By George Kapernaros, founder of YOCTO, a retention agency for subscription and DTC brands, named Klaviyo Elite Master, 2025-2026.
Every subscription brand I work with reviews cancellations weekly. Almost none of them review deferrals at all. That is backwards. Across the 100+ subscription accounts my team analyzed this year, deferrals are often a bigger problem than cancellations. On some accounts they cost four times as much.
What a Deferral Actually Is
A deferral is a delayed payment. A customer skips the next box, pauses for a month, or pushes the billing date out to something that suits them better.
It slips under the radar because it is not a cancellation. The subscriber still counts as active. The churn report stays clean. No save flow fires, no win-back email goes out, and nothing in the Monday numbers looks wrong. The money simply does not arrive.
Retailers have a name and a dashboard for every other kind of loss: shrink, markdowns, returns, failed payments. Deferred revenue is the one line most subscription operators cannot produce on request.
Most subscription platforms will show it to anyone who looks. The forecast view that projects the next 30 days of billings usually separates expected revenue from cancellations, failed payments and deferrals, as three distinct lines. Very few operators open it, and fewer still put it in front of a finance team.
The Revenue That Never Shows Up as Churn
Here is what it costs in practice. One account billing over $10 million a month in subscriptions is forecast to miss a large share of next month’s book. Cancellations account for 16% of that miss. Failed payments account for another 15%. Deferrals are the remaining 69%, roughly $2.5 million a month. A meal-delivery client projected $307,000 in billings over 30 days and collected about $200,000. Cancellations explained $25,000 of the $107,000 gap. Almost all of the rest was people pressing skip.
Neither brand had a retention problem it could see. Both had one they were paying for.
Deferrals Distort More Than Revenue
A cancellation at least tells the business something. It takes a customer out of the forecast, releases the inventory allocated to them and updates the demand plan. A deferral does none of that. The customer stays in the count, the units stay committed, and the cash does not arrive on the date the plan assumed. For a subscription retailer buying stock against a billing schedule, that is a working-capital problem as much as a marketing one. You order for a cohort that is still nominally active, hold product against orders that were quietly pushed a month out, and then find the gap in the bank rather than in the dashboard. Revenue that arrives late, or never, behaves a great deal like a customer who pays late. It is just as expensive and far harder to chase, because nothing in the system flags it as overdue.
Skips Cluster at the Second and Third Order
The timing makes it worse. In the accounts we reviewed, 14% of subscribers skip their second order and 18% skip their third.
That is the exact window where a subscription is supposed to turn into a habit. The first delivery proves the product works. The second and third prove the schedule works. Instead, roughly a sixth of every cohort quietly opts out of the part that decides whether they are still a customer in month six.
A Skip Is Rarely a Pause
The uncomfortable part is what happens next. Very few people who defer come back.
A skip is usually the polite version of leaving. No confrontation, no cancellation flow, no reason given. The customer gets to avoid the decision, and you get to avoid the data. Your subscription platform reads it as retention. Your bank statement reads it correctly a quarter later.
This is also why reported churn and felt churn never match. Our published subscription benchmarks put median annual churn, counting pauses, at 55.8%. Operators who quote me 20% are almost always quoting cancellations only, which is why their forecast and their deposits keep disagreeing.
The behaviour does respond to being taken seriously. On one account, rebuilding the cancellation flow around the reason the customer actually gave, instead of one blanket discount for everybody, moved the save rate from 4.17% to 42.34%. The product did not change and neither did the price. The same logic applies to skips, which almost nobody builds a flow for at all.
What This Means for Canadian Retailers
Canadian operators are already watching shoppers trade down, buy smaller baskets and shop more often. That behaviour does not stop at the subscription box. A household under cost pressure rarely cancels a subscription it still believes in. It skips one month, then another, and by the third skip it has replaced the habit with something cheaper.
If you run a replenishment, meal, wellness or pet subscription, four questions are worth putting to your team this week:
- How much revenue is deferred over the next 30 days, and how does that compare with what we expect to lose to cancellations?
- What share of subscribers skip their second order?
- Of the customers who skipped once last quarter, how many have billed since?
- How many taps does it take a paused customer to restart?
If nobody can answer the first question, that is the finding.
Four Changes That Close the Gap
- Measure it. Put deferred revenue next to cancelled revenue for the next 30 days, in the same report, every week. A number nobody owns is a number nobody fixes.
- Treat a skip like a cancellation. Ask the reason before offering anything, then answer the reason they actually gave. “Too much product” is an interval problem, not a price problem. Discounting someone who did not leave over price just teaches a happy customer to cancel for a coupon.
- Make resuming one tap. If coming back means a login, a password reset and a re-checkout, the skip was permanent and you have not noticed yet.
- Track return rate by skip number. A first skip and a third skip are different customers, and only one of them is still winnable.
The Number You Are Not Looking At
Across the accounts we analyzed, soft churn was the largest single leak and the least instrumented one. This is not a problem retailers are choosing to ignore. It is one their tooling does not show them, because the industry built its dashboards around the customers who announce that they are leaving.
Your churn number only describes the people who bothered to tell you.
