Churn-rate

Churn rate (d)

The share of devices that were not active in a certain period. These are devices with events that occurred during the report period and without events after a different specified time frame (d - days, w - weeks, m - months).

Churn rateFlow metric
Churn rate CACurrent Attribution metric

For example, let's get the churn rate among 100 devices where launches were from 2021.01.01 to 2021.01.14. Specify the churn rate period equal to a week — Churn rate 1w. If you append the churn rate period to each day of the report period you will get control period to check the churn. It's a period from 2021.01.08 to 2021.01.21 in the current example. Let's say that there were no events on 25 of 100 devices during this time, then Churn rate 1w will be 25%.

Shouldn't compare Churn rate and Retention. Metrics describe intersecting but a different audience. Churn rate shows "losses" among active devices of your project and Retention shows "the viability" of new devices.

Let's consider the churn rate after 1, 2 .. 3 weeks following active 2 weeks of devices A, B, C, D:

  • Churn rate 1w = 25%, because after 1 week following activity, there were no events on one of four devices for 2 weeks.
  • Churn rate 2w = 50%, because after 2 weeks following activity, there were no events on two of four devices for 2 weeks.
  • Churn rate 3w = 75%, because after 3 weeks following activity, there were no events on three of four devices for 2 weeks.
Report period
01.01.21 - 14.01.21
Control period 1
08.01.21 - 21.01.21
Control period 2
15.01.21 - 28.01.21
Control period 3
22.01.21 - 04.02.21
A                                                        
B                                                        
C                                                
D                                                
Churn rate 1 of 4
Churn rate 1w = 25%
2 of 4
Churn rate 2w = 50%
3 of 4
Churn rate 3w = 75%

Churn rate (u)

The share of users who were not active in a certain period. These are users who interacted with the app during the report period, but didn't do so on any device after a different specified period (d - days, w - weeks, m - months).

Churn rateFlow metric

For example, let's get the churn rate among 100 users, who ran the app from 2021.01.01 to 2021.01.14. Specify the churn rate period equal to a week — Churn rate 1w. If you append the churn rate period to each day of the report period you will get control period to check the churn. It's a period from 2021.01.08 to 2021.01.21 in the current example. Let's say that 25 of 100 users didn't interacted with the app during this time, then Churn rate 1w will be 25%.

Shouldn't compare Churn rate and Retention. Metrics describe intersecting but a different audience. Churn rate shows "losses" among active users of your project and Retention shows "the viability" of new users.

Let's consider the churn rate after 1, 2 .. 3 weeks following active 2 weeks of users A, B, C, D:

  • Churn rate 1w = 25%, because after 1 weeks following activity one of the four users didn't interacted with the app for 2 weeks.
  • Churn rate 2w = 50%, because after 2 weeks following activity two of the four users didn't interacted with the app for 2 weeks.
  • Churn rate 3w = 75%, because after 3 weeks following activity three of the four users didn't interacted with the app for 2 weeks.
Report period
01.01.21 - 14.01.21
Control period 1
08.01.21 - 21.01.21
Control period 2
15.01.21 - 28.01.21
Control period 3
22.01.21 - 04.02.21
A                                                        
B                                                        
C                                                
D                                                
Churn rate 1 of 4
Churn rate 1w = 25%
2 of 4
Churn rate 2w = 50%
3 of 4
Churn rate 3w = 75%