We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we'll assume that you are happy to receive all cookies on the ESCP website. However, if you would like to, you can change your cookie settings at any time.


Every year, students put Prof. Louis-David Benyayer’s teachings into practice in a corporate context. One group helped L’Oréal define their “make or buy” model by using data and analytics.

Ever heard of Datanomics? That’s the title of the first book about the value of data (in French, sadly) co-authored by strategy consultant, ESCP alumnus and professor Louis-David Benyayer. “Data transforms the way companies create and capture value, and redefines competition. Mastering these Datanomics is paramount for all managers and entrepreneurs to sustain the competitive advantage of their firm. Datanomics and Strategy analyses these changes to support leaders in their strategic decision-making process,” he explains on his eponym blog.

When students start the MSc in Big Data and Business Analytics, their first course is “Data-driven business strategy”. The objective is to give them a global understanding of the forms and sources of value of data, and equip them with skills to design and implement projects to solve business problems using data and analytics in a corporate context. The course is built on two pillars: First, lectures on theories, research and examples of value creation opportunities with data and analytics and the mechanisms at play to turn promises into reality; Second, a company project where they prototype a solution for solving a business problem of a partner company. “The two elements complement each other: it's important students are able to apply the frameworks in a real-life situation,” the scientific co-director of the specialised master’s adds. “The company project is an opportunity to interact with business leaders on their current problems. It's also a good occasion for them to showcase their skills and to meet and work with their classmates. Reversely, our partner companies appreciate getting a fresh view on their problems and having new ideas and propositions to solve them. Frequently, the interactions between the students and the companies continue after the course to implement some of their ideas.”
Over the past four years, professor Benyayer and his students have worked with companies like La Redoute, Safran, Etalab, L'Oréal, Bouygues Construction, Catalina, Nissan and Groupe Up.

A very interesting company project for L'Oréal

For example, a group was asked to help L’Oréal – not for the first time - develop a rationale they could present internally to guide them in the definition of their “make or buy” model for technology. “We need to accelerate our product development and be much more agile in terms of manufacturing and distribution. We must make sure that we leverage data to develop new innovations and personalised beauty,” explains Operations Digital Transformation Director Cristina Monnoyeur. L’Oréal’s objective is to become more data-driven, with strategies to collect data to make objective, rather than subjective, decisions.
“In our fourth Data-driven business strategy session, we covered strategies for leveraging data assets in which Louis-David Benyayer discussed the four options for acquiring them: Make, Source, Buy and Partner. This section of the course was particularly useful to us as it was very specific to our project,” explains Cassandre Sillere, who worked on the company project with Adil Soundardjee, Christophe Sanchis and Yun Zhang. “It was nice to see that there is not just one option that you should systematically use for all parts of the value chain. ‘Make’ offers you complete customisation, but you don't benefit from the scale effect. ‘Buy’ allows for a shorter delivery time, but there is a risk with the execution of the integration.”

It was all the more interesting to them since the example Louis-David Benyayer used in class was that of L'Oréal. A general example of a make or buy situation would be software: L'Oréal could choose between making it in-house or buying it from an external supplier. There are advantages and disadvantages to both, depending entirely on the situation: Whether or not the company has the resources in-house; how much time it has; how much customisation is needed, etc. The acquisition of Beauty Tech company ModiFace when they wanted to reshape the beauty experience through augmented reality is a rather extreme case of 'buy'. ModiFace is now part of L'Oréal, and would now be considered 'make'. “All parts of the course were very useful in the project and certainly helped us come up with a solution to L'Oréal's ‘Make or Buy’ strategy, but it was a particularly memorable example for our group, as we had done some research on them and already had quite a few interviews with L'Oréal employees in which ModiFace was mentioned,” adds the student whose goal is to become a consultant specialising in data-driven business strategy.
“Currently, L’Oréal mainly buys but this is starting to change with the recruitment of more and more engineers. We clearly saw that the engineer-heavy teams try to build as much as possible, so this is helping to balance the make/buy ratio or rather, the make/partner ratio.” In general, L’Oréal works with external partners and favours open innovation, allowing them to expand the team when needed whilst keeping a level of confidentiality which is so important to them especially because most of their business is about innovation. “They do not want to share their data with competitors, as they would then lose a competitive advantage. So when working with external suppliers, they have to be careful that their data is kept safe. One problem that can arise, when outsourcing, is that the external partner can improve their model by training it with your data. This is as serious as your data being stolen because you lose your competitive advantage, if the same external provider uses this model for your competitor.”

The team of students had some misgivings about L’Oréal’s initial briefing and the fact that they specifically asked for a decision tree, which the students found “simplistic” and “very rough in terms of make or buy.” After meeting the BeautyTech team and several start-ups, they scrapped their initial idea of making three decision trees for the different aspects of technology - hardware, software and people. Instead, they were able to present a prototype of an app, keeping the personalisation it offered by using relative weighting coefficients for all criteria. This makes the decision trees specifically adapted to each “make or buy'' situation and should allow L'Oréal to move to a model better suited to their needs without setting a particular ratio of “make to buy”. “We had not clarified our ‘make or buy’ criteria and ordered them in a decision tree, especially with a notion of weighting,” comments L'Oréal Data & Analytics CIO François Nguyen. “We don't have an app, like the students suggested, but we can now make a decision tree for each subject when we find ourselves in this ‘make to buy’ situation.”