Customer churn analysis measures how many customers or recurring revenue you lose during a specific period. It also examines why customers leave and what may cause them to cancel. The goal is to find patterns that show who leaves, when they leave, and why.
Most teams track their churn rate, but fewer take time to understand why customers leave. A rising churn rate can point to issues with pricing, service, product quality, or even customer support.
In this guide, I’ll explain customer churn analysis in simple terms. You’ll learn which churn metrics matter and how each one fits into the analysis. I’ll also cover the main methods and show you how to run the process step-by-step.
Disclaimer: This content contains some affiliate links for which we will earn a commission (at no additional cost to you). This is to ensure that we can keep creating free content for you.
Customer churn analysis is the process of measuring how often customers stop doing business with you. It also examines the reasons behind those losses. The process combines two sides of the same problem. One side shows how much business you lose. The other explains why customers decide to leave.
Churn, also called attrition, refers to lost customers or revenue during a set period. As opposed to reporting, which only shows how many customers left, customer churn analysis examines who left and why.
It goes further by examining customer groups, account value, and lifecycle stages. It can also reveal whether certain groups leave more often than others.
For example, a business may notice higher churn among customers after their first three months. Further analysis could reveal poor onboarding as the main reason. Another business may find that customers leave after a price increase. These findings point to specific areas across the sales funnel that need attention.
Customer churn analysis takes two forms. Descriptive analysis looks at past churn and identifies patterns in customer losses. Predictive analysis uses customer behavior to identify those who may leave next. A strong churn program uses both approaches to understand past losses and spot future risks.
High churn increases pressure on acquisition because more new customers must replace those who leave. Customer retention and churn rates offer two different views of customer loyalty. Retention shows how many customers stay, while churn shows how many customers leave.
A falling retention rate alongside rising churn signals a growing problem with customer loss. These figures can also show whether changes in the customer experience affect loyalty.
In over a decade of helping brands grow, I’ve watched teams pour budget into acquisition while ignoring the leak at the bottom of the funnel. Customer churn analysis is how you find that leak.
Customer churn analysis shows how many customers leave, why they leave, and which groups face the highest churn. It involves the use of descriptive and predictive methods to understand past losses and identify future churn risks.
Churn falls into two main categories based on its cause and what you count. The cause can be voluntary or involuntary, while the measure can focus on customers or revenue. Each type points to different causes, sales funnel metrics, and actions.
Before you analyze churn, you need to know exactly what type of churn you are measuring.
Voluntary churn occurs when a customer chooses to leave your business. They may cancel their subscription, switch to another provider, or stop buying from you.
Involuntary churn happens when customers leave without making a deliberate choice. Failed payments, expired cards, and payment issues are common causes.
These two types require different responses. Voluntary churn may point to product issues, poor service, weak value, or better offers from competitors. Involuntary churn often needs a payment recovery process rather than changes to the product.
This distinction can have a larger effect on churn figures than many teams expect. Recurly’s 2026 research reports a median annual subscription churn rate of 3.60%. About one-third of that churn comes from involuntary losses.
That means some customers counted as churned during customer churn analysis never intended to leave.
Customer churn, also called logo churn, measures the number of customer accounts you lose during a period. Revenue churn measures the recurring revenue lost from those customers. Both metrics tell you different things about the health of your business.
These rates can move in very different directions, which matters when conducting customer churn analysis. For example, losing ten small customers may create high logo churn but have little effect on revenue. Losing one large account could create lower customer churn but cause a much larger revenue loss.
For subscription and SaaS businesses, revenue churn is usually the more honest number, because not every customer is worth the same.
The table below shows how the four types line up.
| Churn Type | What It Measures | Best Fix |
|---|---|---|
| Voluntary churn | Customers who choose to leave | Product, value, and experience |
| Involuntary churn | Customers lost to failed payments | Dunning and payment retries |
| Customer (logo) churn | Number of accounts lost | Retention and onboarding |
| Revenue churn | Recurring revenue lost | Expansion and high-value retention |
Breaking churn into these categories makes customer churn analysis clearer and helps identify the right response.
