Impact: Be able to see the impact on retention

  • 31 January 2020
  • 2 replies

We want to be able to asses how usage of a feature from a few months ago is affecting which users are still active currently. The impact tool is very valuable for looking at changes in engagement, but it would be so much more useful to us if we could measure retention.


All we would need to do is instead of measuring on total number of times they did an event, we would like to have at least one of these options to analyze against:
Total distinct days/weeks/months that they did a specific event
Percentage of people who an event at least x months after the launch date
Time between the launch date and the final time they did an event.

I don’t know if this is the right tool to be analyzing this with, but we think these features would be super helpful if they existing some where.

2 replies

Userlevel 4
Badge +4

Hi @zachToybox,

Can you share a bit about the business question you'd like to answer with this type of report? I'd be happy to do some digging here to see if this can be answered within Retention or another report in the Mixpanel UI.

In the meantime, I'll mention that the Funnels report might be an appropriate environment to explore as well. In Funnels you can see how many users performed a sequence of events, and you can list repeat events to see how many users perform the same event over and over. Here's an example showing retention-like activity after an event is performed twice



Let me know if one of these options looks like it will work for your context!

So this is actually one of the biggest things we want to look at. Our most important metric to us is our retention rather than engagement. So we want to see how many times they come back on a separate day and how far in the future they keep coming back. We are less interested in if they do something 10 times in one day, and more interested if they did at least one thing each day for 10 days.

For instance, there was a specific event that 5% of our users hit 3 months ago.
What we are looking for is to answer a few questions:
1. Did they come back the next day and do our retention event, how about the day after that.
2. Do they continue to come back and do our retention event 1 week later, 1 month? 3 months?
3. We would like to compare the difference between the people who did and didn’t do the impact event.

It seems like this is possible to do with cohorts in the retention, but it just isn’t nearly as easy to get that going if we want to change the triggering event we would have to make multiple cohorts to test each hypothesis about which events are impactful. The impact view is great because it compact and gives a lot of easily digestible information that you can easily change around and filter to what you want, the retention one is a bit harder to do that with.