Building a Customer Churn Analysis for Data Protection Software Company

A leading software company specializing in data backup and recovery solutions was losing an estimated $7M annually due to customer churn. Initial analysis showed that half of the churn was preventable. Although the company maintained steady data streams of real-time events, including telemetry data and customer usage metrics, the analysis of this data was only conducted once a quarter through batch processes. The infrequent processing of data meant many churn events were either missed or identified too late for effective intervention.

To address this challenge, the Aimpoint Digital team deployed Databricks as a central data platform, integrating all the company’s systems into a single source of truth. Our team merged customer data streams to create a comprehensive, cohesive dataset and incorporated a live stream of telemetry data to ensure real-time insights. We also implemented daily ETL (Extract, Transform, Load) processes for ERP and customer data to keep data updated regularly.

By combining these elements, we created dynamic data tables capable of maintaining consistently up-to-date information.  

Our Decision Sciences team then developed a predictive model to ensure the model remained accurate and relevant by continuously retraining and updating. The team also created real-time alerts to automatically notify key decision-makers when a customer showed potential signs of churning.

‍