Creating a Retention Report

The Retention Report analyzes retention rates and the number of retained users who return to your app after a set period from their first visit. Based on the granularity you select, you can view retention rates and the number of retained users by interval for a specific event.

Common use cases

  • Track Day 1, Day 7, and Day 30 retention trends for newly installed users

  • Compare retention rates across ad channels to validate campaign performance

  • Analyze retention patterns at the cohort level, such as users acquired through a specific campaign

  • Estimate Predictive Lifetime (pLTV) for acquired users

Configure the Retention Report

Configure Cohort, Granularity, Date Range, Retention Type, Start Event, Return Event, GroupBy, and Filter to view the data you need.

Cohort

A cohort is the user segment you want to analyze in the Retention Report. In [Cohort], click 'Add' to set a cohort. Define cohort conditions for the Retention Report based on event execution, event type, event execution count, and more. If you don't set a cohort, the report aggregates results for all users.

Granularity

Granularity is the time unit by which retention rates and the number of retained users are aggregated. Retention Report results are divided into intervals based on the selected granularity. Granularity follows the app timezone.

Date Range

The Date Range defines the period you view in the Retention Report. The report aggregates data for users who generated a Start Event during this range. Date Range follows the app timezone. Select a date range condition and dates to set the range. The maximum date range varies by granularity.

  • Minutely: 3 days

  • Hourly: 7 days

  • Daily: 184 days

  • Weekly: 12 weeks

  • Monthly: 3 months

Important

Under privacy policies, the following data from SAN (Self-Attributing Network) channels appears as Unattributed. See FAQ - Actuals Report for details.

  • Major SAN channels: Google Ads (google.adwords), Meta Ads (facebook.business), TikTok for Business (tiktok)

  • Affected data: data older than 180 days from the query date

Date range conditions

Three date range conditions are available.

For 'Last', you can add 'Include Today', 'Include this Month', or 'Offset'. Including today or this month extends the range from n days ago to today or this month. Click 'Offset' and enter a number to shift the range backward by that amount. For example, with granularity Daily and 'Last 30 days', the range covers 30 days ago to 1 day ago. Turn off 'Include Today' and add 'Offset 3 days' to shift the range to 33 days ago through 4 days ago.

Retention Type

Retention Type determines how Airbridge counts users who returned to the service in each interval.

Start Event and Return Event

The Retention Report aggregates retention rates and the number of retained users by checking whether users who generated a Start Event also generated a Return Event.

  • Start Event: An event generated when a user enters the app during the configured date range

  • Return Event: An event used to determine retention for users who generated a Start Event

Start Event

A Start Event is generated when a user enters the app during the configured date range. Users who don't generate a Start Event are excluded from analysis. Available Start Events vary by granularity. You can select multiple Start Events. For example, if you select Install (App) and Deeplink Open (App) as Start Events, only users who install the app or open a deep link are included in the Retention Report.

Return Event

A Return Event determines retention for users who generated a Start Event. When a user generates a Return Event, the Retention Report treats that user as having returned to the app. Users who don't generate a Return Event or generate an Uninstall (App) event are treated as not retained. Return Events can be standard events or custom events collected at least once. Selecting Any Event (App) as the Return Event lets you view retention for users who generated any event other than Uninstall (App).

You can select multiple Return Events. For example, if you select Add to Cart (App) and View Search Result (App) as Return Events, only users who add items to the cart or view a search result are included in the retention aggregation. The maximum number of intervals returned varies by granularity. When granularity is Minutely or Hourly, the Return Event window is fixed at 60 minutes and 24 hours.

Predictive Lifetime

Predictive Lifetime estimates how often a cohort will return to your service in the future. Use Predictive Lifetime to calculate Predictive LTV (pLTV). For details, see Predictive Lifetime.

Measurement Option

The Measurement Option determines how many Start Events count toward Return Event measurement. When a user journey contains multiple Start Events, the Measurement Option changes the number of retained users and retention rate.

GroupBy

A GroupBy divides metrics by a chosen dimension. The Retention Report breaks down metrics based on the selected GroupBy. For example, select Channel to view metrics by ad channel. You can select up to 6 GroupBys. GroupBy applies to Start Events.

Filter

A Filter narrows the report to specific GroupBy values. Filters use the format A is B, where A is a GroupBy and B is the value to filter on. Only values collected at least once in Airbridge appear in the value search results. 'is' is the default condition. All available conditions are listed below.

View the Retention Report

Once configuration is complete, view your results as a chart and a table.

