Case Study: adidas turns customer reviews into actionable insights at scale with Databricks

Driving product innovation through customer feedback

adidas, the iconic German sports brand, needed a faster and smarter way to analyze millions of customer product reviews across 150+ countries and turn them into actionable sentiment and product insights. To solve this, adidas partnered with Databricks and used a GenAI approach with retrieval augmented generation (RAG) to scale review analysis and support product innovation.

With Databricks, adidas built a governed, global-scale solution using Mosaic AI Vector Search, Unity Catalog, Model Serving, and MLflow. The result was 30–40% efficiency gains for analysts, 60% lower latency, and over 90% lower compute costs, while reducing token input size by 98.5% and enabling faster, self-service insights for teams across design, product, marketing, and customer service.

adidas

Rahul Pandey

Senior Solutions Architect