Churn Calculation Pitfalls
1. Don’t Include Growth
Calculating the number of churns during the period should NOT include any new subscriptions or cancellations for those new subscriptions. Here is an example:
|Customers at Start||1800|
|Existing who leave by end of period||102|
|New who churn||18|
The wrong way to calculate:
Since the first calculation includes new subscribers in the formula, we’re not measuring churn anymore, we’re measuring growth. Don't include new subscribers in the formula.
2. How to Calculate Annual Churn
Many people look at their monthly churn rate and think that’s all there is to it, but one’s monthly churn rate is only part of the picture. From the above example we might say that a 5.67% isn’t too bad. And that would be true for an annual churn rate, but it’s not true for a monthly churn rate.
To calculate the annual churn rate first we need to determine the annual retention rate:
(1 - .0567) ^ 12 = 0.496361414
Once we have that we can figure out the annual churn rate:
1 - 0.496361414 = 0.503638586 = 50%
So a monthly (or average monthly) churn rate of 5.67% is actually an annual churn rate of 50%. This means that every year you’d have to replace half of your customers just to stay even.
Why do we raise to the power of 12? My friend Cayce Balara of Pragmatim Consulting Services explains it with a roulette analogy:
With 12 monthly cycles, (1 - monthly churn rate) ^ 12 tells you what percentage you are left with at the end of 12 months, so the reciprocal is your annual churn rate.