Average Churn Rate for Subscription Services: What's Good?

The average churn rate for subscription services in ecommerce sits in a narrower band than most round-up articles suggest, once the categories nobody selling recurring boxes actually competes with are stripped out. The two usable benchmarks below are not measured for supplement, skincare or food-and-beverage subscriptions specifically, and neither is independent research — both come from a billing or subscription-management platform reporting across its own merchant base. That distinction matters more than the gap between the two readings, and it is the first thing this piece is honest about before using either number. A subscription content or membership site — Substack, Patreon, a paid newsletter — runs different unit economics entirely and is not what either benchmark, or this article, measures; the figures and methods below are for a physical-product ecommerce subscription business specifically.

What Is the Average Churn Rate for Subscription Services?

The average churn rate for subscription services in ecommerce is 6.5% to 7.1% a month, depending on which of the two available benchmarks you use — Recurly’s cross-category consumer-goods figure at the low end, a Recharge DTC merchant panel at the high end, both vendor-reported. At 7.1%, the Recharge panel splits further into 4.1 percentage points of voluntary churn, a customer actively cancelling, and 3.0 points of involuntary churn, a declined card that nobody chose. That split matters for what happens next: a voluntary number points at the cancel flow and the product itself, an involuntary number points at the payment stack, and treating both the same way spends the fix on the wrong half of the problem.

Read the 6.5%–7.1% range as a band to sit inside, not a single target to hit exactly. Recurly’s number spans consumer goods broadly — supplements, apparel, home goods, beauty — while Recharge’s panel is built specifically from merchants running Shopify subscriptions through its own app, closer to what most ecommerce operators reading this actually run. Neither publisher has broken either figure out by price point, box cadence or product category, which is the gap the rest of this piece works through directly.

What Do the Published Churn Benchmarks Actually Measure?

The two benchmarks behind that 6.5%–7.1% range measure different populations over different windows — Recurly’s cross-category consumer-goods churn against Recharge’s DTC Shopify-subscription panel — and each source’s own population and window stays attached to its number rather than blending both into one figure that represents neither.

Every figure here is vendor-reported — a billing or subscription-management platform's own read of its merchant base, not an independent academic study. Neither Recurly nor Recharge discloses a sample size, which is why this table carries no confidence interval.

Both primary sources are vendor-reported for the same structural reason: no independent, peer-reviewed study of ecommerce subscription churn specifically could be found for this piece, and an SEO round-up citing “industry averages” with no named publisher is not a source — it is usually the same handful of vendor figures, including these two, laundered through a listicle with the attribution stripped off. Treat any churn statistic that cannot be traced back to one of these two names, or to a platform’s own published methodology page, as unsourced regardless of how confidently it is stated. The full retention-curve breakdown by billing cadence lives on the subscription churn benchmark page; this article works from the same two sources and adds the two things neither publishes — what the percentage costs in dollars for your own subscriber base, and whether price point or cadence moves the number.

One arithmetic note worth stating plainly, calculated from Recharge’s own 7.1% total churn figure and its 3.0-point involuntary share, and not itself a published figure: the Recharge panel’s own involuntary share — 3.0 divided by 7.1, or about 42% of its total churn — sits above the 20%–40% cross-industry range that Paddle and ProfitWell report separately. DTC ecommerce subscriptions may carry a higher involuntary share than subscription software or media, which would track with decline rates on debit-heavy consumer spending, but nobody has published that comparison directly. Read it as an observation from combining two existing benchmarks, not as a third one.

Is Your Churn Mostly Voluntary or Involuntary?

Whether your churn is mostly voluntary or involuntary changes which team owns the fix, and the Recharge panel’s own split — 4.1 points voluntary against 3.0 points involuntary, out of 7.1% total — is the only published starting point for guessing at your own mix before you have measured it directly.

A voluntary cancellation is a decision — too much product, a subscription forgotten and then noticed on a bank statement, a competitor’s offer, a life change. An involuntary loss is not a decision at all — a card expired, a bank declined an unusual recurring charge, an issuer’s fraud filter flagged the transaction — and no amount of product improvement or discounting touches it, because the subscriber never chose to leave.

