How to reduce
subscription churn.
Subscription churn is not one number. It is a handful of moments: the first days, the first charge, every renewal after it, a card that fails, a cupboard that fills up. We build an answer for each one, in your subscription app and in Klaviyo, and measure it on renewals kept.
Where subscription churn actually happens.
Before you reduce subscription churn, find it. Our churn research followed subscribers order by order through the stores we run and audit. At the typical store:
never place a second order. The first stretch holds the biggest single loss.
of those who renew once never renew again. Every renewal costs about a quarter to a third, so one save offer cannot carry the plan.
of cancellations come before the first renewal charge, and 3% on the day the subscription starts.
cancel because they have too much product, ahead of price at 19%: a delivery problem before a loyalty one.
Sell more upfront.
The cheapest renewal to keep is the one the first order already paid for. Every unit, every pack size and every month of plan sold at checkout is a renewal moment that never happens, and so a cancellation that never happens.
The plan as the default
The subscription pre-selected in the buy box, with the saving shown against the one-time price. In our subscribe and save research 69% of brands pre-select it and the median discount is 15%.
Bigger packs and longer plans
Two and three unit packs and a quarterly plan offered on the main path, not hidden on one product page. Fewer billing cycles mean fewer moments to cancel.
A discount you can afford at renewal
A deep first-order discount sells the first order and turns the second into a price rise. Size the saving on the renewal price, and if the first order gets its own offer, say what renews and when.
Upgrades at checkout and after it
The switch to the plan offered in the cart, at checkout and on the thank-you page, and to one-time buyers on the day their order runs out.
The price justified after the sale
“Too expensive” is often doubt about whether it works. The first emails earn the price with the quality, the science and the proof, not only the ads.
How the offer sets the first renewal is on our subscription offer page.
A cancel flow for every stage.
A subscriber who cancels on day two and one who cancels after a year are not leaving for the same reason, so they should not see the same flow. Each leaver is routed once, to the first group that fits, in this order:
| Segment | Why they leave | What the flow does |
|---|---|---|
| In payment recovery | Their card failed; many give “too expensive” or “other” | Asks for a card update. A discount on a card that cannot be charged saves nothing |
| Days 1 to 5 | Did not mean to subscribe, or bought for the first-order offer | The terms plainly, a one-time purchase instead, the first delivery date. Never an argument |
| Saved before | The last save did not answer the real reason | A different answer from last time, never the same offer twice |
| 0 renewals | Has not tried it long enough; too much product before the first charge | What to expect and when, then a later first renewal or a longer interval |
| 1 renewal | Still waiting for results; the second charge feels like a decision | Where they are on the timeline, a skip or a push of one, two or three weeks |
| 2 renewals | Stock builds up; the habit is not set | A longer interval or a smaller pack, the next reward on the ladder |
| 3 renewals | Loyal but bored, or the price adds up | The reward they would lose, a swap, the quarterly plan |
A page before the reasons
Each segment first sees one page written for its stage, with a picture or a video and no offer on it. Most leavers have not thought about what they lose.
Reason, fix, then offer
The reason first; then a fix with no incentive (a skip, a push, a new interval); an offer only if the fix is refused. “I do not want a subscription” gets a one-time purchase, not another plan.
Lead with the deferral that pays
Skip, push, longer interval or pause: read which one leads to more revenue later, put it first, and confirm every change with the new delivery date.
Judge saves on the next renewal
At the typical store the cancel flow saves 16% of the subscribers who enter it. A save only counts once that subscriber pays one more renewal.
A rewards journey they can see.
Every renewal loses about a quarter to a third of the subscribers who reached it, so the reason to stay has to grow with every order. A rewards journey gives each renewal its own gift, keyed on the order number, with the biggest just before the order where your subscribers drop most.
A ladder by order
A gift or a credit on set orders, named by the order (“your third order”), never by the week. A free product where doubt about results is the objection; a credit on the plan rather than a code.
Shown in the billing reminder
Every reminder names this order’s gift and the next one, one rung ahead.
