Churn refers to the attrition rate of your customers. It basically means how many customers leave during a specific time period, which means how much revenue you lose. It is a critical KPI, right up there with the revenue.
During early stages of your startup, churn does not have a serious effect. 4% churn on a customer base of 100 means you are losing 4 customers every month. You can easily find 4 new customers to replace them. However, when you grow big and have, say a million customers, 4% churn means you are losing 40,000 customers every month. Replacing 40,000 customers every month tends to be difficult!
There are basically 2 types of churn: Customer Churn & Revenue Churn.
Customer Churn= (Customers at the beginning of the month- at the end of the month)/ Customers at the beginning of the month
Revenue Churn= (Monthly revenue at the beginning of the month- at the end of the month)/ Monthly revenue at the beginning of the month
It is clear that customer churn and revenue churn aren’t the same.
While calculating customer churn, do not include any new sales from that month. For calculating revenue churn reduce any additional revenue from the existing customers.
For e.g. You had 8lac monthly revenue at the beginning of the month and 6lac at the end. You also booked 1lac in upgrades from the existing customers.
Churn= (8lac- 6lac-1lac)/8lac= 12.5%
Why should you track both the types?
If you have 25 customers paying Rs 100 a month and 25 premium customers paying Rs 1000 a month- you have 20 customers and monthly revenue of Rs 27,500.
If you lose 10 of them, the customer churn rate is 20%.
If 9 out of them are from small account, the revenue churn is ~7% and the monthly revenue is Rs 25,600
If 9 out of them are from premium account, the revenue churn is 33% and the monthly revenue is Rs 18,400.
Apparently the two numbers could be very different and you should understand both to get the complete picture of the business.
Negative churn means that Expansion revenue from existing customers is more than the revenue lost from churning customers.
Do remember not to get trapped in the maths of this. The more you explore, the more complex it becomes. Like this or this
Analysing Churn: Cohort Analysis
Once you have tracked your churn rate, it is important to analyse and improve. Cohort Reports are generally used for this analysis.
Cohort is just a fancy term for the group of customers who purchased your services in the specified time period. For e.g. January 2017 cohort would be all the customers who closed in Jan. It is basically a post-mortem analysis. But it does help you ask pointed questions.
Let’s take an example.
In this cohort report, customers closed in April had 100% retention in month 0, 79% retention in month 1 and so on. Layering a heat map on top of this in excel, gives us clear patterns. You can see a stability in retention from month 4. This could possibly mean that customers who clear that time frame are mature and tend not to churn. August cohort also has a clear change in colour. This could mean that people who bought in August & later have lower churn rates. Find out the reason and improve upon it. Did you release a new version of your product, launched new marketing campaign, stopped overselling, improved onboarding? Also, if you do know that you released a new version, this chart is a good evidence that it is improving retention.
During your cohort analysis, if you notice that the churn is higher in first 2-3 months, try improving on your onboarding process, or training your onboarding person, or improving engagement with your customer. You may notice increase in churn when the annual or half-yearly contracts are up for renewal.
Once you have identified a problem and started addressing it, you would want to know the effectiveness of such actions. Compare customers before the cut-off date and after the cut-off date to see the changes.
You can also create cohorts on the basis of user action, for e.g. who subscribe to a specific feature. You would be able to understand which features retain most customers, your product team can help in improving such features, fix the issues in features with high churn rate.
You can dive as deep as needed. Use external software to track user behaviour or check with your team if they can get you the required data for the button clicks in your app, or a group of button, and what not.
Download a simple cohort analysis template for early stage SaaS, by Christoph Janz of Point Nine Capital VC fund.
There are many softwares available that can help you create models to understand the reasons for your churn and create a churn score for all your customers. Churn score uses all relevant customer measures to identify how likely a customer is to leave your services. This will help you reach out to customers and stop them from churning.
So what is the acceptable churn rate?
Bessemer Venture Partners considers 5-7% annually as an acceptable churn rate. And just to be clear, 5% monthly churn is not acceptable. 5% monthly churn turns to be ~46% annual churn.
Bain & company and Harvard Business School state that increasing customer retention rates by 5% increases profits by 25% to 95%.
Start exploring and see what you can do to reduce the churn.