Baremetrics reports involuntary losses at around 9% of monthly recurring revenue across the subscription businesses it measures (vendor-reported), noticeably higher in dollar terms than the Recharge panel’s 3.0-point share of a 7.1% total churn figure. The two numbers are not directly comparable — Baremetrics measures across all subscription business models including software, and reports a share of revenue rather than a share of churned subscribers — but the direction is consistent with the same point above: involuntary losses are a large enough share of total churn in every reading available that ignoring the payment stack while optimising only the cancel flow leaves real money on the table.

The practical split to run before fixing either voluntary or involuntary churn: filter your own cancellation reasons — Recharge, Skio and most subscription apps capture a reason code or a decline reason at the point of loss — by whether the subscriber initiated the cancellation or the charge simply failed and was never retried into a successful payment. If involuntary losses are a larger share of your number than the panel’s roughly 42%, the payment stack — retry schedule, account updater, processor mix — is the bigger lever; if voluntary losses dominate, the cancel flow and the box itself are.

How Do You Turn a Churn Percentage Into an Actual Dollar Number?

A churn percentage on its own tells you nothing about what it costs — turning 7.1% into a dollar figure needs your own subscriber count and your own average subscription value, because no published benchmark can multiply against a subscriber base it has never seen.

The dollar-math arithmetic is simple once you have your own subscriber count and average subscription value as inputs. Take a program with 1,000 active subscribers at an average $45 a month — both figures invented for illustration, not benchmark data. At the Recharge panel’s 7.1% monthly rate, that is roughly 71 subscribers lost this month (1,000 × 0.071), worth about $3,195 in monthly recurring revenue (71 × $45). Run the same subscriber count through Recurly’s lower 6.5% reading and the loss drops to about 65 subscribers and $2,925 — a $270 monthly difference from six-tenths of a percentage point of churn, which is why the 6.5%–7.1% range between Recurly and Recharge matters more than either single number on its own.

The monthly revenue lost to churn compounds, because a lost subscriber does not resubscribe the following month either — it is a permanently reduced base until new signups replace it, not a one-time deduction that resets. A subscriber lost in month one costs roughly eleven more months of revenue than an otherwise identical subscriber lost in month eleven of the same year, which is why churn concentrated early in the subscriber lifecycle is more expensive than the same churn percentage spread evenly across tenure, even though a single blended monthly rate cannot show that difference on its own.

The actual per-cohort revenue-loss figure for your specific subscriber base and price point is — metric to confirm — no publisher has calculated it in advance, because it depends on your subscriber count, your price and your churn rate together, none of which any benchmark can know before you supply them. The subscription churn calculator runs this arithmetic against your actual numbers rather than the illustrative ones above.

Do Churn Benchmarks Differ by Subscription Price Point or Box Cadence?

No published benchmark breaks ecommerce subscription churn out by price point or box cadence — the Recurly and Recharge figures above are both single blended numbers across whatever mix of price tiers and delivery schedules sits inside their respective merchant bases.

The absence of a price-point or cadence breakdown is not an oversight; it is a harder number to collect than the headline churn figure, because a billing platform would need to bucket its entire merchant base by price tier and cadence before it could report anything, and neither Recurly nor Recharge has published that cut. The reasoning for why it likely moves is straightforward even without the figure: a $15 monthly add-on and a $150 monthly curated box put a different amount of scrutiny on the subscriber at each renewal, and a four-week cadence exhausts a product at a different point in the billing cycle than an eight-week one, which changes when in that cycle a subscriber decides to cancel relative to when the box is actually empty. A fast-consumption category shipped on too slow a cadence produces a different churn pattern than the same category shipped too fast, and a blended industry number cannot separate the two.

The method that substitutes for the missing price-point-and-cadence benchmark is a cohort comparison inside your own data, not a borrowed industry number. Segment your own subscriber base by price tier and by cadence, then compare each segment’s churn at the same tenure — month three against month three, not calendar month against calendar month, since a segment launched more recently will always show a lower cumulative churn simply from having had less time to lose anyone. A cadence that runs ahead of actual consumption — a four-week box for a product that lasts six weeks — shows up in this comparison as higher early-tenure churn concentrated right after the second or third shipment, which a blended, cadence-agnostic churn number would never surface on its own.

