Subscription retention strategies live mostly in the billing mechanics, not the marketing. This article answers one question: in a business that charges customers on a recurring basis, which parts of the subscription itself, from plan design to failed payments to the cancellation page, decide how many customers are still paying a year from now? For an established company of 10 to 250 people selling retainers, memberships, service plans or software, these mechanics are usually the cheapest retention gains available, because they need no new product and no extra staff.
Two kinds of churn, two different fixes
Subscribers leave in two ways. Voluntary churn is a decision: the customer chooses to cancel. Involuntary churn is an accident: the card expires, the bank declines a payment, the direct debit mandate lapses, and nobody notices until access is cut. Most retention advice talks only about the first kind. In many subscription businesses the second kind is a meaningful share of total losses, and it is far easier to fix, because those customers never meant to leave.
Start by splitting your last year of cancellations into those two buckets. If you cannot tell them apart, that is the first problem to solve: your billing records should say whether a subscription ended because of a cancellation request or a payment failure.
Suppose the business loses about 40 subscribers a month, and the split shows 26 cancellations and 14 failed payments never recovered. Say those 14 paid an illustrative £60 a month. Recovering half of them would keep 7 subscribers, or £420 a month, which is about £5,000 of revenue a year for each month's batch saved. That would compound as each saved subscriber keeps paying, with no new marketing spend.
Fixing involuntary churn first
A failed payment should start a sequence, not end a subscription. The sequence has three parts: retry the charge on a sensible schedule, tell the customer plainly what happened, and give them a one-step way to update their payment details.
On the levelpartner.ai platform, the first part is a setting, not a project. Automatic retries are configured once for every subscription, and a person can retry a past-due charge by hand when the automatic attempts run out.
Set automatic retries for failed payments
- Go to Payments, then Settings, then Subscription.
- Select the number of retries, up to three.
- Choose the gap between retries: one, three, five or seven days.
- Save the settings.
- Decide whether a subscription that fails its final retry moves to Cancelled automatically, and set the checkbox to match.
Leave that last checkbox off unless you have a reason. A subscription left as Unpaid can still be recovered by a person; one moved to Cancelled has to be sold again. The settings apply to new and ongoing subscriptions, including ones already part-way through retrying. Retries are available for some payment providers and not others, so check that yours is supported before relying on them.
Retry a past-due payment by hand
- Go to Payments, then Subscriptions.
- Find the failed payment and open its Actions menu.
- Click Retry Payment.
- Check the details in the Retry failed payment window.
- Select a saved card, or click Add New to enter a different one.
- Turn on Use this card for future payments if the customer has given you a new card.
- Click Charge Card.
Only the oldest unpaid payment is retried, and each card allows three unsuccessful attempts before a 24-hour pause. Payments that were missed during the retry window are not charged later on their own; send the customer an invoice or a one-time payment link for them.
Tell the customer, in a workflow
Retries recover the charge when the problem was temporary. When the card has expired, only the customer can fix it. Start a workflow from the Subscription trigger in Workflows so that a failed or past-due subscription sends an email and SMS with a link to update the card, a second message a few days later if nothing has changed, and a task for a person to call higher-value subscribers before access ends. The tone matters. "Your payment did not go through, here is a link to fix it" works. "Your account is suspended" does not, because it reads as a punishment for something the customer did not do.
Imagine a 35-person firm selling a monthly service plan at an illustrative £150, with 12 payments failing in a month. If three retries three days apart cleared 5, card-update messages recovered 4 and a call with a manual retry recovered 2, the firm would lose one subscription instead of twelve. Each layer catches a different cause: a temporary decline, an expired card, a customer who needs a nudge from a person.
Cutting access the moment a payment fails. A customer who loses access without warning often takes it as a reason to reconsider whether they needed the service at all. Give a short grace period, keep access live while you chase, and end it only when the sequence has run.
Plan design: the decisions that shape churn before anyone joins
How you structure plans decides how often customers face a choice to leave. A monthly plan asks the question twelve times a year. An annual plan asks it once. That is why annual plans tend to retain better, though part of the effect is simply that the customers who choose annual were already more committed.