Split churn by its cause, such as voluntary or involuntary, and by what you measure, such as customers or revenue. Involuntary churn is often recoverable, while revenue churn can reveal losses that customer churn alone may miss.
The core of customer churn analysis starts with metrics you can calculate from data you already collect. The three most important are customer churn rate, revenue churn rate, and net revenue retention. With these metrics, you have a base for the rest of your customer churn analysis.
Customer churn rate is the percentage of customers you lose in a period. The standard formula divides customers lost by customers at the start of that period. You then multiply the result by 100 to get the churn percentage.
For example, losing 20 customers from 500 produces a monthly churn rate of 4%.
Choose your reporting period before you calculate the rate. Monthly churn works well for fast-moving subscriptions, while annual churn suits slower B2B sales funnel. The same business can appear healthy or unhealthy depending on the period you measure.
There is another issue that many churn reports overlook. Businesses do not always use the same customer count as the denominator. Some divide lost customers by the starting total, while others use the ending total. Some teams use the average of both figures instead.
Each method produces a different rate from the same customer losses. Choose one calculation method for your customer churn analysis and use it across every report.
Revenue churn rate is the percentage of recurring revenue lost to cancellations and downgrades. You calculate it as revenue lost divided by revenue at the start of the period, times 100.
Revenue churn focuses on the financial impact of customer losses. It is also an important ecommerce KPI for tracking recurring revenue.
Net revenue retention, or NRR, takes the calculation further by including expansion revenue. Start with your opening revenue and add revenue from upgrades.
Then subtract revenue lost through downgrades and churn. Divide that figure by your starting revenue and multiply by 100.
An NRR above 100% means your existing customers are growing faster than they’re leaving. This can happen before the business adds any new customers. Strong NRR can support growth even when customer acquisition remains unchanged.
Here’s a simple example:
A business starts a quarter with $100,000 in monthly recurring revenue. It adds $15,000 through upgrades but loses $8,000 through downgrades and cancellations. Its NRR reaches 107%, so the existing customer base grew without new accounts.
Customer churn analysis should also consider benchmarks when you interpret these figures. Recurly’s 2026 research reports a median annual churn of about 3.6% overall and 3.2% for SaaS. Consumer categories such as media and ecommerce tend to record higher churn rates.
Use these figures as general reference points for your customer churn analysis. Your own churn trend should be your focus, not just an industry average. A rising rate within your business may signal a problem, even when it's below the benchmark.
The table below compares each metric and the specific type of churn it measures.
| Metric | Formula | What It Tells You |
|---|---|---|
| Customer (logo) churn rate | (Customers lost ÷ customers at start) × 100 | Share of accounts lost |
| Revenue churn rate | (Recurring revenue lost ÷ revenue at start) × 100 | Share of dollars lost |
| Net revenue retention | (Start + expansion − contraction − churn) ÷ start × 100 | Growth from existing customers |
The right metric for customer churn analysis depends on what you want to understand about customer losses. You can track these figures together. Review them regularly so you'd spot unusual changes and investigate the reasons behind them.
Track customer churn rate, revenue churn rate, and net revenue retention. Choose one calculation method and use it consistently throughout your customer churn analysis.
Once you can measure churn, methods are how you explain it. The main customer churn analysis methods are cohort analysis, segmentation, behavioral analysis, and predictive modeling. Most teams start with the first two, then layer in the others as their data matures.
Cohort analysis involves grouping customers based on when they signed up or started using your product. You then track how many customers remain active across each period. A retention curve can show when customers tend to leave after joining.
The drop may happen during the first month, after three months, or around renewal. These patterns can point to different problems within the customer journey.
For many businesses, cohort analysis provides one of the clearest views of churn. A sharp drop during the first 30 days may point to weak onboarding. A gradual decline may suggest that customers lose value over time.
