Powering insight-driven dashboards to increase customer acquisition
Average annual growth rate
Increase in data volumes
New data dashboards created
Bagelcode is a global game company with more than 50 million users across 200+ countries around the world. The organization has effectively managed and improved the growth indicators of games —including daily active users, cumulative revenues and churn rates — through sophisticated data analysis work. However, its data platforms comprising Kinesis Firehose and AWS Redshift often became overloaded as more diverse data indicators were required and the number of gamers increased — 50 million registered users to date, and growing. In response, Bagelcode introduced the Databricks Data Intelligence Platform to support virtually all corporate decision-making, using more than 800 automated data dashboards. For example, they are now able to support more diversified indicators, such as a user’s level of frequency and the amount of time they use a specific function for each game, enabling more well-informed responses. In addition, the company is driving high growth through various efforts, including preventing customer churn by predicting game user behavior.
Need for diversified understanding and prediction of game users
Social casino games have continued to show stable growth in response to the new demands of users who want games that provide entertainment and relaxation with widespread internet access and rapid digitization. While most off-line activities were restricted by the COVID-19 pandemic , the global social casino market recorded explosive growth, and is heavily contributing to the expansion of the entire game market. Bagelcode has been responding by providing various social casino games, including Club Vegas.
Bagelcode focused on social casinos because of the continuous growth potential of the market and the judgment that such a move could increase both consumer response and corporate growth based on data, unlike other games. In other words, due to the limited interaction options of users, Bagelcode was certain that it could achieve both the consumer response it wanted and sales growth if it could accurately predict the behavioral patterns of game users based on data and deal with them adequately. To put it another way, Bagelcode concluded that social games are similar to science, where the validation of results predicted with data is possible.
Based on such a conclusion, Bagelcode focused its efforts on the thorough monitoring and managing of game stability based on data analytics indicators such as DAU, MAU, sales per capita, total sales per day and time-of-day. And based on the results drawn from such efforts, it made business decisions about things such as whether to launch an advertising campaign, driving measurable customer acquisition and revenue growth.
Then, the employees of Bagelcode started to ask for more data indicators to better understand and predict game user behaviors. Incidentally, Club Vegas was released around this time. Not only that, the company launched its own publishing service, thus significantly expanding its user base through aggressive marketing activities. In the midst of such changes, however, its existing data platforms turned out to be insufficient to meet new demands because of limited scalability. So Bagelcode determined that it needed a new, highly scalable data infrastructure that guarantees superior scalability by separating the compute layer from the storage layer.
Joohyun Kim, Vice President of Data & AI at Bagelcode, said, “For Bagelcode, data is the key asset for its growth. In other words, data plays a crucial role in solving Bagelcode’s business challenges while supporting all of its business decision-making processes. As Bagelcode is aware of the importance of well-refined data to unlock its values as meaningful data indicators, we are investing heavily in our data infrastructure.”
Data-driven business decision-making of all sizes: predicting game user behavior patterns
Bagelcode chose the Databricks Data Intelligence Platform on AWS as its new data infrastructure because it has excellent scalability and flexibility while supporting various tasks such as data engineering, analytics, BI, data science and machine learning. In fact, as the number of serviced games increased from one to six and the number of data dashboards supporting business decisions increased from five to 800, the amount of data that had to be processed increased by nearly 1,000-fold, from a terabyte to petabyte scale. Databricks nonetheless managed to ensure a stable data infrastructure. This is remarkable, because the frequency of data pipeline updates remains unchanged and keeps the daily and hourly schedule.
The Databricks Data Intelligence Platform also helped to support a variety of game user behavior prediction models that drive significant growth of Bagelcode by preventing customer churn and maximizing sales. The company’s important achievements include the Churn Prediction, which enables intensive care by predicting the potential churn of existing users; the Payer Prediction, which identifies the new users who are likely to purchase; and the Lifetime Value Prediction, which shows each user’s game lifecycle.
Joohyun Kim said, “Thanks to the Databricks Data Intelligence Platform, we were able to unlock value from data in spite of an almost 1,000-fold data increase. In particular, not only did the time and effort required for communication between data experts significantly decrease, but it also enabled them to utilize their expertise in their respective areas, uncovering new ways to unlock data values.”
Driving high revenue growth by utilizing diversified data indicators
Bagelcode has achieved a prominent 100% average annual growth rate since it innovated its data infrastructure based on the Databricks Data Intelligence Platform. This was possible because more than 800 data dashboards are automatically updated daily and hourly, thereby enabling data-driven decision-making of all sizes in game development and its operation, which in turn increased the success rate. And because custom customer care programs are automatically performed based on various prediction solutions, users were able to enjoy Bagelcode games for a longer time, as well as more frequently. In addition, various data obtained during the game operation process can be linked to detect anomalies proactively so that countermeasures can be prepared, and it is being utilized as meaningful statistical data by continuously collecting and comparing A/B test results.
The importance of making a sound decision based on accurate data analysis has never been greater. This is especially true in the gaming segment, including social casinos. However, Bagelcode is moving toward near real-time data analysis based on its experience and the know-how it has gained with Databricks. In addition, the company plans to support faster and more sophisticated corporate decision-making through more diverse data projects.
