Question

Reporting users meeting cohort definition for the first time within a day/week/month

  • 12 October 2021
  • 5 replies
  • 41 views

Badge

I have a working definition of Activated users and have several charts showing our DAU/WAU/MAU or the CUMSUM of this cohort of users. I’d like to build a chart that specifically shows new activated users. I can do this manually by clicking on the cumsum(activated_users) for week A and create a cohort of weekA_ActivateUsers, then repeat for Week B, then create a new chart with a segment where the users are in A but not in B.

But that is tedious and we want to set up a sync of only newly activated users to send an email out to. A sync won’t be reliable if I have to do this manually. 

Does anyone have an automated way to solve this? 


5 replies

Userlevel 6
Badge +4

Hey @jillCasey ,

If I’m understanding your use case correctly, you can leverage the  “New User” option in the event segmentation chart to just focus on users who have been new in the selected time period.

 

If you have access to behavioral cohorts syncs, then the had been new and/or performed the event anytime within X days of first use in the cohort definition would help you in this use case.

 

Hope this helps!

Badge

@Saish Redkar  Thank you!! Do you know if would that give me folks who were new in a cohort? Or just new to my app? My understanding is that would be new to the app. I’m hoping to find new in my cohort. 

 

Gratefully,

Jill

Userlevel 6
Badge +4

It depends how you build the cohort definition. Generally, it would be for users who are new to the app.

If you are trying to exclude the new users on a rolling basis, then I think you can try playing around the who had been/not had been new and the offset clause ( by a week ) to try mimic the segment approach that you were trying initially. I’m not entirely sure if that’s what you are trying to achieve with that approach.

Userlevel 4
Badge +4

Hi @jillCasey ! Just chiming in here to echo Saish’s recent recommendation to use the “offset” clause! 

I’ll be honest in that I’m not entirely sure what you want to achieve but if it is something along the lines of creating a cohort of users who qualify for a cohort for the first time in last X days on a rolling basis, then “offset” should help: https://help.amplitude.com/hc/en-us/articles/360040965352-Behavior-Offset 

For example, take this cohort: https://analytics.amplitude.com/demo/cohort/xxvzsip?source=manage+cohorts+table. This cohort will provide you the users who are new to your cohort by including a “and not who” clause to remove the users who are no longer “new to your cohort” 365 days before the last 30 days. 

Hope this helps!

Badge

Thank you both, you Rock!! I’m reframing my approach to use the offset mechanism. Thank you again! 

Reply