Retention analytics is a powerful tool for understanding user engagement and long-term user behavior in your product or application.
Let's learn how to create a report!
Watch the summarized video or continue reading this article.
1 » Click on the Create Report
2 » Select Your Start Action:
The start action is the action you’d like to use as the foundation of your retention report. The start action is usually a basic action that indicates initial interest in your product. Common start actions include:
- User Registration
- The first interaction with your app (First visit)
- Specific actions: Trying out a new feature.
Choose between events, screens, first visit, or app launch (equal to any action) as your start action.
This step helps define when the retention analysis should begin. You can choose one or more start actions for your analysis and also specify a minimum count of occurrences.
3 » Define the size of your cohorts with 'Group by':
Cohorts are defined by grouping users based on the period in which they performed the start action: daily, weekly, monthly, or quarterly. This grouping creates cohorts of users who initiated the start action within the same time frame. For example, if you choose to group users based on a weekly interval, this means that users who initiated the start action in the same week will be grouped together into a specific cohort.
By defining these cohorts, you can analyze how different sets of users, based on the time of their initial interactions, exhibit variations in retention over time.
4 » Select Your Return Action:
The return action is the action that you want to see your users repeat over time. The return action is essentially the action that best defines that a user is getting value from your product. It could be completing a purchase, playing a movie, or creating a report.
Choose the return action that aligns with your product's goals. Choose between events, screens, or app launch. As with the start action, you can choose multiple return actions, and define a minimum count.
✏️ For multiple return actions, the count means adding up occurrences of any combination of selected actions. E.g., if two distinct actions are selected, and the count is set as “at least 2 times”, the results could include two occurrences of one action or one occurrence of each individual action.
5 » Set the Return Interval:
Determine how often you expect users to perform the return action after their initial interaction (start action). It can be set to each day, week, month, or quarter. The expected interval depends on your product and user behavior.
For example, if you expect users to complete a purchase every week, you should choose each week as a return interval; if you expect your users to log in every day, then you should select each day as your return interval.
Interval periods are, by default, calculated based on rolling periods (e.g. 7 days weeks, 24 hours days), but you can choose to use calendar periods instead.
Choose how retention is measured (Bounded vs Unbounded retention):
- In (Bounded): If you select "In," you are interested in analyzing retention exactly in a specific time frame (day, week, month, quarter) For example, if you chose each week, this means you want to know exactly how many users perform the return action each single week after the start action. You can see how many users came back on week 1, on week 2, on week 3, and so on.
- In or After (Unbounded): Calculates users who return on a specific time unit (e.g., day, week, month) or any time afterward. In or After Retention provides valuable insights into the long-term user engagement with your product, indicating how many users initially used your app and returned to derive more value.
Retention analysis typically focuses on users returning eventually rather than within specific time intervals, but when you expect the return action to happen consistently at every return interval, then you could use IN.
6 » Filters
Filter your reports by any Device properties such as app version, device model, device platform, etc., or by any User properties such as city, country, or other custom user properties.
7 » Groupings
By default, UXCam defines cohorts based on when users performed the start action. However, if you would like to dig deeper and compare behavior based on other properties, you can segment your report by applying groupings.
- Group by device properties such as app version, device model, device platform, etc.
- Group by user properties such as country, city, and any custom user property such as subscription type, gender, age, type of customer, etc.