Churn Rate: How to Calculate It and Reduce Customer Loss

What is churn rate and why does it matter?

Churn rate is the share of customers who leave your business in a given period, measured against the customers you had at the start of that period. Cancelled subscriptions, lapsed buyers and lost accounts all raise it. The higher it is, the more new customers you need just to stay level.

Also, most marketing reports do not show this number. Yet it shapes your ad budget, your sales targets and your product decisions. For example, a campaign can bring in plenty of new buyers and still fail if the same number walk out the back door.

Our team looks at two things when we take over an account. First, what it costs to win a customer. Then, how long that customer stays. Also, carrying water to a leaking bucket costs more than fixing the bucket.

So in this guide we cover the formula, the difference between subscription and e-commerce, and practical ways to reduce churn.

How do you calculate churn rate?

The basic formula is simple. Divide the customers you lost during the period by the customers you had at the start, then multiply by 100. Most teams use a monthly period. However, a weekly, quarterly or yearly period can fit better if your product sells at a different pace.

The most common mistake is adding new customers to the denominator. A customer who joined this month cannot have churned yet. So you track only the group you had on day one. That way the rate measures the base you are trying to keep.

The table below shows the parts of the formula with a sample month. The numbers are an example calculation, not data from a real business.

A monthly 5% looks small. In practice, if it repeats for twelve months, about half of the January base is gone by December. To check percentages quickly, try our percentage calculator.

Also, keep the calculation date and the rules fixed every month. If the rules change, the months stop being comparable. For instance, if you count unsubscribed emails one month and cancelled orders the next, the trend misleads you.

For the yearly figure, use compounding. Multiply the monthly retention rate by itself twelve times, then subtract the result from one. Simply multiplying the monthly rate by twelve overstates the real loss.

What is the difference between customer churn and revenue churn?

Customer churn shows how many people left. Revenue churn shows how much income left with them. The two are often different. If small-plan customers leave more often, customer churn stays high while the revenue loss stays low.

Revenue churn has two versions. Gross revenue churn counts only cancellations and downgrades. Net revenue churn also subtracts upgrades and add-on sales from existing customers.

For example, say you start the month with 100,000 in monthly recurring revenue (example calculation). Cancellations remove 4,000 and downgrades remove 1,500, while upgrades add 3,000. Gross revenue churn is 5.5%, and net revenue churn is 2.5%.

So you should track both. If the customer count looks stable while revenue shrinks, you likely have a pricing or plan problem.

Above all, management cares about this split. Leaders want to see how fast income erodes, not only how many people left. Our team shows customer churn and net revenue churn side by side, so the effect of an upgrade campaign becomes visible.

How do you read churn in a subscription business?

In a subscription, leaving is a clear event: the customer cancels or does not renew. So churn rate is easy to measure here. Software tools, online courses, membership sites and recurring service contracts all fit this group.

A single average can mislead you. New members often leave at a different pace from older ones. Therefore we suggest splitting the rate by membership age, plan and acquisition channel.

Annual contracts look different. Monthly churn seems close to zero because cancellation only happens at renewal. In that case, track the renewal rate separately, otherwise the picture stays incomplete.

Free trials need a decision too. Decide up front whether a trial that never converts counts as churn or as a separate conversion problem. We usually track it separately.

The reason is simple. A lost trial says something about acquisition quality. A lost paying customer questions the product value and the price. Because the fixes differ, the numbers should stay apart.

How do you define churn rate in e-commerce?

In e-commerce, customers do not fill in a cancellation form. They just stop ordering. So you first define an inactivity window that fits your business. For example, if buyers reorder about every 60 days, you might call a customer lost after 180 days without an order.

That threshold comes from your own order data. Check how many days most customers take to place a second order. Then flag the ones well beyond that window as likely lost.

To prepare the data, group your order table by customer ID. Pull each customer's last order date and order count. Then compare those dates with today and build three groups: active, at risk and lost.

Then those three groups guide your campaigns. Active customers get an upsell, at-risk customers get a reminder, and lost customers get a win-back offer.

For one-off purchase categories such as furniture, this model does not work. There, repeat purchase rate and referral rate make more sense. Our e-commerce consulting starts by choosing the right metric for your store.

What is a good churn rate?

No single number is good for everyone. Industry, customer type, price and contract length change the result a lot. A business software sold to companies cannot be compared with a monthly consumer membership.

From our experience, the best starting point is your own history. Take the average of the last six months. Then lower it step by step. Averages you find online only show direction; they are not a target or a guarantee.

That said, a falling rate alone is not enough. If new customer quality dropped, or if the people who left after a price rise were unprofitable, high churn can be a deliberate choice. What matters is whether the loss comes from profitable or unprofitable customers.

Also, margin matters here. Customers with low margins and frequent returns can lower churn without lifting profit. So read the rate together with margin per customer.

In addition, outside events move the number. A new competitor or a platform policy change can shift it overnight. Write such events in your report notes, otherwise you will not remember the cause months later.

What are voluntary and involuntary churn?

Voluntary churn happens when the customer decides to leave. Involuntary churn happens when a card expires, funds run short or a bank declines the payment. The fixes are completely different.

