Solved

Dynamic cohorts based on quartiles or percentiles?

  • 15 May 2023
  • 1 reply
  • 119 views
F

Hello

I have read about creating cohorts based on users having performed a certain event a certain number of times.
For instance, if I were to to look at users that have more than 500 page views in the last 7 days, I could assume this is the most active top 25% or top 10% of my users.

However, in some time that same threshold that right now is the watershed of the top 25%, could return something completely different: maybe top 30%, maybe top 50%.

Therefor, is there a way to have a cohort that will always include the users belonging to a certain relative quartile or percentile, not users above (or below) a certain absolute threshold?

icon

Best answer by Saish Redkar 15 May 2023, 15:29

View original

1 reply

Saish Redkar
Rank
Userlevel 7
Badge +10

Hi @Franz S 

AFAIK, currently for creating dynamic cohort based on the conditions you mentioned isn’t supported yet.

The only options pertaining to event count looks to be absolute in number , except the relative count parameter. But I don’t this this one is applicable to your use case.

 

However, you can always create a static cohort from a datapoint within your selected time range using custom formulas ( like percentile ) or distribution of property value if that fits your use case.

Let me know if I have interpreted your use case correctly. Hope this helps.

 

Saish | Need additional Amplitude help ? Book me on topmate.io/saish_redkar

Reply


Terms & ConditionsCookie settings