Case Study: ASOS achieves real-time personalization and scalable, low-latency order processing with Microsoft Azure

Online retailer uses cloud database to deliver world-class shopping experiences

ASOS, a London-based online fashion retailer serving more than 15 million customers with 85,000 SKUs, needed to modernize its platform to scale globally, speed innovation, and deliver personalized shopping and instant order updates—especially during massive seasonal peaks like Black Friday. The challenge was to support real-time product recommendations and reliable order workflows while handling heavy, bursty traffic.

ASOS migrated to a microservices architecture on Microsoft Azure, using Azure Cosmos DB as a globally distributed, low-latency data store alongside Azure Data Lake, HDInsight (Spark), Data Factory, Service Fabric, and Azure SQL Database. An offline/online recommendations pipeline precomputes models and serves millisecond product relevancy, while Cosmos DB change feed drives event-driven order and returns workflows. The platform handled 167 million requests in 24 hours on Black Friday (about 3,500 requests and 33 orders per second) with ~48 ms average response times, improved returns and tracking, and is moving parts of the stack to serverless functions to boost developer productivity.

ASOS

Bob Strudwick

Chief Technology Officer