Case Study: UrbanSitter achieves 1-second search and scalable, cost-effective infrastructure with Google Cloud Platform

UrbanSitter Helping parents quickly find trusted sitters

UrbanSitter, founded in 2010 in San Francisco, is a web and mobile marketplace that helps parents find babysitters and nannies by surfacing caregiver recommendations from social connections (Facebook, local parenting groups, schools). As its user social graphs grew and traffic rose about 50% year over year, complex searches and an on‑premises, multi‑provider server setup caused slow search times, frequent outages, and costly overprovisioning.

UrbanSitter migrated to Google Cloud Platform—running Compute Engine and Kubernetes Engine, and using Stackdriver, BigQuery, and Cloud Dataflow—leveraging Google Cloud Startup credits to experiment and move to microservices. Isolating search on Kubernetes reduced profile search time from over 30 seconds to under 1 second, enabled fast, cost‑effective scaling for traffic spikes, cut outage frequency (from ~2/month to ~1 every 3 months), and freed engineers to build new features.

Urbansitter

Andrea Barrett

Co-founder and Vice President of Product