Analytics Reference Architectures
These reference architectures describe a variety of analytics use cases and implementations with different alternatives, enabling you to architect your own cloud solution so you can have full control and customization to fit your game design.
Azure PlayFab is a complete back-end platform for building, launching, and growing games. Learn more about PlayFab's out-of-the-box analytics solutions.
Analytics is a broad area with many use cases. First, consider which types of analytics are needed by your game. Here are some examples:
- Real Time Analytics - Commonly used for development, operational health, customer support, and launch-window monitoring.
- Direct Data Exploration - Commonly used to answer questions about user behavior below the surface level.
- Performance Metrics - Commonly used to assess the health of the business against established targets.
- Custom Reporting - Commonly used to build graphs and dashboards from custom events in your game.
Common components of a typical analytics pipeline include:
- Events - The moments captured by your game or services for later analysis.
- Event queues and ingestion - The service that receives the events and dispatches them.
- Storage - Where events are saved for use in reporting or exploration.
- Enrichment - Jobs that transform event data and generate metrics on a scheduled cadence.
- Dashboards - UI that surfaces event level and metrics data in a visual chart or graph
- Query Engine - Compute resources and query language used to explore the data.
- Import/Export - Services that move external data sets between storage location.
- Machine Learning - Enrchiment activities used to create predictive metrics from patterns in your data.
Here are several analytics use cases for you to explore:
Additional potential features
Once you have your analytics pipeline established, you can add things like Azure Machine Learning to supplement this data and learn more. Here is an example using a customer churn analysis project.