What is a good customer retention rate?

It depends on your model. Recurly's July 2026 benchmark puts median annual churn for subscription SaaS businesses at 3.22% and ecommerce subscriptions at 4.25%, which means retention rates in the mid-to-high 90s annually for well-run subscription businesses. A one-time-purchase ecommerce store has a different baseline: a 25-30% repeat purchase rate is a reasonable target for most categories.

How do you calculate customer retention rate?

Use (E minus N) divided by S, times 100, where E is the number of customers at the end of a period, N is new customers acquired during that period, and S is the number of customers you started with. This isolates how many of your original customers stuck around, separate from new growth.

Is it cheaper to retain a customer than acquire a new one?

Directionally yes, because the acquisition spend on an existing customer is already sunk and they convert at a higher rate on a second purchase than a stranger does on a first one. We're not citing a specific multiplier here because we could not find a primary source for the commonly repeated 'five times cheaper' figure. Compare your own numbers: see our guide to customer acquisition cost for the formula.

What's the difference between SaaS and ecommerce retention strategies?

SaaS retention is mostly about usage: a customer who stops logging in is about to cancel, so the lever is health scoring and proactive outreach before renewal. Ecommerce retention is mostly about timing and relevance: getting the second purchase inside a realistic repurchase window and giving a specific reason to come back, not a generic reminder.

How often should you run a win-back campaign?

Trigger it off behavior, not a calendar. For ecommerce, a common window is 2-3x the average time between purchases for that customer segment. For SaaS, win-back usually starts the moment a cancellation is confirmed, with a second attempt 60-90 days later once the reason for leaving may have changed.

What is cohort retention and why does it matter more than one overall number?

Cohort retention tracks a single group of customers who signed up or bought in the same period, and measures what percentage of that exact group is still active at month 1, 3, 6 and 12. A blended, company-wide retention number can look stable while new cohorts are actually retaining worse, because strong old cohorts are propping up the average.

12 Customer Retention Strategies That Actually Move the Number

Customer retention strategies work in combination, not isolation. The companies that hold onto customers best run a specific program at each stage: a real onboarding sequence, a reason to buy again, a loyalty mechanic, a dunning process for subscriptions, fast service recovery and a win-back trigger. None of those on their own moves the number much. Together, they do.

This post gives you 14 strategies grouped by stage, the formulas for measuring retention correctly, and what changes between ecommerce and SaaS.

Key takeaways

Retention strategies only work grouped by lifecycle stage: onboarding, post-purchase, loyalty, subscriptions, service and win-back each need a different tactic.

Measure retention by cohort, not one blended number. A single company-wide retention rate can hide a cohort that’s quietly getting worse.

For subscription businesses, Recurly’s July 2026 data puts SaaS churn at 3.22% annual and ecommerce subscriptions at 4.25%, with involuntary churn (failed payments) making up roughly a third of both.

Involuntary churn, meaning failed card payments, is often the single biggest fixable leak in a subscription business, and it has nothing to do with whether customers like your product.

Don’t quote the “5x cheaper to retain” stat. We looked for a primary source and couldn’t confirm the number; use your own CAC and repeat-purchase data instead.

How to measure retention correctly

Three formulas cover most of what you need.

Customer retention rate

Retention rate = ((E - N) / S) x 100

E = customers at the end of the period. N = new customers acquired during the period. S = customers you started with. Subtracting N matters: without it, new growth can disguise churn in your old customer base.

Churn rate

Churn rate = 1 - retention rate

Or directly: customers lost during the period, divided by customers at the start. For subscription businesses, split this into voluntary churn (customers who chose to cancel) and involuntary churn (failed payments, expired cards). They need different fixes.

Cohort retention

Group customers by the month they first bought or signed up. Track what percentage of that exact group is still active at 30, 90, 180 and 365 days. Plot it as a curve, not a single number.

Metric	What it tells you	Where it’s most useful
Retention rate	Overall health, period over period	Board reporting
Churn rate (voluntary vs. involuntary)	Where you’re bleeding customers and why	Subscription businesses
Cohort retention curve	Whether newer customers behave differently than older ones	Diagnosing a recent problem
Repeat purchase rate	Share of customers who bought more than once	Ecommerce, one-time purchase models

A retention funnel showing onboarding, post-purchase, loyalty, subscription, service and win-back stages feeding into a cohort retention curve

A worked example: why the blended number lies to you

Say a subscription box company has 10,000 customers at the start of the year, adds 4,000 new customers over the year, and ends with 11,500 customers.

Retention rate = ((11,500 - 4,000) / 10,000) x 100 = 75%. Churn rate = 25%.

That 75% looks fine against the industry median. But split it by cohort and the picture changes. Customers who joined 18+ months ago are retaining at 85%. Customers who joined in the last two quarters are retaining at 60%. The blended number of 75% is an average of a healthy old base and a newer base that’s leaking, and nothing in the single company-wide figure would have told you that. Only the cohort curve does.

This is why a retention dashboard needs at least two views: one trailing number for the board, and one cohort table, refreshed monthly, for the team that actually has to fix anything.

14 retention strategies, grouped by stage

Onboarding

1. Build a welcome sequence around one first action, not a tour. The goal of onboarding email isn’t to explain every feature. It’s to get the customer to do the one thing that predicts they’ll stick around: log a second order, invite a teammate, complete setup. Our welcome email sequence guide covers the structure.

