Skip to main content

Events Analysis

Last updated 06/25/2026

Events Analysis is a powerful and flexible model. With simple operations, you can filter or group event data, calculate aggregated metrics for specific user behaviors over a period of time, and display the results in a variety of chart types.

With Events Analysis, you can quickly answer questions such as:

  • How have the daily numbers of players and devices that logged in changed over the past 30 days?
  • What is the payment rate of new users on the day they register?
  • What is the trend of the average payment amount per user for users from Shanghai over the last month?

Quickly configure analysis metrics​

In the Events Analysis model, you first select the specific behavior to analyze. You can select preset events, tracked events, or custom events.

After you select an event, the total number of times the event occurred is calculated by default, that is, how many times a specific event (such as payment or app launch) occurred, and you can view the event's detailed information in event details. You can also select other preset calculation methods, such as uniques or times per user, or calculate based on an event property. To learn the logic of the calculation methods, see Calculation method logic.

Events Analysis also lets you use Formulas to get ratio metrics, such as payment rate and ARPU, with support for the four basic arithmetic operations. You can also switch a metric to a custom formula, or switch a custom formula back to a metric.

If you want to exclude unwanted data, or select events or users with certain characteristics for analysis, such as counting only payments from the App Store, you can filter data.

If you want to compare the results of different source channels, such as the App Store, Huawei AppGallery, and Tencent MyApp, you can compare them with group-by.

After you finish the configuration, click Calculate to see the results.

Visualize results​

Events Analysis offers multiple visualization options, so you can choose the most suitable chart type to analyze the results. If you prefer to view specific values in a table, you can also customize the table display.

  • Line

When the time granularity you select is not Total, Events Analysis displays a line chart by default to help you understand how metrics change over time.

The performance of each analysis metric under each group value corresponds to a line on the chart. You can also switch to a bar chart or a stacked bar chart. If the result data of the analysis metrics differ greatly in magnitude, you can place metrics of different magnitudes on the main axis and the secondary axis to get the best display.

  • Cumulative

When the time granularity you select is not Total, in addition to the line chart, you can use the cumulative chart to see how metrics accumulate over time. Each point is the sum of all data up to and including that time.

Note: If the calculation method you select is uniques, distinct count, maximum, percentile, or a similar method, the cumulative chart may not be suitable.

  • Stacked Line

If you have configured multiple analysis metrics or there are multiple group values, you can switch to the stacked chart.

  • Distribution

When the time granularity you select is Total, the AE system displays the distribution chart by default to help you visually compare the results of different metrics or group values within the selected time range.

If the time granularity you select is not Total, you can also switch to the distribution chart. In this case, the data on the chart is the simple sum of the data for each date, which is Sum in the table's overview settings.

  • Pie

If you want to visually compare the shares of different group values, you can switch to Pie.

Advanced Events Analysis features​

In addition to the features above, you can use advanced features to answer more complex questions:

Was this page helpful?
On this page