Case Study: HomeToGo achieves real-time personalization and faster search results with Snowplow

How HomeToGo Powers Real-Time Personalization and Improved Search Results

HomeToGo, the travel marketplace with millions of users and more than 20 million vacation rental offers, wanted to improve its ML-driven search and ranking with near real-time user behavior. To do that, it needed a way to capture timely event-level data without rebuilding its infrastructure, and it turned to Snowplow CDI to add real-time event tracking alongside its existing Snowflake-based batch setup.

Snowplow helped HomeToGo stream validated customer events into Tecton for real-time feature computation, enabling a unified workflow for both streaming and historical features. With Snowplow’s Event Forwarding (Snowbridge), HomeToGo reduced feature freshness to sub-second latency, deployed the system in just a few weeks with a two-person ML team, and supported 100,000 requests per second with latencies well below 100 ms.

HomeToGo

Stephan Claus

Director of Data Analytics