How L’Oréal France Started Real-Time Big Data

Sébastien Garcin, CMO of the cosmetics group, explains how he simplified the lives of his employees by giving them tools boosted to big data.

To operate the digital transformation of an organization composed of 4,000 employees, one would certainly tend to want to take a hammer and a chisel: destroy everything, to rebuild everything. But in a group with dozens of brands and thousands of references, we will opt for a smoother transition. Convinced that the modernization of his group goes through the optimization of data processing, Sébastien Garcia, the chief marketing officer of L’Oréal France since October 2016, has rolled up his sleeves. Its mission: to simplify the life of all marketers, group leaders and brand managers of the brand by providing them with cutting-edge tools.

The challenges were numerous. L’Oréal France may well have advanced in the digital world, “it was two years ago impossible to know how digital tools contributed to the business,” explains Sébastien Garcin. Another pitfall: the lack of flexibility of the tools. “Most third-party solution providers offer modeling tools that are not very transparent and do not offer real-time indicators.” Problematic when more than 300 campaigns a year are launched in digital, of which more than half in programmatic.

Hybrid tool in-house

To be able to measure the precise contribution of each marketing lever to sales, market share, and brand awareness, the group has built a hybrid tool combining home development and outdoor solutions. The challenge: take advantage of structured data in a very diverse and complex way within a single solution. “We talk about big data when we observe the proliferation of references, signs, and levers. There are often differences in structuring between external and internal data. The first is either very structured or unstructured,” says the specialist. And obviously, some data analyzed on social networks are not structured.

“Helping group leaders or brand managers to look forward to six months”

Sébastien Garcin’s teams have developed a SaaS platform called VIBE, for “Very Intelligence Business Experience”. Able to analyze brand or product data over the previous three years, it is designed to help group leaders or brand managers to look forward to six months. “The idea is to relieve them of the least exciting part of the work, by automating what they were doing on Excel, to give them more time to work on the brand’s capital or listen to social networks. ”

If a product team wants to check the effectiveness of a digital campaign during a promotion, the challenge for Sébastien Garcin is to provide him with a tool that can explain what happened and produce results. recommendations accordingly. “On Gemey Maybelline, we used three years of data to be able to model the next 30 months, and after six months we achieved an accuracy rate close to 90% between modeling and reality.”

The margin of error often stems from unanticipated external fluctuations such as competing brands or new launches. The work of Sebastien Garcin’s teams, composed of marketers and data scientists, is now attracting the attention of other group structures. “The management teams, which process some 27 million lines of data each week, are looking to develop a tool to analyze sales in the panels.” Deployed outside the traditional IT processes, the VIBE platform is destined to disappear, since, at the group level, L’Oréal is preparing its own tool.

Marketers developers

“The fact that we started earlier allows us to go much faster in our data standardization projects, and I think that the role of the IT teams will change in the future and will have to focus on the issues of standardization of data and infrastructure “. It will be up to the marketing teams to hire their own developers to perfect their applications more in line with the complexity of each business. “In other words, this means training more and more of these technicians in the marketing profession,” concludes Sébastien Garcin.

Discover in the new Adtech News

This article is from the latest issue of Adtech News, the quarterly paper supplement dedicated to the latest advertising trends of the JDN and CB News. On the agenda: the CRM revolution with the AI and the data, a comparison of the solutions of header bidding, a preview of the blockchain in programmatic …

Leave a Reply

Your email address will not be published. Required fields are marked *