Your churn rate is one of your most important metrics as a subscription business. It has many different uses; as a general measure of business health, as a measure of customer loyalty, or as the basis of customer lifetime value calculations.
Your churn rate is also a simple idea to grasp, it is the percentage of customers that leave your service over a given timeframe. Calculation of churn, however, is not so simple. And there are dozens of different approaches in use by SaaS analytics services.
Much of this variation is unnecessary and unhelpful over-complication, but it is certainly the case that churn rates can also be over-simplified. At Cleeng, we eliminate as much complexity as possible in calculating churn, but retain a level of nuance that provides a level of balance for the clients we work with.
Cleeng's churn calculation is:
Churned Subscribers / (Subscribers at start of period + New subscribers in period)
We use this calculation because it is easily understandable, and it is also robust in volatile periods. The second point is very important to understand in churn calculation. We will quickly explain why here.
Why not use the most 'simple' churn calculation?
Simple churn works like this. You take all customers churned in a period, you divide by the number of customers at the start, and that's it.
The good thing about this method is that it's very straightforward. The bad thing is that it copes poorly with high acquisition rates and/or high churn rates. This is because simple churn fails to capture change during the time period of the calculation.
So the churn rate it reports at the end of that time period doesn't actually reflect the true size of the subscriber base. This is a significant issue for B2C subscription services where the number of new subscribers in a month can often represent 30-40% of their subscriber base at the start of the month.
Let's take the following scenario:
- 10,000 subscribers at month start
- 4,000 new subscribers
- 1,000 churned subscribers
The simple churn calculation here is 10%, but that calculation really only works if all churn takes place at the beginning of the month. In the following weeks the subscriber base is growing steadily, and by the end of the month it consists of 13,000 subscribers. From this vantage point, the 1,000 subscribers lost represent just 7.6% of your subscriber base.
However, this is also just one point in time during the month, and is not dynamic enough to truly capture what's happening either.
So the simple churn rate does not actually reflect what is happening with your subscriber base. And the distortion increases significantly as the rate of acquisition or churn increases.
Coping with change - The 'adjusted' churn calculation
To better capture this change in the subscriber base, Cleeng uses the adjusted churn methodology:
Churned Subscribers / (Subscribers at start of period + New subscribers in period)
We will take the exact same scenario as above to show how this changes the calculation:
- 10,000 subscribers at month start
- 4,000 new subscribers
- 1,000 churned subscribers
The adjusted churn calculation here is 7.1% (1,000/(10,000 + 4,000)).
By including 'new' customers in the calculation, we capture the dynamic changes in the size of the subscriber base over the course of the month.
We are no longer fixed to a single point in time (start or end), and so the churn rate is not distorted by spikes in churn or new signups. And so your churn rate is a more robust, and truer reflection of the rate of subscriber churn in your business.