The trade-off is conversion. Pushing everyone to annual terms raises the barrier to joining and some buyers will walk away. A common middle path is to offer both, price annual with a clear saving, and invite monthly subscribers to switch once they have been active for a few months and are clearly getting value. That timing matters: an offer to commit for a year lands better after the customer has seen results than before.
Other plan levers worth deciding deliberately:
- A pause option, so customers with a temporary reason to stop can freeze rather than cancel.
- A lower tier, so a customer whose budget has shrunk can step down rather than out.
- Clear minimum terms stated at sign-up, so nobody feels trapped and complains about it publicly.
The cancellation flow
Every subscription business should know exactly what happens when a customer asks to cancel. In many companies, it depends on who answers the email. A defined flow makes the result consistent and gives you data.
A good cancellation flow asks one question, the reason, from a short list plus an "other" box. It then offers one alternative matched to the reason: a pause for "taking a break", a lower tier for "too expensive", a call with someone for "not getting results". If the customer still wants to leave, it confirms the cancellation promptly and cleanly. Making cancellation difficult is both bad practice and, in the US, the UK and other markets, increasingly a regulatory concern for consumer subscriptions, so the flow must never obstruct.
Forms on the platform can capture the reason and route the response, and the reasons feed reporting so you can see month by month why people leave. What to do with those reasons across the wider client base is covered in client retention strategies.
Say a quarter brings 90 cancellation reasons: 35 "too expensive", 25 "not using it enough", 20 "switching provider" and 10 other. If a third of the price group took the lower tier and a third of the low-use group took a pause, that would be about 20 subscriptions kept from 90 requests, without discounting the main plan. The 20 switching provider would go to the product team for review.
What most articles on subscription retention get wrong
The usual advice is to offer a discount to anyone who tries to cancel. That teaches your subscribers that the way to get a lower price is to threaten to leave, and the word spreads. Within a year you have a two-tier base: people paying full price and people who asked. A pause or a step down protects revenue without devaluing the plan, because the customer accepts less in exchange for paying less.
The second error is measuring retention only as a single churn rate. A blended rate hides the fact that churn is usually concentrated in the first few billing cycles. Look at it by cohort: of everyone who joined in a given month, how many are still paying after one, three, six and twelve months. The shape of that curve tells you whether the problem is onboarding, value or price. Our article on the measures worth tracking for retention explains how to set up cohort views.
The first months decide most of it
If the cohort curve shows heavy losses early, the problem is not the cancellation flow; it is that new subscribers are not reaching value quickly. The fix there is a structured first month: a welcome sequence that points to the one action that matters most, a check-in when the first billing date approaches, and a nudge if the customer has not done that action yet. For membership and course products, the platform's memberships area can trigger messages based on progress, so a subscriber who has not started the first module hears from you before they forget why they joined.
Where to start
- Split last year's losses into voluntary and involuntary.
- Set automatic retries, then build the card-update sequence, with a grace period and a human call for high-value subscribers.
- Add a one-question cancellation form with a matched alternative for each reason.
- Introduce a pause and a lower tier if you do not have them.
- Set up a cohort view and review it monthly.
Each step is small. Together they usually do more for a subscription business than a new acquisition campaign, and they keep working without anyone remembering to run them. For campaign ideas to win back subscribers who have already gone, see retention campaign ideas worth testing.
Frequently asked questions
What is the difference between churn and retention?
They are two views of the same thing. Retention is the share of subscribers still paying after a period; churn is the share who left. Track retention by cohort for decisions and churn as a monthly headline.
Should we offer a discount to stop cancellations?
Usually not as the first option. A pause or a cheaper tier keeps the customer without teaching your base that threatening to leave earns a discount. Reserve discounts for specific, high-value cases a person decides on.
Should failed subscriptions be cancelled automatically?
Usually not. Leaving them as Unpaid after the last retry keeps the door open for a manual retry once the customer updates their card. If your plans for retainer clients sit in a renewals pipeline, CRM retention strategies explains how the two fit together.
How long should the grace period be after a failed payment?
Long enough for the reminder sequence to run, typically one to two weeks. Keep access live during that time and make updating the card a single step.
Do annual plans always beat monthly plans?
They usually retain better, but they also convert fewer first-time buyers. Offer both, and invite monthly subscribers to switch once they have seen results.