Google Cloud Platform > Case Studies > King Uses Looker to Make Games More Fun

King Uses Looker to Make Games More Fun

Company Size

1,000+

Region

Europe

Country

Spain

Product

Looker

Tech Stack

EXASOL
Hadoop
LookML

Implementation Scale

Enterprise-wide Deployment

Impact Metrics

Customer Satisfaction
Productivity Improvements

Technology Category

Analytics & Modeling - Big Data Analytics
Analytics & Modeling - Real Time Analytics

Applicable Industries

Consumer Goods

Applicable Functions

Business Operation
Product Research & Development

Use Cases

Predictive Quality Analytics
Real-Time Location System (RTLS)

Services

Data Science Services
System Integration

About The Customer

King is a leading interactive entertainment company that produces popular mobile games such as Candy Crush, Farm Heroes, Pet Rescue, and Bubble Witch. The company currently offers more than 200 titles and games that are played around the world, with 314 million monthly active users as of Q2 2017. Their flagship games are designed to be “bite-sized”—meaning that you can play for a short time, and then pick up the game later—and they’re synchronized across platforms, so you can play anywhere, anytime and from any device. King's hub-and-spoke model is based on a centralized organization that provides data warehouse and business intelligence support to autonomous game studios. The studios each have their own data scientists for the statistical modeling, projections, and finer-grained analysis used by the Business Performance Managers (BPMs), who are like product managers for specific games and who help shape the games to produce the best possible player experience.

The Challenge

King, a leading interactive entertainment company, is in a fast-paced, competitive market where the key to success is making games that are consistently fun to play. The company has always relied on analytics to help developers create the perfect balance between simplicity and difficulty. However, they wanted to improve their ability to deliver the right insights to the right people at the right time. Before Looker, a jumble of tools used by individual studios mixed imperfectly with centralized BI services, so queries could produce inconsistent results. At a point where time was of the essence, BI complexity could make modifying an existing query a sluggish process. The teams at King realized it was time to begin searching for a BI solution to better suit their needs.

The Solution

King identified a need for two different products: a semantic layer that would give them a single source of truth and a common layer for interpreting the data, and a more agile exploration tool. Once the team discovered Looker, they realized that they had found both products in a single solution, with features that met all their criteria. The first step was a two-day Proof of Concept trial in the Barcelona studio, with one data scientist and one BPM. After the successful PoC, they moved onto a 20-license pilot and quickly expanded from there to 70 licenses. The company continues to roll out Looker to its employees on a regular basis. Looker’s flexible approach to analytics has boosted the efficiency of game development. Test games require quick insight into how players interact with the game, as the test may run for only a few days. The centralized teams push out a common definition for KPIs such as the number of players, then each team’s data scientists plug bits into the model that differentiate the game, focusing on economy balancing or measuring the first-time user experience.

Operational Impact

Empowered Business Users: The King BI team expected Looker to help the game studio data scientists and BPMs, but the value of Looker has become evident to a wide range of groups across the company.
A Return to Strategic Focus for Data Scientists: Because the studio teams embracing Looker get the analytics they need instantly through self-service dashboards, studio data scientists are less likely to be called on to write routine queries and instead can focus on the high-value work that only they can do.
Increased Efficiency for Centralized Engineering: The centralized data teams are still very busy, but now they focus on the accuracy, integrity, and quality of their data pipelines and data models written in LookML, Looker’s modeling language.
More Agile Game Development: Looker’s flexible approach to analytics has boosted the efficiency of game development. Test games require quick insight into how players interact with the game, as the test may run for only a few days.
A Single Source of Truth: Whether a team is using R, a custom interface through the Looker API or the Looker layer, having a single source of truth means that they’ll all share consistent answers.

Quantitative Benefit

Years of effort have been condensed into the first seven months using the tool.
Hundreds of reports users have built and scheduled themselves would have taken the centralized teams a year or more to make for them.