When measuring Churn, it’s important to understand if people are leaving because they couldn’t use your product, if they didn’t like it, or if you’re just dealing with happy churn. 

Happy churn is when someone reaches the end of the rainbow. They no longer need your product because they’ve dealt with their problem. Happy cows will say lovely things about you, and they’ll always come back when problem crops up again. If you’re dealing with happy cows then all is well in the world. 

You can segment these three groups by sending out an email after people stop using your app for a while. Send it too soon and you’re just spamming people on holiday. Leave it too late and you probably won’t get a response.

Getting a response from people who've decided not to use your app anymore is an uphill climb. If you do manage to speak to them, and you somehow get them to tell you why they left, it probably won’t be great data. You’re just sampling the people that were polite enough to speak to you.

If people churn because they couldn’t use your app, then you have an engineering problem. You should have alerts and error logs that show you this well before people start leaving. Sending emails to this group is your last line of defence. You should be able to detect problems like this well before they are ready to leave. 

Lastly, you want to make sure you’re only sending emails to people who actually used your app. I’m going to assume they signed up, engaged with the product for a while and then left. If not, we’re talking about an onboarding problem, not a churn problem.