Picture one of your better subscribers. She has taken your coffee every month for half a year. She likes it, she tells people about it, and she has never once opened your cancel page. Then her bank spots a suspicious charge somewhere else, freezes her card, and mails her a new one with a new number.
Your next renewal goes to the old card and gets declined. Your subscription app tries again a few days later, then again, and at the end of its schedule it does what it was set to do on the day you installed it. It cancels her. She gets an email she may or may not read. Your dashboard records one more lost subscriber, filed in the same pile as the people who left because they had too much coffee in the cupboard.
Nobody decided anything. She did not quit. You did not let her go on purpose. A setting did.
That is involuntary churn: a subscriber who wanted to keep paying, lost because a payment failed and nothing recovered it. It is one of the largest and least examined leaks in a subscription store, and most of it is fixable with settings you already have and three emails you can write in an afternoon.
A Third of Your Lost Subscribers Never Decided to Leave
Recurly, a subscription billing platform, publishes churn benchmarks from its network by industry. In its figures for ecommerce subscriptions, updated with July 2026 data, total churn is 4.25% and the involuntary part of that is 1.38 points. Do the division and roughly one in three subscribers an ecommerce store loses was lost to a payment that failed, not to a decision.
That third is easy to miss because of where it lands. Most stores look at one churn number. A subscriber who canceled because they were bored and a subscriber whose card expired both reduce it by one, and from the outside the two look identical. So the owner reads the number, concludes that people are tiring of the product, and starts working on the product, the offer or the cancel flow. All worthwhile work, aimed at the wrong third of the problem.
The cost is also bigger than it looks, because of what kind of sale this is. When a one-time order fails at checkout, you lose that order. When a renewal fails and the subscription ends, you lose that order and every one that would have come after it. We have written about why a subscription is a different business from selling the same product twice, and this is one of the places it shows: the thing that breaks is not a sale, it is a stream of them.
And remember what that stream is paying for. You spent money on ads to win this customer, often more than the first order earned, on the understanding that the later orders would pay it back. As we put it in Lifetime Value Does Not Pay This Month's Invoice, the profit lives in the months you have not reached yet. A subscriber lost to a declined card in month six takes with them the months that were supposed to make the acquisition worth it, and you paid full price to acquire someone who never stopped wanting the product.
Not Every Declined Card Deserves Another Try
Before you change any setting, it helps to know that "declined" covers very different situations, and they call for opposite responses.
Some declines are final. The card was reported lost or stolen, the number is wrong, or the account is closed. Stripe, the payment processor, publishes a list of decline codes it will never retry automatically, including a lost card, a stolen card, a card the bank has asked to be taken back, and an incorrect card number. Retrying these accomplishes nothing except more declines. The only path back is the customer giving you a different card.
Most declines are not like that. The card was replaced after a fraud scare, it simply expired, the account was short on funds the day you charged it, or the bank returned a vague "do not honor" because something about the timing looked unusual. These are customers who can pay and want to. Some of them recover on their own with a well-timed second attempt. Others only recover when the card details are updated, either automatically (more on that below) or by the customer.
Here is the part that surprises most owners: more retries are not better. It feels as if a card that failed once should simply be tried until it works. Vijay Menon, who worked on subscriber retention at Xbox and Dropbox before founding a payment-recovery company, describes what actually happens when a business does that. Its share of approved charges drops, the banks start reading the pattern as the behavior of a risky merchant, and they tighten up on all of its charges, including the healthy ones. Maria Sparagis, who has spent over twenty years in payment processing, adds the other cost: money finally collected on the sixth or seventh try is the money most likely to come back as a refund request or a chargeback, because it came out of an account that plainly had none to spare.
So the goal is a few attempts, spaced sensibly, aimed at the declines that can actually recover. Insufficient-funds declines in particular do better a few days later or near the start of the month, when many people have just been paid, than they do the next morning.
