I’m trying to map the evolution of my population over time, and a such have separated my users into cohorts according to their activity. I have one cohort set up for “others” ie. users which do not fit into any of the other categories + users who have not picked an activity.
Is there a way for me to create a behavior cohort of users not belonging to other cohorts?
Best answer by Saish Redkar
There isn’t a way to currently reference a cohort within a cohort to exclude/include the other’s definition as a whole. You will have to create that “others” cohort using the clauses of not having performed the activities that you are interested from those individual separate cohorts.
However, you can try creating a user segment which doesn’t belong to any of the behavioral cohorts that you have created by using the “not equals” operator if that supports some of your event segmentation analysis. Here’s a sample chart.
W.r.t trying to map your cohort population over time, Amplitude’s cohort population over time feature is out there in beta which you can try using if you haven’t already.
Let me know if I have interpreted your question correctly. Hope this helps!