Jelly Button Saves $240K a Year on Analytics

Jelly Button's existing analytics solution became prohibitively expensive as their games went viral, punishing exponential growth with exceptionally high costs. The system also suffered from monthly shutdowns and API problems that prevented data extraction, leaving the company unable to respond to gamer complaints, bugs, or alerts.

Working with DoIT International, Jelly Button implemented the Google Reference Architecture for Mobile Gaming Analytics. They built a new pipeline streaming data directly into Google BigQuery using Google Kubernetes Engine for autoscaling, Cloud Pub/Sub for low-latency event relay, and Cloud Dataflow for data processing. The modular design stores each game's data in separate BigQuery projects for simplified management.

A key part of Jelly Button's infrastructure is collecting, storing, and processing player data for analytics, which the company uses for research, marketing, and to improve gameplay. Jelly Button demands a reliable analytics backend that is very low latency, does not lose data, and makes information available for analysis quickly, even when millions of new events arrive every hour. But as Jelly Button became more successful, their widely used product analytics solution became expensive, punishing exponential growth with exceptionally high costs.

Using the Google Reference Architecture for Mobile Gaming Analytics as a baseline, Jelly Button began streaming data directly into Google BigQuery from its clients. Autoscaling the solution with Google Kubernetes Engine, the team set up Google Cloud Pub/Sub to relay events with low latency through to Google Cloud Dataflow, which filters, maps, and aggregates the raw data for Google BigQuery, making the data available for analysis in near real time. Because the data and analytics on each game is stored in a separate Google BigQuery project in a modular pipeline, managing the process is simple for administrators.

Licenses for Jelly Button's previous analytics service were expiring, so they had to work to a tight deadline. DoIT International provided its best talent to launch this project in just under five weeks from start to production, with weekly meetings to review progress. It's the latest in several successful initiatives with DoIT International, including cost optimization and a cloud security review.

Beyond speed and savings, the new pipeline offers greatly improved stability. In terms of ETL data streaming, Jelly Button now has the big advantage that they can keep all of their data within the company, instead of on a third party. In the past they had problems with APIs, which meant that they couldn't extract data and experienced days of shutdown without any communication. The new pipeline is much more stable because they have all the data within their own systems. They tested it by shutting down the process for 24 hours, then enabling it again. All the data was queuing within the accumulator and was recovered with no data loss whatsoever.

According to the Jelly Button team, the new Google based solution is more than two-thirds cheaper to run than the system it replaces, saving the company $240,000 per year. In a typical month, processing around 500 events per second, the company spent under $4,000 on Google Cloud Platform services. Using reOptimize, a free cost tracking and optimization tool, Jelly Button easily tracks and controls its Google Cloud Platform costs, keeping budgets controlled and predictable, as well as low.