Shown in the cancel flow
What they lose by leaving, in their own numbers: the gift on the next order, the rung they reached, the price they keep.
Shown in the portal
The whole ladder on every visit, with where they stand and the line that cancelling starts it again.
Long tenure as identity
After the ladder ends, the subscribers who stayed are told what that makes them. “Most people never make it this far” does more than another code.
Involuntary churn, by decline type.
Involuntary churn starts with a failed payment. One scheduled charge in 7 fails on the first attempt at the typical store, and failed payments end 9% of cancellations where the app separates them. Some subscribers see the failed-payment email and cancel themselves, so the reports count them as price churn. A single “update your card” email every three days treats every failure the same; the bank’s reason says they are not.
| Decline | What it means | Retries | What the subscriber is asked |
|---|---|---|---|
| Soft | Insufficient funds, a limit, a temporary refusal. The card works; the timing did not | Several silent retries timed around paydays, before any email | “We will try again on [date]”, not “update your card” |
| Hard | Expired, lost, closed or revoked. Retrying never works | None worth making | Update the card, by email and text the same day |
| Mixed | Reasons that recover by retry only some of the time, such as a generic “do not honor” | A short run of retries while the ask goes out | Update the card, while we keep trying |
Read the reason first
The decline reason sits on each billing attempt in the store and often never reaches the email tool. Class each reason by how it actually recovered, not by its name.
Three emails, not one per retry
The ask, then the reminder, then what they lose: the reward, the price, the rung. The button on the first screen of a phone, three lines of copy.
Pause, never cancel
The last retry pauses the subscription so it can restart in one tap. An incentive, if any, comes late, never on the first ask.
of failed payments come back in the first round of retries at the typical store, from 18% to 70% across stores. The gap is settings.
A portal that sells the next order.
The portal is where subscribers go when something is wrong. If all it offers is “cancel”, that is what they do. The most common reason given for cancelling is too much product (24%), and the portal answers it without a discount.
Every way to stay, one tap away
Skip, push by one, two or three weeks, change the interval, swap, add one, pause. The same controls in every email near a charge, not only a link to the account page.
Reviews
What other subscribers say about the product they are about to stop, at the moment they doubt it.
Education
The questions that come before a cancellation, answered in place: how to use it, when to expect results, short videos and a plain FAQ.
Different for each product
The portal changes with what they subscribe to: its own timeline, its own tips, its own rewards and its own add-ons.
The lifecycle between orders.
28% of new subscribers never place a second order at the typical store. What happens between the first order and the first charge decides more than any campaign.
The upgrade to quarterly
Offered to subscribers on one unit who have not renewed yet, after the core onboarding, a week before the first charge: the saving per unit first, the old and the new price, one tap to switch. Fewer billing cycles, fewer cancel moments. Test it first on the subscribers who said price worried them, with a holdout.
The billing reminder as a gift
The subject is what this order brings, never the charge. It says when the next order arrives, not only when it ships, because a box that lands too soon after the last is why people skip. Skip, push and pause sit in the email, cancel is never named.
Transactional emails with a moment in them
The order confirmation, the shipping and the delivery emails are the most opened emails a brand sends. Each carries one small delight: a note from the founder, a tip for day one, what the gift on the next order will be.
Why other solutions failed
After delivery, the mechanism of the problem: why what they tried before did not work, so the doubt is answered before it comes.
Why this one will work
The mechanism of the product, and an honest timeline of when to notice what, with the first side effect named before it happens.
Compliance tips
Habit stacking: put it next to something that already happens every day, “with your morning coffee”. A product that is not used does not renew.
Usage tips
The uses that make it work better or go further: “apply the cream to your décolletage too, not just your neck”.
Measure subscription churn on what moved.
Retention emails report revenue they did not earn: a renewal would have been charged anyway. Read every change on renewals kept and cancellations avoided, against a group that did not get it.
Incrementality testing
One group gets nothing, the other gets your marketing; the difference is what the marketing earns. Start with the post-purchase journey and let it run for months, not weeks. It is often uncomfortable: the lift can be small, and sometimes negative. A small lift says take bigger swings.
