1. Overview

Retention Analysis focuses on the number or proportion of users who continue to perform Event B within a specified time period after they have performed Event A. It is an important metric for measuring user loyalty and key conversions.

SolarEngine calculates the retention between events. For example, the retention from one "_appOpen" event to another can be seen as active user retention. You can view up to the 365-day retention.

Retention Analysis can be used to answer the following questions:

• How is the next-day retention for new users?
• What is the 7-day retention rate for users who launched the app today?
• How many users who played a certain level for two consecutive days made a purchase the following day?
• What is the number and proportion of users from different channels who purchased a newbie package within 14 days after registration?

2. Usage Guide

SolarEngine retention analysis looks like this:

2.1 Add Initial & Revisit Events

You are required to set an initial event and a return visit event, as Retention Analysis is to calculate user retention between the two. Users who trigger the initial event form the basis (denominator) for retention calculation, while users who trigger the return visit event represent the result (numerator).

Both the initial event and the revisit event support conditional filtering. Filtering the initial event helps filter the denominator, while filtering the return visit event helps filter the numerator.

Note: Only one initial event and one return visit event can be set.

2.2 Use Associated Properties

Use Associated Properties to specify one property for the initial event and another property for the return visit event. Only when a return visit event occurs and its property value matches the property value specified for the initial event, it is considered a revisit event. If a return visit event occurs but its property value differs from that of the initial event, it cannot participate in later calculation.

2.3 Calculate Another Metric

This feature allows you to compute and display additional metrics for the users involved in the initial or revisit events while analyzing retention.

For example, for users who open app on the next day, besides retention rate, you can enable "Calculate Another Metric" to simultaneously display their in-app purchases.

2.3.1 Usage Guide

When using "Calculate Another Metric", you need to specify the user scope for calculating the metric.

• When selecting "Initial User", only the metric of initial users will be calculated.
• When selecting "Returning User", only the metric of return visit users will be calculated.

The "Calculate Another Metric" feature supports composite metrics, which allows you to combine multiple metrics into one for analysis and display.

2.3.2 Calculation Methods for Events & Event Properties

Here are the different calculation methods for events and event properties.

• Cumulative Sum

       The sum of the event property value generated by initial users during the retention period.

       For example, if there are 100 initial users, with retention period as 3 days, "calculate another metric" as purchase, and property value as payment amount, then the cumulative sum should be the total payment amount of 100 initial users within 4 days of retention (D0-D3).

• Cumulative Per User

       Cumulative Sum/Initial Users. The average of the event property value generated by initial users during the retention period.

• Per User:

Sum of all event property values on the retention period / Number of retained users on the retention period

• Sum:

Sum of all event property values on the retention period

2.4 Conditional Filtering

       Initial events, revisit events, and calculate another metric all support data filtering. The selection of filtering options is the same as the event analysis model.

Please see Event Analysis—Conditional Filtering.

2.5 Group Display

     You can group the retention analysis results to categorize the data.

Please see Event Analysis—Group Display.

2.6 Result Display

After completing the conditions configuration in retention analysis, click on "Calculate ", the calculated results will be displayed in the result display area.

2.6.1 Rentention/Lost Period

Based on the selected time granularity, you can set different retention/lost periods.

Click "Edit" and enter number of days, separated by "," or "-". Please do not exceed 30 values at the same time.

Click "Finish" and the report will show days of retention based on your settings.

• If you set the time granularity as "By Day," you can set a maximum of 30 values for the retention/churn periods. The maximum value cannot exceed 365.
• If you set the time granularity as "week," you can set a maximum of 12 values for the retention/churn periods. The maximum value cannot exceed 52.
• If you set the time granularity as "month," you can set a maximum of 12 values for the retention/churn periods. The maximum value cannot exceed 12.

      Switch between "Retention" and "Lost" to switch between "Set Retention Period" and "Set Churn Period".

      In retention analysis, the term "lost/churn" refers to users who triggered the initial event but didn't complete the return visit event within the specified retention period. If a user triggered both initial and return visit events during this time period, they will be categorized into retained users rather than churned users.

2.6.2 Stage Summary

In retention analysis, you can simultaneously display up to three types of metrics: Number of Users, Retention Rate/Churn Rate), and Calculate another metric. You can configure the summary method for each metric type.

In the stage summary settings window, you can also configure whether to display the stage summary.

2.6.3 Retention/Churn for Specified Period

Create a retention/churn rate chart for a specific period where the results will be presented as a line graph, showing the maximum values set for each time granularity within the retention period.

2.6.4 Retention/Churn for All Periods

Create a retention/churn rate chart for the entire period where the results will be presented as a line graph, displaying the results of the same duration across different initial time points.

3. Permission Explanation

Operation Permissions: Super Admin, Group Admin, Analyst, Regular Sub-account with assigned permissions.

Entry: SolarEngine - Product - Analysis - Retention Analysis