The shape of the retention curve tells you where to investigate first. It also lets you compare newer cohorts with older ones and see whether customer retention rate is improving.
Segmentation breaks churn into groups based on shared customer attributes. You can segment customers by plan, industry, company size, acquisition channel, or region. This method answers the question of which customers are most likely to leave.
Churn rarely affects every customer group at the same rate. One segment may account for most customer losses while others remain stable.
For example, a business may report an overall churn rate of 4%. A closer look could show 12% churn from one acquisition channel. Other channels may sit close to 2%.
That difference can change where you focus your customer retention strategies. Combining segmentation with web analytics can also reveal patterns in customer behavior.
Behavioral analysis looks at what customers do before they leave. Common warning signs are fewer logins, lower feature use, and more support tickets. Changes in customer behavior can reveal problems before a customer cancels.
Customer churn risk analysis can also use support data to identify early warning signs. A rise in complaints may show growing frustration before it leads to cancellation.
Predictive modeling uses these signals to estimate which customers may churn next. You don't need a large data science team to start using basic predictive rules.
For example, a CRM for SaaS can flag accounts with no login for 30 days and an open complaint. More advanced teams can use logistic regression, decision trees, or survival analysis.
Survival analysis adds a time element to churn prediction. It estimates when a customer may leave, not just whether they may leave.
Here's an overview of how the three customer churn analysis methods compare.
| Method | What It Reveals | When to Use It |
|---|---|---|
| Cohort analysis | When customers drop off | Always start here |
| Segmentation | Which customers churn most | To target the biggest leaks |
| Behavioral analysis | Early warning signs | To catch risk in real time |
| Predictive modeling | Who is likely to churn next | When your data is mature |
Each method answers a different question about customer losses. Start with methods that match the data you already have. Add more advanced analytics tools when you have enough reliable customer behavior data.
Cohort analysis shows when churn happens while segmentation shows who churns. Behavioral analysis shows the warning signs and predictive modeling shows who’s next.
Running a customer churn analysis follows six repeatable steps. You define churn, gather the data, calculate the metrics, segment for patterns, flag at-risk customers, then act on what you find.
These customer churn analysis steps use data you already collect. Here’s how each step works in practice.
Define churn, gather data, calculate metrics, segment for patterns, flag at-risk customers, then act and re-measure. The loop matters more than any single report.
A good customer churn analysis should lead to clear actions, not just another report. Start by finding when customers leave, which groups leave most, and why they leave.
For example, a SaaS company may have 1,000 customers and lose 30 during one quarter. Its quarterly churn rate is 3%. That number matters, but it does not explain the reason behind those losses.
Break the 30 lost customers into groups based on plan, signup date, acquisition source, and behavior. You may find that 15 came from one campaign. Further checks may show that many never completed onboarding.
This way, customer churn analysis points to a clear problem. The company may need to improve the campaign message and customer onboarding software and process.
A customer churn analysis dashboard keeps this visible instead of buried in a spreadsheet. Strong customer churn analysis visualization makes these patterns obvious to a whole team.
Here are the metrics some useful dashboards track.
| Dashboard View | What It Answers |
|---|---|
| Churn-rate trend | Is churn rising or falling over time? |
| Cohort retention curve | When do customers drop off? |
| Churn by segment | Which segments leak the most? |
| Voluntary vs involuntary split | How much churn is recoverable? |
| At-risk account list | Who needs attention this week? |
Many customer success platforms include these views out of the box. HubSpot Service Hub customer success tool surfaces customer health scores, at-risk accounts, and renewal trends on one dashboard. That keeps the signals from your analysis in front of the people who can act on them.
Image via HubSpot
Customer churn analysis dashboard should also show changes from the previous period. You can spot new problems before they become larger issues.
Once you find the main causes, choose one action for each major problem. Don't try to fix every churn cause at once. Follow these tips:
Run your customer churn analysis again after making these changes. Compare the results with your original figures and check which actions worked.
