By Will Novosedlik
Due to its strong commitment to innovation and rigorous cost control, L’Oréal has just pulled off its best financial performance yet. Fiscal 2021 saw record growth of 16.1%, twice that of the beauty industry as a whole.
While we’ll leave the cost control discussion to the accountants in the room, there’s lots to discuss when it comes to innovation and beauty tech.
The latest development has been the creation of its Business Data Lab. To learn more about that, we sat down with Ludovic Bégué, the company’s Montréal-based head of data science and CRM.
What is the thinking behind the Business Data Lab?
When I joined L’Oréal Canada in 2020, part of my mission was to improve our use of data. With 40 brands, a plant and a distribution centre, we have a massive amount of it. We have competitive data, customer and consumer data, and product information data. Last year we assessed how we are currently using all these different types of data and how it can improve customer experience and business operations.
Our idea was to access the assets we have here in Montréal in terms of post-secondary institutions and start-ups. We put together a team and challenged them to develop an innovative data solution that would change the way we do business. In this way, we’ve created a business-driven innovation hub.
What are the key opportunities for this hub?
We discovered three key takeaways. The first was that we did indeed have massive business opportunities with our data, whether it’s for sales forecasting, inventory management or personalization. We could see opportunities for data to help in all these areas through automation, algorithms and machine learning.
The second opportunity was in the area of data education. In order to do proper analytics, users need to understand what quality data looks like and how best to structure it. So there is an opportunity to better leverage our internal tech talent.
The third was that while we’re primarily focused on the present and on learnings from the past, we don’t have enough of a view to the future.
You also have something called the Digital Services Factory. What’s the difference between that and the Business Data Lab?
The way we see it is that the consumer will start her beauty journey with us online. If she agrees to share her data, we will use it to understand her needs and then curate a highly personalized experience for her. That’s the Digital Services Factory, the part that the consumer actually experiences.
To enable that, the Business Data Lab will cross this information with any sociodemographic and touchpoint data that we have on the consumer to make sure that the experience was sufficiently personalized and that the consumer does not go back to being unknown to us. We want to make sure that we bring the right beauty solution to every single one of our clients.
One of the things you say you can do is to change product descriptions to suit the language of your consumers. How do you do that?
Our Amazon algorithm is the best way to illustrate this. Everything on Amazon starts with a search. Let’s say you’re looking for men’s body wash. We need to buy some keywords. Until a few months ago, we were very brand and product centric. But sometimes this is not the best way to make a connection between the beauty concern of the consumer and our product. We use natural language processing to classify keywords, then we rank them so that we are able to use them to sell a skincare product. The top keywords of 80% of the searches on this type of product are “hydration” and “moisturization.”
We also look at trends. You might have 500 or 600 keywords that are not ranking high today, but you can observe the velocity and frequency of their appearance over time. What was ranked 1,300 a week ago may be ranked 500 today. That tells us that it might be the right moment for us to place our bets on it. We can then adapt our product description page to make sure that we speak the same language as the consumer.
Sometimes we avoid keywords we don’t think are beauty related. But, for example, we have a very good product that can help with a skin condition similar to eczema called keratosis pilaris. There was a case where the algorithm told us there are a lot of women who are looking for a solution to this. So we started introducing this keyword and suddenly we saw sales jump by 70% because the product was speaking the language of the consumer.
Can we talk about how you use open innovation?
L’Oréal Canada has used its tech to host more than a hundred live events across nine brands. The last time we did it, there were several teams. One was focused on sustainability, another on gamification and another on philanthropy. We’re lucky to have so many start-ups in Canada. We also work with McGill University. We’ve done two use cases with them on churn, about trying to predict who will stay with the brand and who will not. We have also worked with students from the Masters of Marketing program at McGill. It’s humbling to work with students. They’re much more advanced than us in many ways.
As for the start-ups, what kinds of skills are you looking for?
You have basically two kinds of people: very advanced analysts who are comfortable talking about technology, and people who will talk about the business challenges that you are trying to solve. I prefer the latter. An example would be our partner Rocket Science Development. They always start with the business context.
What about social impact?
When we talk about social contribution, it’s really about diversity, equity and inclusion. For our brands, diversity is extremely critical and we can also use data to ensure that we provide the right products to the right people. We really want to take into account the fact that everyone is unique. Canada is particularly diverse and tolerant. So it’s much easier for us to actually test campaigns or new ways of approaching the customer because in Canada, diversity is part of the landscape. Diversity really helps us speak the right language to the right person at the right place and time. What works here will very likely work anywhere else.
What’s next for the Business Data Lab?
Two things. With data and AI, I want to make sure that from a consumer experience standpoint, we have some kind of horizontality, so you are always treated the same way, wherever you are, whatever channels you engage with. And that means working in coordination with retail and the IT team. The second thing for me is foresight. That’s really something that we can improve. How can we provide our executive with the right insights? How can we look not three or six months ahead, but two to five years ahead? So your earlier question about foresight was extremely relevant. That’s somewhere I want to go, a capacity we need to develop.