Involuntary loss often grows quietly, because the customer never wanted to leave. Fixing your payment retry flow reduces it without changing the product. So check the share of failed payments first.

With voluntary loss, you need the reason. People leave because they dislike the product, find it too expensive, switch to a competitor or no longer need it. A short question at cancellation turns this into data.

• Voluntary: price, dissatisfaction, a competitor, the need ended.
• Involuntary: expired card, low balance, bank decline, technical payment error.
• Mixed: the customer notices a payment error but does not think it worth fixing.

A cancellation survey with four or five options and one click is enough. Long forms lower the answer rate and push the customer further away. After three months of data, you will see the most common reason clearly.

How are churn rate and retention rate related?

Retention rate is the mirror image of churn rate. If monthly churn is 5%, retention is 95%. The two always add up to 100%, but they give a different point of view.

Churn frames loss as an urgent problem. Retention highlights the strength. For teams, this is a communication choice: saying "reduce the loss" leads to different actions than saying "grow the ones who stay".

Google Analytics 4 supports this view too. The Retention overview report shows the percentage of users who return each day in their first 42 days. For online shops and content sites, that is a free starting point.

Also, calculate retention by segment. For example, customers who made their first order with a discount may hold on less than full-price buyers. That gap can change your campaign strategy.

Also look at the shape of the retention curve. If it flattens at some point, the remaining group is loyal. If it does not flatten, customers keep wearing away over time.

Why does cohort analysis show churn more accurately?

Cohort analysis follows customers who joined in the same period as one group. January arrivals and February arrivals appear on separate rows. So you can catch a problem that an average hides, such as customers from one campaign leaving fast.

The overall churn rate may look like it is falling while new cohorts hold on worse. The reason is that loyal old customers pull the average up. A cohort table makes that illusion visible.

In GA4, cohort exploration lets you compare user groups by the day or week they were acquired. If GA4 is not yet set up, start with our GA4 guide.

When you read a cohort table, compare rows, not columns. Each row is an acquisition period. If newer rows darken faster than older ones, acquisition quality has slipped.

Also, add the ad channel to the cohort. Customers who arrive in the same month from different channels often show very different retention curves. Then you can move budget to the channel that brings lasting customers, not only cheap ones.

How does churn rate affect customer lifetime value?

Churn rate is a direct multiplier of customer lifetime value. In the simple model, average customer lifetime equals 1 divided by monthly churn. A 5% monthly churn means an average lifetime of 20 months.

Example calculation: with 500 in monthly revenue per customer and 5% churn, the revenue-based value is about 10,000. If churn falls to 4%, lifetime rises to 25 months and value to 12,500. One point of improvement lifts value by 25%.

This calculation leaves out margin. For a real decision, multiply by gross margin. For the full formula and its variations, read our CLV guide.

Here is a second example. A customer paying 300 per month with 5% churn gives a 20-month lifetime and 6,000 in revenue value (example calculation). If acquisition costs 1,500, the first months become critical for profit.

Still, use the simple model with care. In reality, churn is not constant: it runs high in the first months and lower later. Still, it is a good start for quick decisions with one number.

As a result, a small drop in churn raises the ceiling on what you can spend on ads. That gives you more room in competitive auctions.

How are churn rate and blended CAC connected?

However low your acquisition cost is, a customer who leaves fast stretches your payback period. So CAC and churn should never be read apart. One explains today's spend, the other explains tomorrow's return.

Payback time is CAC divided by monthly margin. If the customer leaves before that time, the acquisition spend becomes a loss. In our blended CAC guide we explain how the combined cost works.

The same logic applies by channel. A channel that looks cheap is expensive if it brings customers who leave quickly. Instead of optimizing acquisition cost alone, track CAC and churn in one table.

In practice, a useful habit is to build two tables by channel over a three-month window. Put CAC in one and the sixth-month retention rate in the other. Side by side, they make budget decisions much stronger.

When you plan a new budget, ask one question: how long can we keep this customer? Without an answer, you have no basis for an ad limit.

What are the early warning signs of customer loss?

Customers change behavior before they leave. They log in less, order less often, contact support more or stop opening your emails. If you catch these signs early, you have time to prevent the loss.

Every business has its own signals. Our team does not guess; we look at the last 60 days of customers who already left. If you find a shared pattern, you turn it into an alert rule.

• A clear drop in login or order frequency.
• More support tickets, or complaints left unresolved.
• Returning to the pricing or cancellation page.
• Not opening newsletters or notifications.
• A payment method close to expiring.

So adapt this list to your own data. Even a simple rule beats no tracking at all.

You can also score the signals. Give each behavior a weight and move the customer to an at-risk list when the total crosses a threshold. A complex model is not required; a basic table tells your team when to act.

How do you reduce churn rate?

To reduce churn, first separate the causes, then fix the biggest loss. Attacking every problem at once produces results in none. The table below lists common causes and a first action for each.

Also, each row is a separate experiment. Apply one row, watch the result across a few cohorts, then move to the next. The method looks slow, but it shows clearly what works.

