Case Study: Bazaarvoice achieves 90% cost reduction and serverless scalability with MongoDB Atlas

Adopting a Serverless Approach at Bazaarvoice with MongoDB Atlas and AWS Lambda

Bazaarvoice, which powers user‑generated content like product reviews and social posts for many online retailers, replatformed its Curations service after scaling pain points became untenable. The original Python/Django monolith on EC2 with per‑customer MySQL clusters cost roughly $60,000/month and created heavy operational overhead as each new client required another cluster, making debugging, patching, and data management difficult even at massive traffic peaks.

The team rebuilt Curations as a serverless, microservice architecture backed by MongoDB Atlas—using AWS Lambda and Kinesis for collection and autoscaling Elastic Beanstalk for enrichment and display—leveraging Atlas’ managed operations and performance tools. The move cut monthly costs by about 90% to ~$6,500, delivered strong performance and consistency, reduced inbound support tickets by ~80%, and let the company onboard many more clients without a proportional increase in operations.

Bazaarvoice

Ani Hammond

Senior Staff Software Engineer