Why Does Billing Cadence Change What “Average” Means?

Billing cadence changes what an average churn rate means because a monthly-billed subscriber and an annually-billed subscriber are measured against completely different renewal opportunities, and the Recharge panel’s own retention curve shows the size of that gap directly: about 28% twelve-month retention on monthly billing against about 62% on annual billing.

A monthly subscriber faces twelve separate cancellation opportunities in a year; an annual subscriber faces one, already paid for the next eleven months regardless of second thoughts. That is most of the 34-point gap between 28% and 62% twelve-month retention — not necessarily twelve times the product satisfaction, but twelve times the exit doors. A brand comparing its own monthly-only churn rate against a blended industry average that includes annual subscribers is comparing against a figure that structurally cannot apply, because it is averaging two different exposure windows into one number.

A quarterly prepaid option sits between monthly and annual billing and is worth naming separately, because it removes eight of the twelve renewal decisions rather than eleven, which should place its retention curve somewhere between the monthly and annual figures — neither Recurly nor Recharge publishes a quarterly reading, so that placement is reasoning rather than a sourced number.

The practical takeaway is to benchmark like against like: a monthly cohort’s churn rate against the Recharge panel’s monthly reading, not against a blended figure that assumes some share of annual billing it may not have. If your subscription program has no annual or prepaid option, the 6.5%–7.1% range in the earlier table is closer to what you are actually working against than any blended average that folds annual retention in — and adding a prepaid path is one of the more direct ways to move a monthly-only number toward the annual side of that curve, since it removes renewal decisions rather than trying to win each one individually.

What Should You Do If Your Churn Rate Is Above the Benchmark?

A churn rate above the 6.5%–7.1% range is worth investigating in a specific order — involuntary losses first, because they are the cheapest to fix and require no judgement call about the product, then voluntary losses, which need an actual reason code before a fix makes sense.

Start with the involuntary share, because a retry schedule and an account updater are configuration, not strategy: confirm your subscription app is actually running an account-updater service against expired and reissued cards, check how many retry attempts happen after a first decline and over what window, and compare that schedule against what Recharge, Skio or Bold Subscriptions document as their own default — a store that never changed the out-of-the-box setting is often running a shorter retry window than it assumes it is.

Only once involuntary churn is under control does the voluntary churn number mean anything on its own, because an unmeasured involuntary loss inflates the total churn figure and hides how the cancel flow is actually performing. From there, the reason code a subscriber gives at cancellation decides the next move: a cadence complaint gets a pause or a frequency change offered before a discount, a price objection gets a smaller size rather than a percentage off that trains every future renewal to expect one, and a product complaint gets routed to whoever owns the catalogue rather than absorbed as a problem the retention team cannot actually fix from its own side. Discounting every cancellation regardless of the stated reason is the fastest way to make next quarter’s blended churn number look better while making the underlying reason worse.

If both the voluntary and involuntary shares are higher than usual at once, fix the involuntary side first anyway and re-measure before touching the cancel flow — an inflated total makes the voluntary percentage look worse than it is, and a retention team that starts discounting against a number that is partly a payment-stack problem ends up buying back subscribers who were never actually trying to leave.

Neither the involuntary nor the voluntary fix is diagnosis-proof without ongoing reporting, because a churn rate measured once a quarter tells you where you stood, not which fix is working right now. Voluntary and involuntary churn need separate, continuously updated figures — not one blended number recalculated occasionally — because the fix for one does nothing for the other, and a single improving trend line can hide one getting worse while the other improves faster. That is a systems problem before it is a benchmark problem: a cancel flow that captures a real reason code, a retry schedule that is actually configured rather than left at its default, and a reporting layer that keeps the two churn numbers separate are the mechanics behind subscription retention, not a one-time comparison against Recurly’s or Recharge’s published range.