Churn Analysis: How to Figure Out Why People Are Churning & What to Do About It

Introduction

You worked so hard to build a great product and acquire users and then you end up losing some of them, what gives? You might have got a cancellation request by email or a notification from Stripe or maybe you saw some numbers on a dashboard go down. Either way you started the first step of churn analysis and we are here to help you understand why people are churning, what you can do about it and get those charts going up and to the right!

Measuring the Extent of Churn aka: How much churn do I have or how big is my churn problem?

Before you can address churn, you need to quantify it. This is important because you will also want to measure how your efforts to reduce churn are going. 

The key metrics you are looking for are your subscriber churn rate (what percentage of subscribers cancelled in the last month) and how much revenue churn you have, in monthly recurring revenue (MRR) or Annual Recurring Revenue (ARR) if that is a better fit for your subscription lengths.This will give a good indication on what it is costing the business in revenue on a monthly basis and whether you are losing a few big fish or a good percentage of your customers.

If you use Stripe or a similar billing provider they often have excellent charts showing these key metrics.

It is worth looking at the trend over time and seeing whether this is a normal or abnormal amount of churn. 

Once you know how much churn you have, you might want to compare it to some benchmarks to see whether you are doing well against otherwise of a similar size, average customer value or industry. 

Figuring Out Why People Are Churning

Quantitative Methods for Churn Analysis

Running the numbers aka Quantitative methods is the best place to start with understanding churn. It helps you focus on the right areas which have the biggest impact and make sure you are asking the right questions and getting the right answers when you move onto qualitative analysis. 

Involuntary vs. Voluntary Churn

This is the first place to start with churn analysis. Are you customers leaving on purpose or are they leaving because of payment failures which is also known as involuntary churn. 

• Involuntary Churn: Customers who leave due to circumstances beyond their control, usually payment failures.
• Voluntary Churn: Customers who actively decide to stop using your service, often due to dissatisfaction or better alternatives.

You might feel that this is naturally not a major contributor to churn, but it is usually between 20-40% of churn with 5-15% of payments attempted failing. 

Involuntary churn can be initial payment issues after a trial, card expiry, payment processor declining due to risk, their bank declining, insufficient funds or many other reasons. 

If the majority of your churn is involuntary there are some great tools that automate retries (also known as automated dunning) or suggest users change payment methods when their payment fails. Your payment processor or subscription management system may have these as options you can enable. 

Segment Analysis

Looking at voluntary churn, which is people choosing to leave. It is good to start by breaking it down by a few different factors to hone in on where the problem is. Often your payment processor will have good tooling for this or you can use advanced tools like Upollo which will show you specific factors on a case by case basis. 

Breaking down churn data by specific segments can highlight problem areas, some of the best places to start are:

• Plans: Are customers on certain pricing plans more likely to churn?

• Cohorts: Do customers who joined during a specific time frame have higher churn rates? This can highlight challenges with onboarding or product issues introduced at a certain time
• Countries: Is churn higher in certain regions due to cultural differences or local competition or methods of payment having higher failure rates (India & Brazil are a common examples)?

Activation Metrics

Assess whether customers are fully utilizing your product's features. Those who haven't activated key features may not see the value and are more prone to churn.

You can use product analytics tools like Mixpanel, Amplitude or Posthog to look and see if people are using the features you believe are key to impacting churn. 

Not sure which ones have an impact positively or negatively? Upollo does the analysis for you and you can look through which features statistically have an impact positively or negatively depending on whether they are done or not!

Event and Factor Correlation

Analyze user behavior data to identify patterns that precede churn. For example, a decrease in login frequency or engagement with certain features could signal dissatisfaction.

This gets into data science space. If you would like to avoid lots of queries you can spot check users who churned and see if you can spot patterns or use tools like Upollo which will tell you the factors that do actually correlate. 

Qualitative Methods for Churn Analysis

Now that you know from the numbers which areas for focus on you can dive into the details of what is going on from a users perspective. 

For instance, if you see users on a certain plan are having trouble you might want to look specifically at their exit surveys or review. Similarly if you saw people weren’t activating you might want to look at user session recordings to see what is happening. 

Below are a few of the best ways to collect qualitative data to help understand why people are churning. 

Exit Surveys

Implement exit surveys when customers cancel their subscriptions. Keep them short to encourage participation. For instance, Stripe offers a simple survey that captures essential feedback without overwhelming the user.

High level categorization is useful, but the raw feedback is often the most impactful, so if there is an option for that we suggest you enable it. 

User Flow Recordings

Tools like Hotjar, FullStory or Clarity (which is free!) allow you to watch recordings of user interactions on your platform. This can uncover usability issues or points of frustration leading to churn. Often you will see situations which are obviously causing issues and you or your team can quickly rectify them. 

Net Promoter Score (NPS)

NPS surveys can be a great way to measure customer happiness. A declining NPS can be an early warning sign of increasing churn rates. It can also be a good way to get feedback about what is making customers unhappy. 

Customer Reviews

Monitor reviews on platforms like G2 Crowd, Trustpilot, the App Store or the Play Store. Negative feedback can highlight widespread issues affecting customer satisfaction and give you a way to narrow down what needs to be solved.

Support Ticket Analysis

Analyze the frequency and nature of support tickets. Recurring complaints about specific features or services can indicate problems that need addressing.

Taking Action Based on Churn Analysis

At this stage you should have a hypothesis on where your problems with churn lie. Are they involuntary or voluntary? Are they because people are running into specific issues with the product, being unengaged or just simply not being able to get to the ahh hah moment / realizing value?  

Generally the root causes of churn fall into one of the following categories:

1. Wrong customer fit - The customer simply isn’t a good fit for your product. 
2. Never found value - They couldn’t work out how to solve the problem they had with your product 
3. Is no longer finding value -  They may no longer have the problem, moved to a competitor or forgotten about the product
4. Had a poor experience with product - This may be a bug, missing functionality, slowness or other issues with getting their core task done
5. Had a poor interaction with support or other staff
6. They are not getting enough value to justify the price - This may be compared to competitors or other ways of solving the task

There are many different methods of reducing churn from reactive methods like dunning, payment retries and offers to the proactive like churn prediction, customer success outreach, onboarding and personalized marketing. 

We wrote a guide on what to do for some of the most common churn problem with real world examples. Now that you know what the cause of churn is, you can use this guide to take a stab at solving them. 

Remember this is an iterative process and even the best products have churn! 

Conclusion

Churn analysis is a critical component of customer retention strategy. By combining quantitative data with qualitative insights, you can uncover the root causes of churn and implement effective measures to reduce it. Proactive churn prevention not only saves costs but also fosters long-term customer loyalty, driving sustainable business growth.