Analysis Modes

Cohort analyses give you the ability to toggle between different modes of analysis that measure user behavior in different ways.

First Time vs. Recurring Behavior

• First Time: Only takes into account the first time a user performed the target behavior during the selected date range. Use the first time mode to model the distribution of time it took for users to first perform a target behavior. For example, if the selected time interval is daily, and a user performed the event on three different days, they will only be counted in the data for the first day on which they performed the event.
• Recurring: Shows the number of users who perform the target behavior over time. Users who performed the target behavior multiple times over time are considered in this analysis. For example, if the selected time interval is daily, and a user performed the event on three different days, they will be counted in the data for each of those three days.

To toggle between the First Time and Recurring cohort analysis modes, click on Recurring in the Query Builder:

Complete vs. Incomplete

In queries that use a generational breakout, such as hour, day, week, or month of the event, the setting for Only Complete/Include Incomplete controls how the cohort visualization accounts for periods with incomplete data.

If the query is set to Only Complete, only cells where all users have had a chance to complete the target behavior in the analysis will be shown.

For example, consider a query showing users who first performed the event Blog View and returned to perform the event Subscribed, grouped by Week of Blog View.

With the visualization set to show Only Complete, results from this week will not be displayed. This is because users who entered the analysis by Blog View in the week have not yet had a full week to be able to complete Subscribed and show up in the completed analysis.

If the visualization is set to Include Incomplete, an extra series of data will be displayed in the cohort. These numbers will typically be lower than the completed figures, because there has not yet been enough time to capture the full extent of users who are completing the parameters of the query.

By utilizing the Only Complete/Include Incomplete setting, you can customize your results so that you’re viewing data as it comes in, or so that you filter incomplete data from your analysis.

Cumulative vs. Non-cumulative

The Non-Cumulative and Cumulative cohort analysis modes determine how your cohort data is analyzed over time. To toggle between the non-cumulative and cumulative cohort analysis modes, click on Non-Cumulative in the menu bar, just below the Query Builder.

• Non-Cumulative Percent: Displays the percentage of users who completed the target behavior for each selected time interval or breakout.
• Non-Cumulative Count: Displays the count of users who completed the target behavior at each point in time defined by the selected time interval.
• Cumulative Percent: Displays the percentage of users who completed the target behavior for the first time as a running total over time.
• Cumulative Count: Displays the count of users who completed the target behavior for the first time as a running total over time.

Filters

Filter a query to include or exclude users who meet a certain condition. You can filter on event properties, user properties, or user segments. There is no limit to the number of filters you can apply to a single query row.

In the following example, PetBox wants to exclude any users who downloaded their app on Android. Therefore, the user selects the filter where function, and creates a filter where device type is not equal to Android.