Keep reviewing customer retention as customer behavior and needs change. Pairing these efforts with steady retargeting keeps lapsed customers within reach.
Use your customer churn analysis data to find the main reasons customers leave and choose actions for each problem. Improve onboarding, target at-risk customers, recover failed payments, or review pricing based on your findings. Then measure churn again to see which changes worked.
Q1. What is a good customer churn rate?
A. A good customer churn rate depends on your model, but lower is always better. Recurly’s 2026 research puts the median annual subscription churn rate near 3.6%, with SaaS closer to 3.2%. Compare yourself to your own trend and your industry, not a single universal target.
Q2. How do you calculate the customer churn rate?
A. Divide the number of customers lost during a period by the number of customers at the start. Then multiply by 100. Lose 20 of 500 customers in a month, and your monthly churn rate is 4%. Just keep the same denominator basis in every report.
Q3. What does a 20% churn rate mean?
A. A 20% churn rate means you lose one in five customers over the measured period. If that’s monthly, it’s severe, since you’d lose most of your customer base in a year. If it’s annual, it may be normal for some consumer or early-stage businesses.
Q4. How do you measure customer churn?
A. First, you choose a metric such as customer churn, revenue churn, or net revenue retention. Then choose a period, such as monthly, quarterly, or yearly. Define what counts as churn and use the same definition each time. Pull the data from your CRM or billing system.
Q5. What is the difference between customer churn and revenue churn?
A. Customer churn measures the number of accounts you lose. Revenue churn measures the recurring revenue lost from those accounts. The two numbers are different. Losing several small accounts may raise customer churn but have little effect on revenue.
Q6. How would you use Excel to analyze customer churn?
A. Export your customer and subscription data into Excel. Use pivot tables to group churn by cohort, plan, or customer segment. Excel can work for smaller datasets and basic analysis. For larger datasets, you’ll need a CRM or product analytics tool.
Q7. What is the best model for customer churn prediction?
A. There’s no single best model for any business. You can use logistic regression, decision trees, random forests, and survival analysis, and the right one depends on your data. Clean, complete data matters more than the algorithm you choose.
Q8. What data do you need for a customer churn analysis?
A. You need subscription data such as signups, renewals, and cancellations. Product usage, support history, plan, and industry data can also help. The quality of your data affects how well you can find churn patterns and warning signs.
Q9. How often should you run a customer churn analysis?
A. Track your main churn rate each month to spot changes early. Run a deeper analysis each quarter to review segments, cohorts, and churn drivers. Repeat the analysis after major changes to pricing, onboarding, or your product.
Q10. Is some customer churn unavoidable?
A. Yes. Some customers leave because their needs change, they outgrow your product, or even their business closes. Some payment-related churn can also be difficult to prevent. So, the goal is not to have zero churn but to understand why your customers leave and keep avoidable churn rate low.
So, what is customer churn analysis? It’s the process of measuring customer and revenue loss, diagnosing the causes, and acting before they compound. Get the metric and the method right, and your customer retention strategy has a clearer direction.
Strong retention teams treat churn analysis as an ongoing process. They measure churn, find the causes, take action, and review the results.
HubSpot Service Hub is one option to consider if you need a tool to track customer health and at-risk accounts. If you’d rather have a team turn those insights into a retention and conversion plan, see how Attrock’s conversion optimization services can help.
Disclaimer: This content contains some affiliate links for which we will earn a commission (at no additional cost to you). This is to ensure that we can keep creating free content for you.
Marketing attribution is how you decide which touchpoints get credit when a customer converts. That…
Customer survey software is used to create and send questionnaires to customers to measure their…
Email deliverability decides whether the message you just sent reaches someone's inbox or their spam…
Customer retention rate tells you how many existing customers you keep over a set period.…
Customer onboarding software guides new clients through your product so they get value fast. It…
AI data analysis tools use AI and machine learning to examine data and identify patterns,…