Churn rate is the most depressing of the relevant metrics to be aware of. It represents the percent of your total users each month who leave the game. This is obviously a regular phenomenon as the life cycle of players is always finite. For clarity this metric can understood as follows.
If you have 100 users and 10 users leave the game you have experience a churn rate of 10%
Churn Rate Formula
10 / 100 = 10%
Users who leave / Total Users = Churn Rate
Users Who Will Stay
The inverse of this metric holds a ton of value as well. If you think about it the other way churn rate allows you to predict how many users will stay. (Example – 90 / 100 = 90%)
If you had a fairly consistent average for how many users were leaving on a monthly basis can you predict the volume of users you can expect to have (without any positive growth) in a number of periods.
Number of users in 3 periods; n = number of periods
.9(^n) * 100
.9(^3) * 100 = 81 users remaining
Average User Life
Going even further we see we can use the churn rate to estimate the likely lifespan of a user.
1/churn rate = Average user life
1 / .1 = 10 months
We have to be very careful when looking at this computation because we’re making an assumption to get a figure instead us just looking at the raw figure. The reason I say this is because if you have access to the churn rate of your game, you more than likely have your actual average user life direct from your database. So what I suggest you do is compare the value yielded from 1 / churn rate and compare it to your actual average user life.
Let’s say your average user life that you pulled from your database was 8 months.
8 months compared to 10 months is a fairly big difference, so what do you do with this? To keep it simple (i’ll do advanced optimization on this later) we’ll use it to look at variance.
If you took the past 3 months of churn rate (1 / .1 = 10 months) and saw generated an average user life longer than your database pulled number (8 months) you would know that you’re experiencing a lower churn rate during these 3 months compared to your average.
Conclusion on this point;
When your average user life from churn is > than the database = your experiencing churn lower than average
When your average user life from churn is < than the database = your experiencing churn higher than average
Churn rate can be alarming if you look at it in isolation.
Generally a social or virtual game will experience between 20% – 30% churn per month
Given the nature of ease of entry to a F2P or social game, their churn is significantly higher. For a premium title you can expect closer to 10% per month churn