Hong Kong Oasis Game Network Technology Co., Ltd. was founded in November 2011. It is one of the largest overseas distributors of online games in China. Its products are available in more than 130 countries and regions around the world, covering over 200 million users. The company’s headquarters is located in Beijing, and it has branches in Turkey, Brazil and other places.

industry

game

Case background and challenges

    The composition of Oasis game data is divided into four levels: The first level is the data from forums and media, which is data outside the game; the second level is operational data; the third level is players’ behavioral data. Whether players log in, chat, release skills, etc., all of these will be recorded in the form of logs, and the amount of this data is extremely terrifying. The fourth level is regular business data. In terms of management, collection, and processing, these data are all pain points.

The figure below shows that BigQuery is used as the single data source for analysis and regularly receives data from various sources. In Looker, Looker Blocks are combined with custom first – party LookML models to analyze data from different users.

The architecture is as shown in the figure below:

Description of the solution architecture

      In terms of data collection, BigQuery provides a highly scalable and cost-effective option that enables us to capture and analyze data in real-time, including game runtime error logs, player behavior, and data on topics outside the game. Using Memorystore for Redis managed instances, it is easy to manage and maintain a sorted list of scores while providing the uniqueness of elements. Player profiles are another piece of information that may be frequently accessed. Redis hashes can be used to quickly and easily store and access profile data. Google Cloud Looker is used to achieve data visualization. By building a live information center on this platform, the difficulty of data analysis is greatly reduced, and goal tracking is presented more intuitively. To better prioritize error logs, in order to handle, resolve, and minimize customer interruptions, technical project managers, product managers, and developers need to measure the impact of issues in their applications. This saves a lot of manpower and costs for data management.

Use the product

• CLoud BigQuery
• Looker
• Cloud SQL
• CLoud Memorystore

Features of the plan

    In terms of data collection, Bigquery can integrate data from multiple aspects, including game runtime error logs, player behavior, and data on topics outside the game. Looker can monitor and analyze data in real-time and in batches, greatly reducing the difficulty of data analysis. It presents goal tracking more intuitively. It can better confirm the priority of errors. In order to prioritize handling, resolving, and minimizing customer interruptions, technical project managers, product managers, and developers need to measure the impact of problems in their applications. Metrics that help address this use case include crash issues encountered by high-value customers and issues triggered during chat support.

Customer benefits

    In terms of data collection, the client uses BigQuery for cross – data – source analysis and management, running queries that process terabytes of data within seconds. In terms of data visualization, the client can use Looker to visualize data, presenting the analysis results of various data intuitively, enabling the client to have a thorough understanding of the data situation and manage it more easily. In this way, the client can keep abreast of the latest needs of game users at any time, and thus quickly adjust their product and operation strategies. Through wise business decisions and strategies to continuously improve the player’s gaming experience, the user churn rate is reduced by 5%, and the lifetime value output of users indirectly increases the revenue of the product department of “Oasis Games”.