Your Subscription App Already Retries Failed Cards. Nobody Checked the Settings.
Every subscription app retries failed payments on some schedule. The question is whether that schedule is one you chose. On most stores it is whatever was in place the day the app was installed, and nobody has looked at it since.
Here is what the most common options do, according to their own documentation:
- Shopify Subscriptions. Retries on a schedule you set, then takes a final action you choose. You control the number of retries, the days between them, and whether to skip, pause or cancel at the end.
- Recharge. Picks retry timing automatically from the decline type and past results, does not retry hard declines, and keeps a default recovery window of 120 days. When retries run out, it cancels; its support team can change the timing.
- Skio. Retries on your schedule, with optional Smart Retries that choose the time of day and day of month. Skio suggests starting at 3 days between tries and 7 tries in total.
- Loop. Retries on your schedule. You set the number of retries, the spacing, and whether to skip, pause or cancel at the end, and Loop recommends pause.
- Stripe Billing. Smart Retries choose the best times within a window you pick. Stripe's recommended default is 8 tries within 2 weeks.
Shopify does not publish the default numbers for its own app, which is reason enough to open the settings and read them. In the Shopify Subscriptions app, they sit under the app settings in a section for payment method failure. Write down what is there before you change anything, so you know what you are comparing against.
Three settings matter more than the rest.
The final action. This is the decision that turns a failed card into a lost customer, and it is the one most worth changing. Canceling ends the relationship. Pausing keeps the subscription in place, stops the shipments, and lets it pick up again the moment a working card arrives. Loop's own help center calls pause the best default for most merchants. Recharge goes part of the way on its own: when a customer updates their payment method, it reactivates the subscriptions it canceled for running out of retries. Either way, the principle is the same. A paused subscriber is one click from coming back. A canceled one has to decide to sign up again, and many never will.
The number of tries. A handful, spread over a couple of weeks, is where the platforms themselves land. Skio's suggested seven tries three days apart and Stripe's eight in two weeks are both well short of the fifteen-attempt approach that hurts approval rates. If your app lets you pick, stay in that range.
The spacing. Back-to-back daily retries mostly hit the same empty account or the same suspicious bank on consecutive days. Several days apart, ideally with one attempt landing near the start of a month, gives an insufficient-funds decline a real chance.
Visa and Mastercard Will Send You a Subscriber's New Card Number Automatically
The single most common reason a healthy subscriber's renewal fails is also the most boring: the card on file is out of date. It expired, or the bank replaced it after a fraud alert or a lost wallet. The customer has a perfectly good card in their wallet. You just do not have its number.
The card networks built a fix for exactly this. Visa Account Updater and Mastercard Automatic Billing Updater let a merchant ask for the latest details on a stored card before charging it. When the customer's bank issues a replacement, it shares the new number and expiration date, and the next renewal goes through on the new card without anyone lifting a finger.
If your store takes payment through Shopify Payments, this is already working for you: Shopify says it works with the major card networks to update saved card details automatically when customers receive new cards, and that this is widely supported in the United States. Stripe does the same and says it can update most American Express, Visa, Mastercard and Discover cards issued in the US.
If you bill through a different gateway or processor, do not assume. Ask them, in writing, whether account updating is turned on for your account and whether it costs extra. Sparagis's advice to merchants is blunt: it is a short request, and it does not always happen unless you make it.
Two limits are worth knowing. First, it covers most US cards but not all of them, and international support varies. Second, as Stripe notes, there is no way to tell in advance which cards will update. So the updater shrinks the problem without ending it, and some customers will still need to give you a new card themselves. That is what the emails are for.
One Failed-Payment Email Is Not Enough. Each Stage Needs Its Own.
Most stores send one failed-payment email, and it usually says something like "Your payment failed. Please update your card." Read that from the customer's side. They did nothing wrong. Their bank sent them a new card, or they had a tight week. And the message from a brand they like implies they are behind on a bill. Menon makes this point directly: the standard notice puts the blame on the customer for a problem that is often on the merchant's side, and it erodes the trust a subscription depends on.
