Case Study: Houseparty cuts a 40% growth drag and boosts invites up to 64% with Segment

Reducing a 40% drag on your business how Houseparty experiments

Houseparty, a group video‑chat app with tens of millions of users generating billions of events each month, faced the common but hard problem of becoming truly data‑driven at scale: collecting terabytes of event data, running rapid experiments, and avoiding costly product changes based on intuition. The team needed an architecture and process that made analysis fast, consistent, and trustworthy so product decisions wouldn’t drag growth.

They built a pipeline (Segment → Taplytics & Amazon Redshift → Periscope), standardized event schema, and ran SQL‑backed experiments with a weekly review cadence. That approach let them run 100+ tests last year, ship 60% and kill 40% that would have hurt the business, and produce clear wins—most notably increasing invites sent by 64% on iOS and 27% on Android—driving stronger engagement and growth.

Houseparty

Jeff Needles

Houseparty