Retention Rate Formula: How to Calculate CRR

Retention rate = ((customers at the end of the period minus new customers acquired during the period) / customers at the start of the period) x 100. Start a quarter with 200 customers, finish with 195, and win 15 new ones along the way, and your retention rate is ((195 - 15) / 200) x 100, which is 90 percent. The step almost everyone skips is subtracting the new customers, and skipping it is what turns a mediocre quarter into a flattering one.

Retention gets treated as a customer success metric and reported once a quarter in a slide nobody argues with. That undersells it. Retention is the single largest input to customer lifetime value, which means it silently sets the ceiling on what you are allowed to spend acquiring anyone. This guide covers the formula, the worked arithmetic, the revenue-based variants that SaaS boards actually ask for, and the three ways the number gets quietly inflated.

The retention rate formulas, and what each one answers

There is no single retention rate. There is a family of them, and picking the wrong member of the family is how two people in the same meeting quote different numbers for the same quarter.

How do you calculate customer retention rate?

Pick a period, then count three things: distinct customers at the start (S), distinct customers at the end (E), and customers acquired during the period (N). Retention rate is ((E - N) / S) x 100. Everything hard about this is in the counting, not the arithmetic.

Work an example. A B2B software company begins Q1 with 200 paying accounts. During the quarter it signs 15 new accounts and 20 of the original accounts cancel. It ends with 195 accounts. The naive calculation, 195 divided by 200, gives 97.5 percent and is wrong, because 15 of those 195 accounts were never part of the original group. The correct figure is ((195 - 15) / 200) x 100 = 90 percent, and the matching churn rate is 10 percent for the quarter.

Two rules keep this consistent. Use the same period length every time, because a quarterly retention rate is not comparable to a monthly one and cannot simply be divided by three. And count distinct customers, not orders or seats, unless you are deliberately measuring revenue retention instead.

What is the difference between retention rate and churn rate?

They are the same fact stated two ways. For an identical period and an identical customer basis, churn rate equals 100 minus retention rate, so 90 percent retention is 10 percent churn. Neither is more accurate than the other.

Churn is the more useful of the two for financial modeling, because customer lifespan is 1 divided by the churn rate. A 10 percent quarterly churn rate is roughly 34 percent annually, which gives a lifespan near 2.9 years, and that lifespan is the input the customer lifetime value formula needs. Retention is the more useful framing for a team, because people work harder to keep something than to reduce a loss percentage.

The confusion starts when one number is measured on customers and the other on revenue. A business can lose 8 percent of its accounts while growing revenue, if the accounts it loses are small and the ones it keeps expand. Logo churn of 8 percent alongside net revenue retention of 105 percent is not a contradiction. It is a description of a business that is losing its worst customers.

Net revenue retention and gross revenue retention

Subscription businesses are usually judged on revenue retention rather than customer counts, because a customer who halves their plan has not churned but has still cost you money. Gross revenue retention counts only the losses. Net revenue retention adds expansion back in, which is why it can exceed 100 percent.

The arithmetic: GRR is (100,000 - 4,000 - 5,000) / 100,000 = 91 percent. NRR is (100,000 + 12,000 - 4,000 - 5,000) / 100,000 = 103 percent. Both describe the same quarter honestly, and reporting only the flattering one is the most common form of retention theater. GRR tells you how leaky the bucket is. NRR tells you whether the tap is running faster than the leak.

New revenue from brand new customers never belongs in either figure. It is acquisition, not retention, and folding it in produces a number that looks like retention and behaves like a growth rate.

How do you calculate repeat purchase rate?

Repeat purchase rate is the share of customers who have bought more than once: customers with two or more orders divided by total customers, times 100. It is the natural retention measure for ecommerce, where there is no subscription to cancel and a customer never formally leaves, they simply stop coming back.

The measurement window is the whole argument. Calculated across all time, repeat purchase rate always rises, because customers acquired years ago have had years to buy again while last month's cohort has had a month. That produces a number that improves on its own and tells you nothing. Fix the window instead: measure the share of a given month's cohort that placed a second order within 90 days, and compare cohort to cohort. Now a change means something.

What is a good customer retention rate?

There is no useful cross-industry answer, and most published averages compare businesses with incompatible period lengths and customer definitions. A monthly consumables brand and an annual enterprise contract will produce wildly different percentages while being equally healthy.

The question worth asking is what retention does to your unit economics. Retention sets customer lifespan, lifespan sets lifetime value, and lifetime value divided by customer acquisition cost gives the ratio that actually decides whether you can keep spending. Run the arithmetic and the leverage is obvious. Take an $85 order value bought 3.2 times a year at 60 percent margin. At 40 percent annual churn the lifespan is 2.5 years and lifetime value is $408. Cut churn to 30 percent and the lifespan becomes 3.33 years, lifting lifetime value to $544, a 33 percent gain from a ten point retention improvement and no change whatsoever to acquisition.

That is why retention beats almost every acquisition lever available to you. Ten points of retention permanently raises the ceiling on what you can afford to bid; ten percent better ad creative raises it for as long as the creative lasts. The full reading of the resulting ratio is in the LTV:CAC ratio and what good looks like, and the cash timing question sits in CAC payback period.

Three ways retention rate gets inflated

Leaving new customers in the numerator. Dividing ending customers by starting customers without subtracting acquisitions turned a real 90 percent into 97.5 percent in the example above. In a fast-growing business the distortion is larger than the churn being measured, which means the faster you grow the healthier your retention appears.

Switching period length when the number looks bad. Monthly retention of 95 percent sounds excellent and compounds to roughly 54 percent over a year. Quoting the monthly figure to a board while modeling lifetime value on the annual one produces a lifetime value that is roughly double reality.

Reporting NRR and calling it retention. Net revenue retention above 100 percent is a genuine achievement, but it can hide a business losing a third of its logos every year while a handful of large accounts expand. Report GRR beside it. If the gap between them is wide, your growth depends on a small number of accounts that are themselves a concentration risk.

Where the retention number should actually live

Retention is usually calculated in a spreadsheet by whoever owns customer success, while acquisition cost is calculated separately by whoever owns marketing. The two numbers then meet once a quarter in a deck, by which time both are describing a business that has moved on. Meanwhile the ratio between them, the one that governs how much you can spend, has been drifting the whole time.

When retention slips, the cause is rarely in the marketing at all. It sits in onboarding, billing, and the support operations that decide whether a customer sticks after the first invoice. Marketing sees the symptom as a rising acquisition cost that never seems to pay back, which sends teams to optimize ad creative when the leak is somewhere else entirely.

Putting retention on the same board as spend and revenue fixes the diagnosis. MixedMetrics connects Shopify, Stripe, Klaviyo, GA4, Google Ads, Meta, and TikTok read-only, then computes blended ROAS, blended CAC, MER, lifetime value, and the LTV:CAC ratio from the same reconciled figures. Retention stops being a quarterly slide and becomes an input you can watch move. The wider set of numbers worth tracking together is laid out in the marketing KPIs that decide budget.

The short version

Use ((E - N) / S) x 100 for customers, and the revenue variants for subscriptions, always reporting gross revenue retention alongside net. Keep the period length fixed, count distinct customers, and measure by acquisition cohort where you can. Then stop judging the percentage against someone else's average and judge it against your own lifetime value: retention is worth measuring precisely because it is the cheapest way to raise the number that decides how much you can afford to spend.