtual try-on service can be used.

When online and in-store customers use technology that lets them virtually try on beauty products before purchasing them, L’Oréal sees three times the rate of customers purchasing a product, Ms. Rochet said.

The company has been working with Modiface’s augmented reality technology since 2012, Ms. Rochet said, and decided to buy the business after coming to appreciate its potential.

“We realized that AR and AI would really change the consumer experience for L’Oréal and the industry,” Ms. Rochet said. “We decided that partnership was not enough. We really wanted to source that capability internally.”

It is difficult to determine how certain products, especially hair color, will look on a person until they try them on, Ms. Rochet said. “There’s a real advantage to guiding your consumer to find the right color,” Ms. Rochet said.

And while retailers may offer makeup samples in stores, customers might not have time to try on multiple shades and wipe them off, or they might not want to experiment with samples that have already been used.

L’Oréal is among several cosmetics companies turning to augmented reality and artificial intelligence to boost sales. Cosmetics retailers that offer similar virtual try-on services include

Ulta Beauty Inc.,

which recently acquired two startups that specialize in augmented reality and AI, and Sephora, owned by LVMH Moët Hennessy Louis Vuitton SE.

Sales for L’Oréal between January and September of this year were about €22 billion, up by about 8% compared to the same period in 2018. The company said in October that one of the main drivers of growth was increased e-commerce sales, which grew about 49% during the first half of 2019 compared to the same period last year. The company does not break out e-commerce sales.

Shoppers are spending more on L’Oréal products online, partly thanks to ModiFace’s technology.

A customer using the ModiFace tool typically tries on more than 20 different shades of makeup, Ms. Rochet said. Clicking a “try on” icon next to a product lets the app use the customer’s laptop or phone camera to generate a preview.

Simulations of makeup and hair color have gotten more true-to-life with advancements in quality of the camera on smartphones, as well as processing power in the phones, said Parham Aarabi, chief executive of ModiFace, based in Toronto and founded in 2007.

“It took a long time for the technology to get better,” Mr. Aarabi said.

Websites such as

Amazon.com Inc.

and Sephora allow shoppers to virtually try on L’Oréal makeup products including lipstick, eye shadow and foundation using ModiFace’s technology, Mr. Aarabi said. The same feature is also offered on in-store tablets at bricks-and-mortar stores including Sephora.

The ModiFace technology uses an AI system called a neural network to automatically identify parts of the user’s face and apply the right shade of makeup on the right part of the face. Mr. Aarabi said. The AI systems also analyze the background lighting to simulate the makeup color correctly.

Neural networks are composed of layers of interconnected artificial “neurons” that automatically learn about the features of a specific object based on large amounts of training data. The ability to create such systems with multiple layers has led to advances in speech recognition and computer vision.

Write to Sara Castellanos at sara.castellanos@wsj.com