Customer feedback loops
A few questions at the moments that predict a cancellation, each answer stored on the profile and acted on in the flow:
| When | What we ask | What happens with the answer |
|---|---|---|
| After the order | Their goal, their worry (price, results, taste, the habit), what they used before | Each worry gets its answer in onboarding; price worries go into the upgrade test |
| Around day 14 | “Do you feel anything yet?” Yes, a little, not yet, in one tap | Yes asks for a review; not yet gets the fixes and the guarantee |
| Before the first charge | Do you feel the benefits? | Low answers go to help, never to the cancel page |
| After a few orders | Are you happy? Stars as the button | Happy goes to a public review, unhappy to a private form |
| At cancellation | The reason, required, and their own words | Routes the save, and the words are read every month |
Documenting learnings
Every test written down: the date, the change, the result, what it taught. The record is how a team, and any agency after it, shares one picture of who the customer is, and stops relearning the same lesson.
Subscription retention strategies we build.
Churn read
Your subscribers order by order against our benchmarks, the reasons they give, where your failed payments go, and the order to fix things in.
Offer and plans
The default, the packs, the quarterly plan and the discount, set up in Loop, Recharge or Skio.
Cancel flow and portal
The segments, the pages, the reasons and the offers, and the portal controls.
Rewards journey
The ladder by order, in the reminder, the cancel flow and the portal.
Payment recovery
Retries and emails by decline type, in your subscription app and in Klaviyo.
Lifecycle and measurement
The upgrade, the reminder, the post-purchase journey, the feedback questions and the holdout, with every test written down.
Proof on named accounts.
subscription churn, Frøya Organics. Read the case study
quarterly growth from the subscription engine, Zapply UK. Read the case study
Real client accounts, named and linked. All case studies.
Our email and campaign conversion is up 4× since we started. Your team just gets the women’s-health customer.
Repeat customer revenue is up like 50%. So that’s really good. There’s lots of positive things from last month.
We really saw a huge impact, especially in the beginning - a huge boost, both in revenue and revenue share. You guys are true professionals, you have a very creative team.
Super happy. Best retention agency in the business.
Find where your subscribers leave.
We read your subscription app and your orders against our benchmarks and show which moment costs you the most. Free, run by the founder, no contract.
Questions founders ask.
How do you reduce subscription churn?
Find the moment it happens first, then answer each one: sell more upfront, route every cancellation to a flow for its stage, give each renewal a reward, recover failed payments by decline type, make the portal answer the reason before the cancel button does, and measure it all against a holdout.
What is a good subscription churn rate?
Count it by orders, not months: at the typical store 72% of new subscribers place a second order and 20% reach a sixth. A store that bills often asks the stay-or-go question more often, so a monthly rate flatters quarterly plans.
When do most subscribers cancel?
Early. At the typical store 38% of cancellations come before the first renewal charge, and every renewal after that loses about a quarter to a third of the subscribers who reached it.
Should a cancel flow offer a discount?
Last, not first. Ask the reason, offer a fix with no incentive (a skip, a push, a longer interval), and only then an offer. Judge the flow on whether saved subscribers pay their next renewal, not on the save rate.
Skip or pause: which should come first?
The one that brings more revenue back later in your store. Read it, put it first in the portal, the reminder and the cancel flow, and confirm every change with the new delivery date.
How do you recover failed subscription payments?
By decline type. Soft declines get silent retries timed around paydays before any email; hard declines get a card-update ask the same day; mixed declines get both. The last retry pauses the subscription instead of cancelling it.
How do you know if retention marketing is working?
Hold a random group out of it and compare renewals and cancellations. Opens and clicks do not say whether a renewal would have happened anyway.
Keep reading.
Subscription churn benchmarks
Where subscribers leave and why, order by order, with the reasons they give.
Subscription ecommerce statistics
Revenue mix, retention by order and failed payments at the typical store.
Subscription offer strategy
How the buy box sets how many shoppers subscribe and how many renew.
Your churn,
taken apart.
We read your account against these benchmarks and give you the list of what to change, in the order it pays.