Better is three emails, each with its own job, sent at the moment it applies.
The heads-up, before the card expires. This is the one failure you can see coming. Every card has an expiration date printed on it, and cards usually last three to five years, so in any given month some of your subscribers are approaching theirs. Subscription apps generally know which ones; Loop, for example, sends a card-expiring event to Klaviyo for exactly this purpose. A short note a few weeks out, something like "Your card on file expires next month. Want to update it now so your next box ships on time?", fixes the problem before a single charge fails. No urgency, no warning tone. It is a favor.
The plain note, after a charge fails. Keep it calm and factual: the renewal did not go through, here is the date of the next attempt, here is a link to update the card. Two details matter more than the wording. The link should take the customer straight to the card form, already signed in, rather than to a login screen where some of them will give up. And the email should make sense on its own, because the customer probably did not read the last one. The team at Loop puts it well: treat the sequence less like a series and more like a set of states, where each message describes exactly where this subscriber is right now.
The welcome-back, once the subscription is paused. By now the retries are done, and functionally this is a win-back. It is also the only one of the three where an incentive belongs. A discount on day one teaches customers that a failed card earns a coupon, and it hands money to people who were going to fix their card anyway. Saved for the last step, it goes only to the subscribers who needed a reason.
If you already run the post-purchase and win-back flows we described in Your Store Sends One Automated Email. It Should Be Sending Five., these three sit alongside them and use the same tools. Most subscription apps send their own basic notices too, so check what is going out already before you add more, and switch off any that would double up.
Failed Payments Are Costing You a Number You Have Never Looked At
None of this is worth doing on faith. Here is how to see the size of the problem in your own store.
Count the subscriptions that ended because of a payment. Most subscription apps record why a subscription was canceled or paused. Recharge, for example, labels the ones it canceled after running out of retries with the reason "Failed Payment flow max retries." Pull the last 90 days and count them.
Compare them with all the subscriptions you lost. Divide the failed-payment losses by your total cancellations over the same period. If the share comes out near the third in Recurly's ecommerce figures, or higher, this is one of the biggest retention problems you have.
Look at how many failed payments you got back. Recurly's analysis of 76 million subscribers across 2,200 businesses found that about 53% of failed monthly payments were eventually recovered. That figure spans many kinds of subscription businesses, so do not treat it as a target. Treat it as a sense of what is possible. If you recover far less, your settings are the first suspect.
Then put a dollar figure on it. Here is an illustrative example, not a benchmark. Say you have 300 subscribers paying $40 a month. Recurly's 2026 subscription report, as cited by Shopify, puts the share of renewal charges declined at 9%, so call it 27 failed renewals a month. If your current setup recovers half of them, about 13 subscribers a month are lost to a card. If the average subscriber would otherwise have stayed another eight months, each one is worth $320 in future orders, and those 13 are about $4,200 a month in revenue that never had a chance. At a 60% gross margin, that is roughly $2,500 a month of profit, every month, lost by default.
Your numbers will differ, and that is the point of running them. The fixes in this piece cost little: a final action switched from cancel to pause, a retry schedule you chose, account updating confirmed, and three short emails. Every subscriber they save is one you already paid to acquire, and as we covered in Your First Sale Barely Breaks Even. The Second One Is the Profit., the repeat orders are where the money is.
The Subscribers You Counted as Churn Were Never Leaving
The subscribers in this piece are the easiest customers you will ever keep. They chose you, they kept choosing you, and they are still willing to pay. All they need is for your store to stop treating an expired card as a goodbye.
If you would rather have someone go through your subscription settings, your failed-payment emails and the numbers behind them with you, it is the kind of work we do for online stores every week. And if you would rather do it yourself, start by opening your app's failed-payment settings today and reading what is actually there.