2. Add a proactive check-in at day 7, 14 and 30 for SaaS. Don’t wait for a support ticket. A short email or in-app message asking “did you get X set up” catches confusion before it becomes a cancellation. Say a project management tool notices a new account hasn’t invited a teammate by day 7. A single email nudging that one action, with a two-click invite link, does more for retention than a feature announcement ever will, because it targets the specific gap instead of broadcasting to everyone the same way.

Post-purchase

3. Follow up on delivery with usage guidance, not a review request. The first email after an order arrives should help the customer get value from what they bought. A skincare brand might send application tips. A software signup might send a 2-minute setup checklist.

4. Time the review request to actual usage, not order date. Asking for a review the day an order ships gets a rating on the unboxing, not the product. Wait until there’s been time to use it.

Loyalty

5. Run a points or tiered program that rewards the second purchase, not just volume. Say a $40 skincare brand gives a small credit on a customer’s second order inside 60 days, separate from its main loyalty tier. That’s a retention mechanic, not a loyalty mechanic, and it’s worth testing before building the full tiered version. See loyalty program ideas for structures that work at different budget levels.

6. Add a referral incentive that rewards both sides. Customers who refer someone and get rewarded for it are more invested in staying. Our referral program ideas post has templates by business type.

7. Layer channels instead of relying on email alone. Per a 2019 Omnisend analysis of its customer base, marketers using three or more channels saw a 66% retention rate against 34.8% for marketers using a single channel. That data is dated, so treat it as a directional argument for diversifying channels, not a number to plan around.

Subscriptions

8. Fix involuntary churn before you touch voluntary churn. Recurly’s July 2026 benchmarks show involuntary churn (failed payments) making up roughly a third to a half of total churn across every industry it tracks. Card retry logic, updated-card prompts and dunning emails recover revenue with zero product or pricing changes.

9. Give subscribers a pause option before they cancel. A customer who pauses for two months is still a customer. A customer who cancels because pausing wasn’t offered is gone, and reacquiring them costs more than the pause would have. Put the pause option on the cancellation page itself, not three menus deep. If a customer has to search for how to pause, they’ll just hit cancel instead.

Service

10. Resolve the first contact instead of escalating by default. A customer who has to repeat their issue to a second agent is measurably more likely to churn. Track first-contact resolution as a retention metric, not just a support metric.

11. Build a simple health score and act on drops. For SaaS, this can be as basic as login frequency plus feature usage. When a healthy account’s score drops two tiers in a month, that’s a retention trigger, not a data point to review next quarter. You don’t need a data science team for version one. A spreadsheet that flags “logins down 50% month over month” and routes that account to a human is a health score. Build the dashboard once the manual version proves the alert is worth acting on.

Win-back

12. Trigger win-back off behavior, not a fixed calendar date. For ecommerce, that means watching when a customer’s typical repurchase window has passed by 2-3x. For SaaS, start the clock at cancellation, not three months later.

13. Ask why they left, and change something because of the answer. A one-question survey at cancellation (“what made you decide to cancel”) that never changes product decisions is just data collection. Close the loop with at least one visible fix a quarter.

Across the lifecycle

14. Put one person’s name on the retention number. Acquisition almost always has an owner. Retention often doesn’t, and metrics without an owner don’t move. Pair this with the customer acquisition cost you’re already tracking so retention and acquisition get judged against the same dashboard.

Ecommerce vs. SaaS: the real differences

	Ecommerce	SaaS
Primary churn signal	Time since last order	Drop in login/feature usage
Median annual churn (Recurly, July 2026)	4.25% (subscription ecommerce)	3.22%
Biggest controllable lever	Repeat purchase timing, loyalty	Involuntary churn, onboarding activation
Typical retention metric	Repeat purchase rate	Net revenue retention, logo churn
Win-back trigger	2-3x average reorder window	Cancellation date

Repeat purchase rate for ecommerce is simple: customers with 2 or more orders, divided by total customers, in a given period. It’s the single most useful number for a store that doesn’t run subscriptions, because it tells you whether the business is growing on new customers alone, which is expensive and fragile, or on a mix that includes people coming back.

The category matters more than most people assume. A skincare brand selling a 60-day moisturizer has a natural repurchase window built into the product itself. A furniture store doesn’t, and chasing a high repeat purchase rate there usually means pushing the wrong second product instead of accepting that retention for that category looks more like reviews, referrals and brand recall than repeat transactions in month two.

A 30-day retention audit template

Run this before you add a new tactic. Half the time, the fix is something already broken, not something missing.

Area	Check	Owner
Onboarding	Is there a welcome sequence tied to one first action?
Post-purchase	Does the first post-order email help them use the product?
Loyalty	Is there any mechanic rewarding a second purchase inside 60 days?
Subscriptions	Is card-failure recovery automated, or manual?
Service	Is first-contact resolution tracked at all?
Win-back	Is the trigger behavior-based or calendar-based?
Reporting	Is retention tracked by cohort, or one blended number?

The bottom line

Retention strategies fail when they’re treated as one initiative, usually a loyalty program, bolted on and left alone. They work when each lifecycle stage gets its own small, specific tactic, and when you’re measuring cohorts instead of a single company-wide average that can hide a declining newer group behind a strong older one.

Start with the audit table. Fix what’s broken before you add what’s missing.