In time spent building campaigns
Bayard, a medium-sized French publishing company in both B2B (stores, kiosques, schools, etc.) and B2C (subscribers), has always been in the business of targeting and nurturing clients with the newest books and publications.
With its many titles and collections that target different audiences, including J’aime Lire for children and La Croix for adults, this was a largely time-consuming and complex process. Their data team of eight data analysts and scientists thus welcomed an important data transformation initiative with the arrival of Dataiku in 2021.
After switching to Dataiku, Bayard saw their full-time job application intake increase from 15 candidates (when using SAS) to over 200 candidates (when using Dataiku)!
With over 20 automated campaigns, teams reduced their time spent on building campaigns by 42% and therefore, were able to spend more time on high-value ML projects. Automation enables them to produce more and spend more time on higher-value projects.
Bayard’s data team has also seen a decrease in issues thanks to their ability to reuse work and an increase in reactivity and productivity thanks to alert messages.
Before Dataiku, the data team did not do any ML so this opened a whole new realm to them. Dataiku’s visualizations features really helped the team increase awareness of the need of governance and data quality. This led to the development of a new governance process.
Interested in learning how they got there? Read on for a detailed description for Bayard’s data journey.
Bayard’s data team, which groups all the company’s analysts and data scientists, is part of the central IT team. This means they easily diffuse best practices and ensure consistency, no matter what markets different analysts or data scientists might be targeting.
Four years ago, the team welcomed new leadership and, along with that, an all-encompassing modernization effort. The team migrated to the cloud and adopted new tools, including Dataiku, new recruitment strategies, and new machine learning (ML) topics, including studies, client analyses, and more.
Bayard decided it was time to modernize and meet market standards to remain competitive. What did this mean concretely? They needed to find a tool that was visual, multi-user, centralized, and easy to implement and adopt for technical teams. Indeed, Bayard’s team had certain profiles that knew only SQL, Python, or SAS. So they needed a tool that was not specialized in one of these, but could be used by all. After a market benchmark, the decision stood between open source and Dataiku — and Dataiku became the platform of choice.
With the Dataiku Academy, the data team upskilled smoothly and new members were onboarding easily. The first week of new joiners was entirely dedicated to product onboarding to ensure complete understanding of the tool.
Bayard saw an opportunity in this transition period to upgrade their whole database and move their data stack from Oracle to AWS. They also reviewed their entire database and created a new data engineering team.
This entire transformation journey lasted a year and a half, but the migration of projects onto Dataiku was very fast. All the more so with the hiring of a Dataiku lead who managed the project and held daily checks for the team to share best practices and new discoveries. They also had communication help channels to quickly support those who were blocked. Bayard also worked with a specialized consultant from Lincoln Group who helped with the migration.
Bayard made the change official in February 2023 — marking the end of SAS and the beginning of Dataiku. Despite teams feeling a bit slowed down by changing the automations at first, once sharing best practices became a habit, they all felt processes were a lot smoother and their worries faded. Moving to such a visual tool helped enormously in convincing the teams and removing any fears.
Everyone knows it’s going a lot better. We spend less time discussing data issues and nobody in the team has mentioned wanting to move backwards.Claire Utiel Data & CRM, Bayard
The team’s main Dataiku use case remains the usual targeting activities they had been engaging in before. However, Dataiku sped up these efforts with automation features. Bayard now automates targeting and emails as part of larger marketing campaigns. For example, Bayard has automations in place to remind subscribers of a payment method or subscription about to reach expiration.
Just as important as automation are the monitoring features enabled by Dataiku. The data team leverages Dataiku’s monitoring capabilities to ensure that all automations are going as planned. This includes setting up alert messages when emails are not delivered for example.
With Dataiku, the Bayard data team also benefits from simple visualizations that make it easy to share findings with teams. Bayard pulls their datasets from Dataiku and into Tableau to present visual and understandable dashboards. These initiatives give business leaders a lot more autonomy and transparency, and help them make more data-driven decisions.
With their newfound efficiency thanks to Dataiku, the team is able to take on a larger range of projects, dipping their toes into ML more than ever before. Additional use cases include customer understanding, ML studies, segmentation of customer portfolios, digital usage segmentation, affinity scores of different products, and cross sell.
In our team, our pillar is to share — nobody should be irreplaceable, and I believe Dataiku allows this.Claire Utiel Data & CRM, Bayard
Bayard has already seen notable improvements with the adoption of Dataiku, but their journey does not stop here. The data team is working on improving their data lineage with Dataiku and has plans to build a Dataiku app for recurring targeting that would optimize data analysts’ time on production.
In the spirit of democratization and transparency, Bayard is also hoping to increase their Dataiku licenses for data stewards or hybrid project manager profiles working in their data governance team. They would be able to explore datasets and build visual dashboards — something that would not have been possible on SAS.
The strength of this tool is that it creates a community. Before this, we were not sharing findings or best practices. Now, we have made a habit of and set clear communication channels both within our data team and outside our team, with the larger organization.Claire Utiel Data & CRM, Bayard
Each of Heraeus’s 20 operating companies has its own lead identification and qualification process. See how Heraeus uses Large Language Models (LLMs) in Dataiku to support these processes, ultimately saving time and increasing sales conversion.
Read moreOrange has accelerated their data science practice with Dataiku, building a call load detection and triage model in just one month, with more use cases on the horizon.
Learn MoreDiscover how LVMH centralized and customized deployment of AI algorithms for its luxury goods houses.
Learn MoreDataiku is GRDF’s handy toolbox for code management, collaboration, and more.
Learn MoreHear how Engie's IT team partners with Capgemini and Dataiku to support business teams across the company on their data journeys.
Learn More