To find the biggest loss segment, split churn by channel, plan and customer age first. Often most of the loss comes from a small segment. One fix there moves the overall rate the most.

How does onboarding in the first 90 days lower churn?

Most losses happen in the first weeks, because the customer has not yet seen the value. Onboarding is the plan to show that value as fast as possible. It is not a training video; it is the path to the customer's first concrete win.

Here is a practical approach. Find the one action that brings customers to the first "aha" moment. Then make sure new customers take that action in their first days. In e-commerce, this is a well-timed suggestion for a second order after the first.

Next, a short message after delivery helps. A quick check on satisfaction lets you catch a problem before a return or complaint. So the customer talks to you instead of leaving silently.

Tie your onboarding messages to behavior. If a customer does not use the product on day one, send a short reminder on day three. For those who do use it, show the next step. That way everyone sees their own path, not one generic flow.

How do price and plan design affect churn?

A customer who finds the price too high may move to a smaller plan instead of leaving. If you do not offer that option, cancellation is the only exit. Therefore a lower plan or a pause option can prevent a loss, at the cost of a smaller revenue drop.

An annual payment option also lowers the rate, because decisions happen less often. The customer decides once in twelve months, not every month. However, weigh your margin when you offer a discount on annual plans.

Be careful with price rises too. If a rise arrives with poor communication, cancellations jump right after. Explaining it together with a change that adds value softens the reaction.

Loyalty programs also come in here. Still, points and discounts alone do not keep a customer. Holding on to a product without real value with a discount lasts only a short time. Fix the experience first, then add the reward.

How do you recover failed payments?

The fastest way to cut involuntary churn is to retry failed payments in a systematic way. When a payment fails, tell the customer right away, retry within a short window and send a link to update the card.

The tone of the message matters. Help, do not blame: "There was a problem with your card; you can update it in one click." Keep the update flow short, and do not send the customer through a full login screen.

Also check the retry and card update features of your payment provider. They differ by provider, so test the settings with a small group before you go live.

Then track recovery results as a separate metric. Knowing what share of failed payments you win back shows the real size of involuntary loss. If that share is low, review your technical setup and your message wording.

How do you use email and CRM to keep customers?

Email is the cheapest channel against churn. However, sending the same newsletter to everyone does not work. Split customers by behavior and write a different message for each segment.

1. New customers: a first-use guide and first-week support.
2. Active customers: new features, content and upgrade suggestions.
3. At-risk customers: a personal outreach and an offer of help.
4. Lost customers: the reason to come back and a special offer.

To run this flow, your customer data needs to live in one place. CRM integration provides that link. For writing the messages, see how AI can help in our post on AI in email marketing.

Also, keep message frequency balanced. Too many emails push people to unsubscribe; none at all lets them forget you. Set weekly and monthly limits per segment, then tune them with open and click data.

How do you win back customers with ads?

Ads are the second way to reach lost customers. You can upload your customer list to Google Ads and build a separate audience. Google offers Customer Match for these lists, and it defines re-engagement modes for lapsed customers.

You set the inactivity period yourself. The Google Ads help pages say to define it for your own business, for example no purchase in six months. The retention goal page also describes how to build campaigns for this purpose.

Also, match the message to the reason for the loss. Telling a customer who left over price about a new feature does not work. If you choose the message that fits the reason, the return rate rises clearly.

For setup details, read our remarketing guide. If you want help building this in your account, our Google Ads management service covers it.

What are common mistakes in churn analysis?

The most common mistake is looking at one overall average. That single number does not tell you which segment is losing customers. So we suggest never drawing conclusions without a channel, plan and cohort split.

• Adding new customers to the denominator.
• Changing the loss definition between periods.
• Mixing up voluntary and involuntary loss.
• Treating the loss of an unprofitable customer as a problem.
• Watching only the rate and never asking why.

Another mistake is reading ad results apart from churn. A campaign that looks good in the ad panel can lose money in the long run if customer quality is low. So weigh short-term measures like the ROAS calculator together with repeat purchase data.

Meanwhile, reporting has its own trap: showing the rate only to management. Support, product and marketing each touch a different part of the loss. When everyone sees the same table, responsibility becomes clear.

How do you start tracking churn?

The first step is to write down your loss definition. In a subscription, it may be a cancellation or a missed renewal. In e-commerce, it may be more than a set number of days without an order. When the definition stays fixed, month-to-month comparison means something.

Second, collect six months of past data. Customer ID, first purchase date, last purchase date and acquisition channel are enough to begin. If your records are messy, plan the first week for cleanup.

1. Write the loss definition.
2. Calculate the monthly churn rate for the last six months.
3. Build the cohort and channel split.
4. Pick the biggest cause of loss.
5. Start one improvement experiment.
6. Compare the result with the next cohort.

You can finish these six steps in a month. When we set this up with clients, we fix the measurement first, then move to action. Because you cannot prove an improvement in something you do not measure.

Finally, set a rhythm. In a monthly meeting, look at only three questions: where did the rate go, what caused it, and which experiment do we run this month? A short, regular rhythm beats long, rare reports.