Churn rate: how to calculate it and genuinely reduce customer attrition

What churn rate is and how to calculate it without mistakes

Churn rate is the share of customers who stopped using a service or product in a given period, relative to all customers at the start of that period. In B2B it is usually calculated monthly or annually, and the difference between these two periods can be misleading, because a monthly churn of 3% looks harmless until you convert it to a yearly figure.

The basic formula looks like this:

Churn rate = (number of customers lost in the period / number of customers at the start of the period) × 100%

Example: at the start of the month you have 80 customers, and 4 left during the month. Churn is 4/80 = 5%. Simple. Now the same number in practice: a customer signed a contract in January, stopped replying to emails in March, and submitted a termination notice in April. Nobody on the team noticed that they had already stopped logging into the panel in February. Formally they left in April, but in reality they disengaged two months earlier. The problem starts with interpretation, because the same formula leads to completely different conclusions depending on what you count as a "lost customer".

Three versions of churn worth calculating separately

• Customer churn (logo churn) – how many customers left, regardless of how much they paid. Good for assessing the work of the support team and the quality of relationships.
• Revenue churn – what percentage of revenue left along with the customers. If you lose three small customers and one large one, logo churn will show 4, while revenue churn may show 40%.
• Net revenue churn – takes into account expansions among existing customers. If a customer worth 10 thousand left, but two others increased their scope by 15 thousand, net churn is negative. This is the metric that most honestly shows whether the company is shrinking or growing.

In service companies, including marketing agencies, there is one more problem: a customer has not formally left, but has reduced the scope to a minimum and is waiting for the contract to end. This "silent churn" will not appear in any of the formulas above until you look at the engagement of the customer's team on their side: the number of replies to emails, attendance at meetings, and the pace of approving proposals.

If you want to compare churn across months, settle on one definition and stick to it for at least four quarters. Changing the definition halfway through the year is the easiest way to show the board an improvement that does not exist.

What actually drives customer churn in B2B

Customer churn rarely rises for a single reason. In practice, four groups of factors overlap, and only by separating them can you plan effective action.

1. Factors on the sales promise side

If the sales team promises results that operations cannot deliver, churn is only a matter of time. In B2B this shows up most clearly in outcome-based selling: the client buys a specific number of leads or revenue growth, and two months later reality looks different. The client then leaves not because of a bad service, but because of a gap between expectation and delivery.

2. Factors on the onboarding side

The first 30 to 60 days determine whether the client stays longer. If implementation drags on for weeks, the client sees no early results and starts to doubt. In agencies, the most common mistake is having no single person responsible for onboarding, so the client talks to five different specialists and does not know who makes the decisions.

3. Factors on the perceived value side

Clients judge not how much work the team did, but what that work produced. A report with 40 hours of work and not a single sentence about business impact sounds like an excuse, not a result. It is worth showing value in the client's language: new inquiries, shorter handling time, lower acquisition cost.

4. Factors on the client-side change side

Some churn is outside your control: the client changes strategy, loses budget, gets acquired. That does not mean you cannot influence it. Regular contact with the decision-maker, not just the operational contact, lets you catch signals of shifting priorities before a termination notice appears.

A well-designed B2B marketing strategy accounts for these four areas at the planning stage, not only when the client announces they are leaving.

Realistic churn rate ranges in B2B companies

There is no single number that is a "good churn" for everyone. Ranges vary widely between industries and billing models, so every value needs to be set against a specific context.

The ranges below are indicative estimates based on market practice and conversations with service and SaaS companies, not data from a single study. Treat them as a starting point for your own comparisons, not as hard benchmarks.

Treat these values as a reference point, not a target. A company with 4% churn in a segment where 1% is typical has a problem. A company with 4% churn where 6% is typical may have a healthy model. What matters is comparing against your own result from a year ago and against a specific customer segment, not against the overall market average.

Annual churn is a separate matter. If monthly churn is 3%, annual churn is not 36%, it is around 31% (because you are calculating from an ever-smaller base). At 5% monthly, annual churn is already around 46%. This shows why monthly figures around 5% are alarming in B2B, even if they look innocent at first glance.

How to reduce churn: actions you can implement within a quarter

Effective churn reduction does not rely on one big project, but on several parallel changes in the process. Below are specific actions that deliver measurable results in service companies within one to two quarters.

1. Introduce a customer health score

Collect 5-7 signals that most often precede a departure: a drop in email response rates, no logins to the tool, delays in approving proposals, a change of contact person, shorter meetings. Assign a weight to each and calculate the score weekly. A customer with a low health score goes on the proactive outreach list before they raise a problem themselves.

2. Separate onboarding from ongoing service

The first 60 days should have a separate process, separate goals, and separate reporting. The customer needs to see the first result quickly, even if it is small. In an agency, this often means changing the order of work: a quick win first, then building the foundations.

3. Review meetings with the decision-maker, not just the operational contact

Once a quarter, talk to the person who signs the contract about what has changed in their business. This is not a sales meeting. It is a meeting where you learn about a shift in priorities before it becomes a reason to end the collaboration.

4. Respond to signals within 48 hours

A customer who reports a problem and waits a week for a response has already made up their mind. Speed of response is cheaper than any rescue efforts undertaken after the contract is terminated.

5. Measure churn by segment

A single overall metric says little. Break it down by industry, customer size, acquisition channel, and account manager. It often turns out that the problem is not evenly distributed: one group of customers leaves three times more often than the rest, and the cause lies in how they were acquired.

It is also worth calculating how much it costs to acquire a customer relative to their lifetime value. If acquisition costs only pay back after 18 months while the average relationship lasts 14 months, no sales optimization will fix the model. The relationship has to be extended first.

Churn and business scaling: when the metric starts blocking growth

In the early stage, a company grows because it acquires more customers than it loses. At some point this mechanism stalls: acquisition costs rise, and churn eats up an ever larger share of new revenue. This is the moment when further investment in sales without work on retention stops paying off.

A practical rule: if monthly churn exceeds 5% in a service model, scaling sales usually deepens the problem rather than solving it. The sales team delivers new customers, but operations cannot retain them with the current process. Fix retention first, then increase spending on acquisition.

It is also worth looking at churn through the lens of time. A customer who leaves after 3 months generates a loss. A customer who leaves after 3 years is a normal occurrence. If the average relationship length drops year over year, it signals that something has changed in the quality of the value delivered, even if the churn rate itself looks stable.

A good control question is: how many customers who left in the last quarter had contact with the decision-maker within the last 60 days before leaving? If the answer is "few", the problem lies in the process, not the product.