Implementing Consent Mode at L’Oréal, a solution to reconcile customer experience and privacy

Since 2023, the L’Oréal group has been deploying Google’s Consent Mode across all its websites worldwide. Thomas Bianconi, Head of Data Technologies at L’Oréal, Alexis Vuillermoz, Technical Program Manager for Global Client Solutions at Google, and Baptiste Laurenge, Lead Analytics & Conversion at Converteo, collectively reflect on this ambitious project.

What is Consent Mode, and how does it help address advertisers’ challenges while respecting users’ privacy?

Alexis Vuillermoz, Google: Google Consent Mode allows Google tag users to communicate users’ cookie or app identifier consent status to Google. Our solution receives users’ choices from consent banners or widgets and dynamically adjusts the behavior of our products, such as Analytics and Ads, according to these preferences.

Two implementation options are available: 

• Basic Consent Mode, which prevents Google tags from loading until a user interacts with a consent banner and only transmits data if the user consents. This setup transmits no data to Google prior to user interaction with the consent banner and when the user doesn’t consent, no data is transferred to Google at all – not even the consent status (Google tags are completely blocked from firing).
• Advanced Consent Mode, loading Google tag when a user opens the website or app and sending consent state and cookieless pings when consent is denied.

The advanced mode enables improved modeling compared to the Basic one as it provides an advertiser-specific model as opposed to a general model. As a result, our clients can benefit from more accurate probabilistic reporting for opt-out users, enabling them to continue optimizing user journeys through more precise conversion modeling.

Baptiste Laurenge, Converteo: Indeed, Consent Mode addresses two major market trends: increasingly stringent regulations regarding user privacy and the decreasing availability of consented and observable data in media and analytics platforms. On average, we observe a 30% reduction in data among our clients. 

In this context, Consent Mode allows for modulation of what is or isn’t collected according to the user’s choices, transmitting only the data processing purposes to which they have consented. It compensates for the loss of observable data with modeled data, all while ensuring compliance with regulations like the GDPR and the DMA.

Why did L’Oréal choose to integrate this solution in 2023?

Thomas Bianconi, L’Oréal: At L’Oréal, our goal is to bring the very best in beauty to every person, everywhere in the world. We also believe that beauty is a personal concept, unique to each individual. Our websites and content must therefore reflect this vision through personalization and a consumer-centric approach. To offer this experience, we need to collect data on usage and users, all within a framework that respects their privacy.

Google’s Consent Mode, combined with Google Tag Manager Server, now enables us to follow the recommendations of European privacy authorities. This ensures that we never collect personal data from our opt-out users. Furthermore, with the Digital Markets Act (DMA) coming into force in Europe in 2024, enabling Consent Mode has become a prerequisite for using Google’s tools, and therefore for enhancing our users’ experience.

Baptiste Laurenge: Solutions like Google Tag Manager Server tagging allow us to control very precisely which data is transmitted to third-party platforms. For example, in the classic “client-side” collection approach we’ve used for several years, the user’s IP address is necessarily associated with any pixel sent to a third party. However, with “server-side” collection, it’s possible to systematically remove this type of sensitive data. Advertisers now have the upper hand in this area.

This is one of the reasons why so many clients, like L’Oréal, are opting to deploy advanced Consent Mode while equipping themselves with server-side solutions. This prepares them for the end of third-party cookies and ensures that their media tracking solutions comply with regulatory requirements.

L’Oréal group has a worldwide presence, with numerous brands and different markets: what challenges were involved in implementing this solution?

Thomas Bianconi: From a technical or organizational standpoint, there were no particular complexities: this deployment posed the same challenges as other global projects. We were able to rely on the quality of our existing infrastructures, and our technical stack was ready for implementation. To ensure a successful global rollout of Consent Mode, we proactively engaged with our DPOs, providing comprehensive training and resources on its features and compliance aspects. We adopted a phased approach, starting with the basic Consent Mode in regions where further adaptation was needed. This allowed us to address any questions thoroughly and tailor the implementation process to each region’s specific requirements.

Baptiste Laurenge: To successfully educate stakeholders, it was important to proceed gradually, with the implementation of POCs and pilots on half a dozen sites in different countries. By running tests in parallel with the existing system to compare observed and modeled data, L’Oréal was able to demonstrate the solution’s benefits—both in terms of privacy, by showing that no unnecessary data passed through the solution, and in terms of modeling efficiency. With this concrete proof, we were then able to carry out a first wave of implementations in the most receptive markets.

Alexis Vuillermoz: While this approach of POCs and pilots with certain countries may seem fairly standard, in L’Oréal’s case, it was highly innovative and above all, highly scientific. The teams sought to compare very precisely the benefits of our solution for an identical scope. Being able to provide “proof by the numbers” was decisive, and there was extensive discussion on the subject.

Looking back, what were the success factors of this deployment?

Thomas Bianconi: In all our markets, there were very high expectations regarding data privacy, but also concerning the quality of the user experience, with the need to regain complete visibility of how our websites are used. We therefore needed to provide a data collection solution that would reconcile both aspects while proving its effectiveness. Adapting to changing regulations, thanks to GTM Server advanced Consent Mode, means regaining a form of simplicity and data quality.

Baptiste Laurenge: The most important factor was demonstrating the reliability of the modeling and building confidence in the solution. We are transitioning from a world where we relied on 100% observed data to one that relies on probabilistic models. It was crucial to show that statistical methods and algorithms based on real data can provide high-quality insights that are as valuable as ever.

Alexis Vuillermoz: On our side, collaboration and transparency were key. It was important for us to support a player like L’Oréal in deploying the solution, explaining both theory and practice, and providing means to verify the quality of the solution. The involvement of various stakeholders—notably legal teams and DPOs—was also crucial.