Case Study: Razorfish achieves scalable, cost-effective clickstream processing and 500% return on ad spend with Amazon Web Services

Razorfish - Customer Case Study

Razorfish is a digital advertising and marketing firm that builds customer segments from large click‑stream data to drive targeted advertising and cross‑selling. Their challenge was scaling batch data‑mining across seasonal and sudden traffic spikes—processing datasets from gigabytes to terabytes—without investing in costly headroom, facing a potential $500k hardware bill, long procurement times, extra staff, and the risk that daily processing would slip well past SLA limits.

Razorfish moved its pipeline to AWS, using Amazon Elastic MapReduce with Cascading integrated into its ad‑serving systems to run existing algorithms at scale without infrastructure overhead. The elastic, managed environment eliminated upfront hardware and extra operations hires, enabled a six‑week rollout for the first client, automated processing for about $13,000/month, and helped deliver a 500% increase in return on ad spend for that campaign.

Razorfish

Mark Taylor

Program Director, Razorfish