How to Calculate Customer Churn Rate?

By Eman Azzoz - May 11, 2020

First, let's introduce this concept for those who don't know much about it. If you visit Wikipedia, you will find the term "attrition" instead of "churn", both of them refer to the same thing which is the loss happened in your business at a certain period. But experts differentiate between revenue churn and customer churn.

In previous articles, we've talked about customer churn reasons and reduction. In this one, we discuss another ambiguous concept related to customer churn, which is how should you calculate it to get a number that really represents your churn.

Customer churn and revenue churn difference

Customer churn is the percentage or number of customers who stopped their business with you at a certain period. On the other hand, revenue churn is about how much revenue you lost at a period of time due to cancelation or downgrading, it's sometimes called "MRR Churn".

  • Customer churn is about losing customers.
  • Revenue churn is about losing cash.

Customers may leave you; that's obviously a customer churn. But they may also downgrade their subscription, which is reflected in your revenue even if the customer is still considered a retained one. Customers' behavior (whether it's a cancelation or a downgrading) affects your revenue and, accordingly, your MRR. Thus, they call it "MRR churn".

Revenue churn is highly associated with LTV; thus, you can find interesting facts when tracking it. For example, you may find that the techniques you apply to keep lots of customers affect your most valuable ones, or you may find that one valuable customer covers all the losses that happened at a specific time. So you better focus on those customers and give him/her special treatment.

Calculating customer churn

Most subscription metrics don't have an agreed-upon definition; instead, each organization does its best to reach the most accurate formula that represents these metrics and guides its decisions.

Calculating customer churn is a complicated thing as it depends on the complexity of your system, the duration you are calculating at, the policies of your organization, and the purpose of your calculations. Thus, there are different formulas and different comments about their accuracy and reliability.

The simple formula

Churn Rate Calculation
Number of Churned Customers

Let's take an example!

Customers at the beginning of October1000
Customers at the end of October600
Churn rate(1000-600) / 1000 = 40%

But there isn't a business that doesn't acquire any customers during the month! So, let's think of those numbers; do they represent your business correctly?

The total number of customers

Say that the customers at the beginning of the month are those who renewed their subscription, what about new customers who joined your business during the month? Should you add them or not?

What if your subscription is weekly. That means, during October, some new customers may be added, and some of those added customers may leave you.

So, how should you calculate the total number of customers, should you calculate it at the beginning of the month? Or at the end? Maybe you should take the average?

Believe it or not, numbers can trick you and give you unreliable outputs.

The churned customers

According to what mentioned above, the formula we mentioned to calculate the number of churned customers isn't accurate enough. It only applies when there aren't added customers, so you have only customers who renewed and who didn't renew at the end of the month (which is odd).

Steven H. Noble's formula

There is another way suggested by Steven H. Noble from Shopify, which is to represent the total number of customers according to customers' opportunities to churn. So each day, a customer may churn at is an opportunity. 

This method depends on calculating:

  • The number of customers at the beginning of each day in the month.
  • The number of churned customers each day of the month.

Final thoughts and suggestions

Who can churn?

When calculating the total number of customers, remember that you only add those who can churn during the period X you choose.

Say the period you are calculating your churn at is a month (the easiest way), you have 600 customers at the beginning of this month. One hundred of them have an annual subscription, another 50 of them have a quarter subscription. Thus 150 of your customers won't renew during this month because they are stuck for a longer period.

Our suggestion is to not include all the customers in your calculations; you only include those who can leave you within the period you are calculating at.

So the real number of customers who can churn at the beginning of this month is (600 - 100 - 50 = 450).

Instead, you can track inactive customers during this period as there is a higher property for them to churn. You can use some techniques to reactivate them and warm them up.

Simplicity or accuracy

Sometimes trying to figure the most accurate formula leads to complications and makes tracking your churn a burden to your organization. Though something as easy as this may give you a good perspective over your churn (if you take into consideration the concerns we mentioned above):

  • The number of total customers = the number of customers at the beginning of the month (Those who can churn during the month).
  • The number of churned customers = the number of customers who left you (from those who started the month with you).
  • Without including added customers (as they will be added on the next month)
  • You also don't include those who subscribed and then churned in the same month (if your polices allow this).