Case Study: Similarweb achieves sub-second analytics over 1 trillion rows with Firebolt

How Similarweb uses Firebolt to deliver sub-second analytics over more than 1 trillion rows

Similarweb, a company that provides detailed analytics on website traffic and user behavior, faced a significant technical challenge. They needed to power a complex "Segment Analysis" feature for their customer dashboards, which required running dynamic, ad-hoc queries over a massive dataset of more than one trillion rows stored in their AWS data lake. Their existing solutions, including Presto, were unable to deliver the required sub-second query performance needed for a responsive user experience without costly and time-intensive pre-processing.

To solve this, Similarweb implemented the Firebolt cloud data warehouse. The Firebolt solution required no additional data pre-processing, allowing raw data to be loaded and immediately queried. This integration enabled Similarweb to deliver sub-second analytics at scale, with some UI features loading in approximately one second even while scanning terabytes of data. Firebolt's architecture also provided workload isolation for consistent performance and offered a superior price-performance ratio, allowing Similarweb to deploy this complex feature rapidly and cost-effectively.

Similarweb

Yoav Shmaria

Vice President R&D, Platform