What it means
Every business with repeat customers has two engines of growth: acquiring new customers and keeping the ones it has. Acquisition gets the attention, because it is visible and easy to attribute to marketing spend, but retention usually determines whether growth compounds or leaks away.
A business that adds 1,000 customers a year and loses 900 grows by 100; a business that adds 1,000 and loses 300 grows by 700. Both have the same acquisition performance.
The difference is entirely retention, and it shows up in the numbers over years rather than quarters, which is why it is easy to neglect. The measure is defined carefully to separate the two engines.
The formula takes the customers at the end of the period, subtracts those acquired during the period, and divides by the customers at the start. The subtraction matters: without it, a business could hide poor retention behind strong acquisition.
The period must be chosen to suit the business. A subscription software company measures monthly and annually; a car dealer measures over years; a retailer defines a "customer" as someone who has bought within the last twelve months and measures whether they buy again in the next twelve.
Whatever the definition, it must be fixed and applied consistently, or the trend is meaningless. Retention can be measured by logo (the number of customers) or by revenue.
Logo retention treats every customer equally; revenue retention weights them by what they pay. The two diverge when the customers who leave are systematically larger or smaller than average.
Subscription businesses track gross revenue retention (opening recurring revenue less revenue lost from churned and downgraded customers, as a percentage of opening revenue) and net revenue retention (the same, but adding expansion revenue from customers who upgraded). Net revenue retention above 100% means the existing customer base grows on its own, before any new customer is added, which is the mark of a very strong business.
Cohort analysis is the most revealing way to look at retention. Instead of a single rate for the whole base, the customers acquired in each month or quarter are followed as a group, and the percentage still active after one, three, six, twelve and twenty-four months is plotted.
The curves show where customers are lost (usually early, in the first weeks or months, when the product has not yet become a habit) and whether later cohorts are retained better than earlier ones, which is the evidence that product and service improvements are working. A cohort curve that flattens at a healthy level shows a business with a durable core; one that keeps sliding shows a business that must keep buying customers to stand still.
The financial link runs through lifetime value. If churn is constant, the average customer lifetime is one divided by the churn rate: at 15% annual churn the average customer stays about 6.7 years, at 10% about 10 years.
Every point of retention therefore lengthens the stream of margin from each customer, increases the value of the existing base, and raises the amount the business can afford to spend to acquire the next customer. Retention is also cheaper to improve than acquisition in most businesses, because the levers (onboarding, product quality, service responsiveness, pricing fairness, proactive contact with customers showing signs of disengagement) are within the company's own control.
In practice
Real-world examples.
Example
A gym with 3,000 members at the start of the year signs 1,400 new members and ends with 3,200; retention is (3,200 minus 1,400) / 3,000 = 60%, and the owner realises that the marketing budget is replacing leavers rather than growing the club.
Example
A software company with 92% logo retention but 84% gross revenue retention discovers that its largest customers are the ones leaving, and redirects its account management effort towards them.
Example
An insurer measures retention at policy renewal and finds that customers who made a claim during the year and were handled well renew at a higher rate than those who never claimed, which changes its view of claims handling from a cost to an investment.
Think of it
“Retention rate shows what percentage of your customers stick around-your ability to keep them.
Formula
Calculation
Customer retention rate = (Customers at end of period minus Customers acquired during period) / Customers at start of period x 100
Churn rate = 100% minus Retention rate
Average customer lifetime (years) = 1 / Annual churn rate
Gross revenue retention = (Opening recurring revenue minus Revenue lost to churn and downgrades) / Opening recurring revenue
Net revenue retention = (Opening recurring revenue minus Revenue lost plus Expansion revenue) / Opening recurring revenue
Worked example. A subscription software company starts the year with 4,000 customers, acquires 1,200 during the year and ends with 4,600.
- Retained customers = 4,600 minus 1,200 = 3,400
- Retention rate = 3,400 / 4,000 = 85%; churn rate 15%; customers lost 600
- Average customer lifetime at 15% churn = 1 / 0.15 = 6.7 years
Revenue view. Opening annual recurring revenue was $12,000,000 (an average of $3,000 per customer). The 600 customers lost were smaller than average, with $1,500,000 of recurring revenue between them ($2,500 each); retained customers upgraded by $900,000.
- Gross revenue retention = ($12,000,000 minus $1,500,000) / $12,000,000 = 87.5%
- Net revenue retention = ($12,000,000 minus $1,500,000 plus $900,000) / $12,000,000 = 95%
Value of improvement. If retention rose to 90%, the company would lose 400 customers instead of 600, keeping 200 customers and about $500,000 of recurring revenue each year that would otherwise have been lost. At a 75% gross margin, each retained $3,000 customer contributes $2,250 a year. Lifetime value per customer at 15% churn is $2,250 x 6.7 = about $15,000; at 10% churn it is $2,250 x 10 = $22,500. The five-point improvement in retention raises the value of every customer, existing and future, by half.
Cohort example. The January cohort of 500 customers has 410 still active after twelve months: 82% twelve-month retention, compared with 76% for the January cohort of the previous year, evidence that the improved onboarding introduced in between is working.Case study
Seen in the real world.
A meal-kit subscription business grew quickly by heavy advertising and reported its customer numbers proudly each quarter. Its monthly retention rate was 92%, which sounded good to the board until the new finance director converted it to an annual figure: 0.92 to the twelfth power is 0.37, so only 37% of the customers acquired in any month were still subscribing a year later. With a customer acquisition cost of $95 and a monthly contribution of $22 per customer, the average customer lifetime of 12.5 months (1 / 0.08) produced lifetime contribution of about $275, a return that looked acceptable on paper but that vanished once overheads were allocated, and that depended entirely on advertising never becoming more expensive.
The cohort analysis showed where customers were lost: 30% cancelled within the first eight weeks, most citing the difficulty of skipping weeks or changing recipes. The company rebuilt its onboarding and its subscription controls, made skipping a week a single tap, and introduced a check-in message in week three for customers who had not yet rated a recipe. Monthly retention rose to 95%, which converted to 54% annual retention (0.95 to the twelfth power) and an average lifetime of 20 months.
Lifetime contribution per customer rose to about $440, a 60% increase, without any change to the acquisition cost. The board's quarterly pack was changed to lead with cohort retention curves rather than gross customer numbers, and the marketing budget was rebalanced towards retention for the first time. The finance director's observation was that the business had been pouring water into a leaking bucket, and that mending the bucket was worth more than any increase in the flow.
Watch out
Common mistakes.
- Counting new customers in the closing number without subtracting them, which lets strong acquisition disguise weak retention and overstates the rate.
- Reporting a monthly retention rate without converting it to an annual equivalent; 92% a month sounds healthy but means losing nearly two thirds of customers within a year.
- Tracking logo retention only, when the customers leaving may be the largest or the most profitable; revenue retention and cohort analysis show what logo retention hides.
Questions
People also ask.
How is retention rate different from churn rate?
They are complements: retention is the percentage kept, churn the percentage lost, and they add to 100% over the same period and definition. Businesses usually quote whichever sounds better, so read both.
What is a good retention rate?
It depends on the business. Enterprise software often retains 90% to 95% of customers a year; consumer subscriptions 60% to 80%; retail and hospitality far less. The useful comparison is with the company's own past cohorts and with direct competitors on the same definition.
Why does retention matter more than acquisition?
Because it compounds. Retained customers cost little to keep, buy more over time, refer others, and extend the period over which acquisition spending is repaid. A business with poor retention must keep spending to replace leavers and never builds a base.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%