Hey @seb !
Interesting use case for a cohort.

I tried giving this a try using the “and not who” approach

This would give the users who were new in the last 3 days and those who used it until 48 hours if my day numbers are correct in there.

You can try tweaking the “Last X days”  to test this out.

Also, since you mentioned “signed up” you can try exploring the “performed’ clause and the ‘within’ clause

Hoping that I have interpreted your question correctly!

Let me know if this helps.

Hi @Saish Redkar ,

Thank you for your answer. I think it solves my issue relatively well, but not entirely.

Your answer picks any new user in the last 3 days which have not use the app in the last day, but I need the users who have had at least 24 hours after signup.

Here is an example

User          | Day 1 | Day 2    | Day 3 |
 
User A       Start    |  NA       | NA      |

User B        Start   |  A          |  A        |

User C        -         |   Start     |  A        |

If I were to create the cohort on Day 3 using your method, the cohort would include user A since they are a new user with less than 3 days, and have not been active in the last 24 hours. It would ignore user B and C as they have been active in the last 24 hours.

What I need is a way to include user A AND B, since they both have been users for more than 24 hours. C should not be included since they have only started 24 hours ago. If I were to create the cohort in Day 4, then all 3 users would be included.
 

My objective is to create a cohort for users who have been on the platform at least for 24 hours. If their joining timestamp is less than this amount ago, then you are not part of the cohort. I dont need them to not have been active for x amount of time, just for them to have signed up at least 24 hours ago.

Kind regards,

Seb

Hi Seb,

This is an interesting case could you share more on your use case for this? The best solution would be to implement a new event property that shows a timestamp of when a user is signing up. If you have a timestamp property of when the user performed ‘Signed Up’ you can create a cohort that looks how values greater than that value.

If you want to look at users who have ‘Signed Up’ at least 24 hours in the past without timestamp and without them having performed another event you can try to stagger your Cohort based on the present day. So for example today is March 5th 2021. I can make a cohort that look at users who ‘Signed Up’ in between days that a full 24 hours have passed. For example this set up below will show users that did ‘Play Song or Video’ between February 2nd 2021 and March 3rd 2021. Since it is March 5th right now any user that did ‘Play Song or Video’ in this time frame would have had at least 24 hours on the platform.

Hope this helps let me know if I misunderstood you at all!

Sydney

Hi @sydney.koh 

Thank you for your answer. We do have a timestamp property, but Amplitude does not allow me to use this in a dynamic way.

There is no way for me to say Timestamp > CURRENT DATE - 24 hours. I have to put in a number and cannot use a formula.

I need to create a cohort that has the following rule:

On any given day, the cohort will contain users that have signed up AT LEAST 1 day ago AND AT MOST 7 days ago.

I dont want to every have to manuall change values to make sure that is true. This should be calculated dynamically, every time I need to use it.

Kind regards,

Seb

Hi Seb,

If you want a dynamic cohort that looks at users who signed up during the last 1 day or at most the last 7 days you can use a cohort like above. The time frame will continue to move and only look at users who did Sign Up during the last 1 day or the last 7 days not including the last one day. The offset feature is the feature that moves the 7 day time frame over by one day. You can remove that if you want to include the last 1 day in your 7 day window.

Otherwise, you are correct you cannot dynamically compare the timestamp property value to the current date. I can submit a feature request for you if you share your use case!

Best,

Sydney