Case Study: AppsFlyer achieves 30% cost reduction with Granulate

AppsFlyer Decreases Costs By 30% Leveraging Granulate For Spark and Kafka Consumers Clusters

AppsFlyer, a global attribution leader in mobile attribution and marketing analytics, was facing rising AWS compute costs as its business and infrastructure grew. To improve efficiency across its environment, AppsFlyer looked for a solution that could support fast scaling without requiring code changes or extra R&D effort, and selected Granulate to optimize its Spark and Kafka consumer clusters.

Using Granulate’s real-time continuous optimization, AppsFlyer achieved immediate performance improvements and lower resource usage across both clusters. The Spark cluster saw about a 30% CPU reduction and 33% cost reduction, while the Kafka consumers cluster saw a 25% CPU reduction and 30% cost reduction, contributing to an estimated annual savings of about $320,000.

AppsFlyer

Eli Zilbershtein

Platform Group Team Leader