Cohort Analysis
Cohort Analysis gives you the option to segregate this group of people according to the acquisition date, country or apps under the same publisher account.
You can set Min Cohort Size while grouping for convenient analysis. For example, while choosing country type cohort, you want to see minimum 100 people downloads from X country, so Min Cohort Size option gives you chance for this.
Metric option gives you the opportunity to sort cohorts. When you choose ARPU as a metric highest ARPU values will be on the top of the list while sorting, also still you can see other.
Cohort table shows Install Day as day-0 for people who install app between dates you select, this means when you select dates between 2nd of July to 20th of July; day-0 is installing day of 2nd of July downloads to every install until the end day. Day-1 is the 1st day of 2nd of July downloads, 3rd of July downloads… to 20th of July downloads.
You can see different shades of blue in the table, darker blue means there is higher retention, its’ shade getting lighter while retention decrease.
You can see “All Time” values at the end of the table, All Time column shows the average value of ARPU for day-0 to day that you create a table.
Cohort Analysis helps you to group people in different ways that you set:
Cohort Type – Acquisition Day: According to this selection cohort is a group of users who download the app in the same day. The system counts as day 0 for download day, and analyze cohort behavior until the day you set. You can see how many people come back to the app on the second day, and how many of them come back on the third day an so on.
Cohort Type – Country: Country selection provides data for geographical strategy. You can group people by their country and see how many of them are coming back. This is a critical data for ROI based on country.
Cohort Type – App: It is a helpful selection for publishers who have many apps, or want to compare iOS and Android version.