Solved

Is there anyway to view the cohort population over a longer time horizon?

  • 1 September 2021
  • 2 replies
  • 26 views

Badge

I’ve setup a cohort on uninstalls as we have quite a few users that reinstall our app (so the query is based on uninstall events being > install events, to correctly capture uninstalls). I’d like to see this over a longer term time horizon than 30 days, or alternatively if possible, add the cohort to an event segmentation so i can run it that way. Is this at all possible?

icon

Best answer by eddie.gaona 3 September 2021, 01:09

Hello @RosieC. 

I am understanding that you want to create a cohort where the uninstalls are greater than installs to correctly measure uninstalls for a time frame greater than 30 days.

In cohorts you can look at a max of 365 days(one year). Is this what you are looking for: https://analytics.amplitude.com/demo/cohort/0c8vzfh?source=redirect%3A+cohort+created?

The cohort is looking in the past 365 days how many users have purchased more songs than played a song.

View original

2 replies

Userlevel 6
Badge +4

Hey @RosieC. ,

Currently, the cohort population over time is in beta and doesn’t support any adjustments in date range or granularity in the cohort view.  

Also, cohort population gives the number of users who meet a cohort definition on each day over a predefined time. So afaik, an event segmentation chart might not give a similar result compared to cohort population view if you try to chart it over 30 days. 

Would love to hear from others and the Platform Specialist Team if there is a way around this.

 

Userlevel 3
Badge +4

Hello @RosieC. 

I am understanding that you want to create a cohort where the uninstalls are greater than installs to correctly measure uninstalls for a time frame greater than 30 days.

In cohorts you can look at a max of 365 days(one year). Is this what you are looking for: https://analytics.amplitude.com/demo/cohort/0c8vzfh?source=redirect%3A+cohort+created?

The cohort is looking in the past 365 days how many users have purchased more songs than played a song.

Reply