Case Study: Publicis Groupe achieves up to 50% campaign revenue lift and 5x faster data processing with Databricks

Delivering revenue-generating experiences with data and ML

Publicis Groupe, a global communications and marketing company serving retail clients, faced major data and collaboration challenges that limited its ability to deliver personalized omni-channel experiences. Massive, inconsistent datasets and multiple HDInsight clusters led to slow analytics and fragile ETL, while tools like Jupyter hindered code sharing and cross-team reuse—blocking timely, consistent personalization at scale.

By adopting Azure Databricks and Delta Lake, Publicis unified ETL, analytics, and ML into a single, autoscaling platform with interactive notebooks, enabling real-time activation and processing of over 2.5 billion transactions. Results included 45–50% campaign revenue uplifts for clients, a 5x speedup in processing (36 to 5 hours), a 22% reduction in pipeline operating costs, ML cost cuts (from ~$5,000 every 2–3 days to $800/month), and roughly a 30% productivity improvement across data teams.

Publicis Groupe

Krish Kuruppath

Senior Vice President