Chart

Review interval values for the selected metric in a line chart. Hover over empty space on the chart to see all values for that interval by item. Hover over a single line to see only that item's value. Click an item in the legend below the chart to show or hide it.

Table

Review interval values for the selected metric in a table.

Top rows and sub-rows

The table consists of top rows and sub-rows. Also see Retention Report calculation method.

  • Top row: Shows retention data for the entire date range. Top rows extend from the configured start date through the day you view the report, up to the maximum number of intervals.

  • Sub-row: Shows retention data broken down by granularity.

Top rows and sub-rows share the same maximum number of intervals. Depending on when you view the report, top rows may provide more intervals than sub-rows. Some intervals may also be shorter than the standard interval unit. When granularity is Minutely or Hourly, sub-rows are not provided. Airbridge provides only the aggregated total of sub-row data instead.

Retention rate and the number of retained users

View Retention Report results in the table. Each cell shows the retention rate and the number of retained users for a specific interval after the user generated the Start Event. For example, the Day 3 retention rate is the rate 3 days after Day 0, when the user generated the Start Event.

Click the expand (>) icon on the far left of the table to view retention rates and the number of retained users by date. The expand icon changes to a collapse (⌵) icon. When granularity is Minutely or Hourly, the expand icon is hidden because date-level breakdowns are not available.

Retention rate formula

  • Top-row retention rate (%) = Σ {(sub-row retention rate) × (weight of users who generated the Start Event in each sub-row)}

  • Sub-row retention rate by interval (%) = (number of retained users in the interval of each sub-row) ÷ (number of Start Events in each sub-row) × 100

Number of retained users

  • Sub-row retained users: The number of users who generated a Return Event among those who generated the Start Event in that interval.

  • Top-row retained users: The number of unique users who generated a Return Event across all sub-rows. As a result, the top-row count is less than or equal to the sum of the sub-row counts.

Predictive Lifetime

View Predictive Lifetime in the Predictive Lifetime column.

Share the Retention Report

Share your results using the following features.

A Sharelink is a URL for downloading a CSV file with Actuals Report data. Generate a Sharelink by clicking the share button to the right of the report name. Sharelinks are only available in [Saved Reports]. For details, see Share retention data in real time via Sharelink.

Specifications and limitations

Specifications and limitations of the Retention Report.

Supported features

The following features are common across reports. Some features are not available in every report.

Important considerations

Keep the following in mind.

Meta Privacy Policy

Under Meta's privacy policy, some Meta Ads (facebook.business) data viewed during the configured date range is masked if you have not agreed to the AMM (Advanced Mobile Measurement) terms. Masking applies under these conditions.

  • The sum of Impressions and Engaged Views generated by the Meta Ads campaign is less than 1,000

  • The data is provided through Meta Ads channel or cost integration

Masked data is excluded from aggregation in the report and displays as follows.

Note

To view all Meta Ads campaign data in reports, agree to Meta's AMM Terms. For details, see FAQ - Meta Business.

If Meta Ads data still appears masked after agreeing to the terms, try the following.

  • Change or extend the date range. Only data generated after agreement is provided in original form.

  • Change or remove some GroupBys and Filters.

Date-level data not available for some granularities

Whether GroupBy results are available at the date level depends on granularity. When granularity is Minutely or Hourly, date-level data is not available in GroupBy results.

Archived components (unused after migration)

Data available in the Retention Report

The Retention Report only shows App event data; Web and iOS SKAdNetwork event data are not available.

The Open (App) event performed on 2024-03-05 is attributed to both Channel X and Channel Y.

  • The Open (App) event is measured as the Return Event of Day 4 of the user journey initiated by the Start Event on 2024-03-01, which is attributed to Channel X.

  • The Open (App) event is also measured as the Return Event of Day 2 of the user journey initiated by the Start Event on 2024-03-03, which is attributed to Channel Y.

Refer to the table below.

 

The Open (App) event performed on 2024-03-05 is attributed only to Channel Y. The user journey initiated by the Stat Event that occurred on 2024-03-01 is no longer tracked.

  • The Open (App) event is not attributed to Channel X.

  • The Open (App) event is measured as the Return Event of Day 2 of the user journey initiated by the Start Event on 2024-03-03, which is attributed to Channel Y.

Refer to the table below.

 

Attention

All values displayed in the Revenue Report with the GroupBy and Filter options applied are based on users who performed the Start Event.

Note

Starting February 4, 2025, the Overview Dashboard, Actuals Report, Trend Report, Retention Report, Revenue Report, Active Users Report, and Audience Manager will only display events and metrics measurable using the collected data.