Hello,

I’m struggling with a retention chart over a cohort that uses “any time within x days” filter.

I’m trying to understand how frequent a user should perform a specific action (AddToWishList) in order to be retain as a user on the platform. So I built a cohort to capture all users that perform the metric on a weekly basis for the first month of usage.

The above cohort is built according to the following filters (all with “And also” clause):

1. users performed “AddToWishList” with count ≥ 1 time any time within 7 days of first use
2. users performed “AddToWishList” with count ≥ 2 time any time within 14 days of first use
3. users performed “AddToWishList” with count ≥ 3 time any time within 21 days of first use
4. users performed “AddToWishList” with count ≥ 4 time any time within 28 days of first use

The above definition can be found at the following link: https://analytics.amplitude.com/vestua/cohort/pd31xt5

Then I built a retention analysis using the above cohort. I would expect to have a high weekly retention for new users with “AddToWishList” retention event for at least the 4 first weeks. This should be by the definition of the cohort, right? But as the chart shows, retention is drop to 14% by the 1st week.

The retention analysis can be found on this link: https://analytics.amplitude.com/vestua/chart/frf7c78/edit/04g1lv6

What am I missing on my analysis?

Thanks in advance to anyone that could help me with this.

EDIT

I just found the missing part of the equation and feel so stupid!

The cohort considers users that perform the event 4 times within the first week, and this action meets all the cohort conditions.

Regards,

Santiago.