How L’Oréal Wins Big Online, One Shopper at A Time

Lots of products to tag, lots of unique customers to engage

The product line runs deep at L’Oréal. Finding a way to efficiently identify product attributes was a big challenge for shu uemura. But it had to be done if L’Oréal wanted to achieve their main goal of recommending products to individual customers.

In the past, L’Oréal’s Data Team had to identify VIP customers one by one from their CRM database. Then they had to customize different EDM marketing campaigns. The process took a lot of time and money.

Realizing this was not a sustainable SOP, the Data Team sought out Rosetta AI and applied machine learning algorithms to automate product tagging, identify customer preferences, and provide personalized recommendations.

Automated tagging

The Rosetta AI solution automatically creates product tags for individual preference profiles, a rich source of data for you to create truly personalized recommendations. This solution saves time and money because the AI can look at the product catalog like a human Data Analyst and create tags.

The automation relies on computer vision trained by Rosetta AI’s beauty and fashion industry experts, ensuring the tags are written to industry standards. Product category tags can be set (#brands #feature #material #skin #position, for example), and tags can be written into product descriptions. Beyond that, the AI takes over, identifying tags from catalog images:

Personalized recommendation boxes

The Rosetta AI service uses deep learning technology to analyze consumer preferences and onsite behavior. As shoppers visit, the AI analyzes the context of the experience and delivers personalized recommendations.

The visitor may be on the landing page, category page, product detail page or cart page, but the algorithm always ensures that just the right products are presented in recommendation boxes. On their product detail page, shu uemura uses “Recommended just for you.”

Preference profiles include the favorite colors, preferred makeup attributes, and past purchase history. This data is used to select the best products for that individual customer at that particular time.

Personalized pop-ups

Rosetta.ai can create personalized pop-ups that offer one-to-one recommendations or discounts. These pop-ups engage customers as individuals on the shu uemura site and convert much more than generic pop-up offers, especially with hesitant shoppers.

The personalized discount pop-up presents special offers on products that match the shopper’s preference profile.

Hesitant shoppers make up 10%~20% of all website visitors, but with personalized pop-ups, shu uemura has been able to encourage them to buy more and conversion rates are up 140%.

More cross-sells

The extra tagging work on the backend enriches individual preference profiles, making recommendations more accurate. Whether it's a recommendation box or a pop-up, the right offer at the right time makes each experience unique and drives more cross-sells. Customers click on more products, stay longer and buy more. For shu uemura, the average order value went up 109%.

L’Oréal’s goals for adding personalization to their online marketing included improving CVR, AOV and revenue, as well as saving time and lowering costs. In less than a year they were able to achieve all those goals with the Rosetta AI personalized marketing solution.

The large gains speak for themselves, with increases of 140% for conversion rate, 109% for average order value and 149% for revenue.

Business growth for Rosetta AI clients like L’Oréal and shu uemura is steady and strong because the solution offers real one-to-one personalization, not segmentation.

Customers are emotionally engaged by the results and the AI-driven automation frees up human resources behind the scenes to focus on other tasks.