Skip to main content

I want to be able to create a cohort of returning users on a rolling basis. These would be users who do any event in the app, who had done the event ‘create an account’ on a previous day. I only seem to be able to create a cohort of users who had done the event before TODAY, not on a rolling basis so that I can see how many users on a given day were ones that had created an account on a previous day.

Some example use cases of this:

  • One of our partners sent out a marketing email last week, we want to see how effective this was at driving existing users back to the app.
  • In our database we are able to see the number of users we had in a month, that had created their account in a previous month. We have seen a big increase in the number of returning users last month. We want to understand what brought them back to the app and what they did.

Does anyone have any suggestions on how to do this? 

Hello Annabel,

Thanks for writing to Amplitude Support. Happy to help!

I understand you are looking to create a cohort of returning users on a rolling basis. At the moment, you have to specify particular date ranges in a cohort definition so the exact definition you have described is not possible out-of-the-box using cohorts. That said, here are some options you can use to get close to your desired definition:

- You can use offsets in the cohort definitions to offset the date range for when users create an account so that it looks at dates prior to the current day/date range. Here is an example cohort looking at users who performed any event in the last day, and who also performed the User Create Account Success event in the last 30 days offset by 1 day. This cohort will grab users who performed the account creation event prior to today's date.
- You can also use offset in user segments, like in the example chart here. For example, the April data point should be interpreted as 'in April'24 there are 67 users who performed any event who also performed User Create Account Success in March'24'.

Additionally, regarding the marketing email use case, I think a retention chart would solve for this use case. The retention chart shows you how often users return to your product after taking a specific action. Here is an example chart where we are looking at the users who clicked in email in the last twelve weeks and seeing in what week they returned to perform a page view (e.g. the week 2 data point is users who returned on their 2nd week after performing the email click event).

Apologies that there isn't a feature to solve exactly for your use case at the moment! I would encourage you to submit a feature request to our Product Team to document your use case. We sincerely appreciate you taking the time to share your thoughts on how we can make Amplitude a better product.

I hope this helps! Please let me know if you have any more questions!

Best,

Jacenda


P.S. Checkout upcoming events and user meetups on our events page.

Reply


Terms & ConditionsCookie settings