Failed Payments Are a Retry Schedule Problem, Not a Churn Problem
Friday afternoon, a renewal fails. The customer is still using the product on Monday, still getting value from it, and has no idea anything went wrong. By the time anyone notices, the subscription has quietly cancelled itself and the customer is gone, not because they decided to leave but because nobody ran a retry schedule. That is the most common way subscription stores lose money they already earned.
The first thing to get right is the split between a soft decline and a hard decline, because the two need opposite responses. A soft decline is a timing problem: insufficient funds, a network error, a bank flag that clears. The card is fine, the moment is wrong, and another attempt on a different day often goes through. A hard decline is a facts problem: the card is lost, stolen, cancelled, the number is wrong, or the issuer demands authentication the customer has not completed. No retry schedule in the world collects from a stolen card.
What Smart Retries actually does
Stripe Smart Retries is the default on new accounts for a reason, and the default is specific: up to 8 retries inside a 2 week window, with the timing picked by the system instead of a fixed calendar. It watches signals like device activity and the best time of day to attempt the charge, and it reads the decline code before deciding anything. The window can stretch to 1 week, 2 weeks, 3 weeks, 1 month, or 2 months, and a custom schedule caps at three retries, each set a number of days after the previous attempt.
What it will not do matters just as much. Stripe does not retry when there is no payment method on file, when the card is India-issued, or when the issuer returns a hard decline code: incorrect number, lost card, stolen card, pickup card, revocation of authorization, authentication required, highest risk level, transaction not allowed. For those, the schedule keeps ticking but the charge only runs once the customer adds a new payment method. That is the design, not a bug. Eight attempts against a stolen card do nothing except burn goodwill and issuer fees.
Stripe reckons failed payments drive 20 to 40 percent of total churn for subscription businesses. That range is wide because the mix matters: a store with mostly soft declines recovers a lot, a store full of expired cards recovers almost nothing from retries alone. The number to watch on your own account is not the recovery rate, it is the split. If most failures are soft declines, fix the schedule. If most are hard declines, fix the card-update flow.
The emails that do the work
Retries without dunning emails leave money on the table, because a customer whose card expired needs to hear about it from you, not from a silent retry loop. The sequence that works is boring and specific: a pre-dunning note about 30 days before the card expires, then a first failed-payment email within 24 hours, then follow-ups on roughly day 3, day 7, and day 14. Stripe sends a failed payment email after each failed subscription payment when the setting is on, and sends an expiring card email one month before the card on file expires.
Each email needs one job and one link. The first says the payment failed, names the card, and links straight to the update page. The second says the retry is coming and offers help. The third says access changes on a date. The fourth says it changed. Stores that write four versions of the same vague nudge get ignored four times. The invoice.payment_failed webhook carries an attempt_count and a next_payment_attempt date, so your support team can answer "when will you try again" with a real date instead of a guess.
What to check on your own account tomorrow
Open Billing, then Revenue recovery, then Retries, and confirm Smart Retries is on with a window that matches how your customers pay. Then pull the last 30 days of failed invoices and sort by decline code. If soft declines dominate and your window is one week, stretch it. If hard declines dominate, stop polishing the retry schedule and put the card-update link in every email, in the account page, and in the first failed-payment notice. The retry schedule is the easy half. Getting a new card from a customer who has not noticed anything is the work, and it starts with telling them quickly, plainly, and more than once.
Keep reading: Shopify New Customer Accounts: What Breaks When Classic Accounts Go Away, Free Returns Are Not Free: The Cost Per Order Nobody Budgets, and Shopify Checkout Extensibility: The Deadlines Passed. Now What? cover the nearby cases in this series.
Frequently asked questions
How many times will Stripe retry a failed subscription payment?
By default, Smart Retries attempts up to 8 retries within a 2 week window, with timing chosen by the system. You can stretch the window to 1 week, 2 weeks, 3 weeks, 1 month, or 2 months, or switch Smart Retries off and set a custom schedule of up to three retries, each a set number of days after the previous attempt.
What is the difference between a soft decline and a hard decline?
A soft decline is temporary, like insufficient funds or a network error, and a later retry can succeed. A hard decline means the card is lost, stolen, cancelled, or needs authentication, and Stripe will not collect until the customer provides a new payment method, no matter how many times it retries.
When should I email a customer about a failed payment?
Send a pre-dunning note about 30 days before a card expires, then a failed-payment email within 24 hours of the failure, with follow-ups around day 3, day 7, and day 14. Each email should name the card and link directly to the page where the customer updates it.
Why did my subscription cancel if the customer never chose to leave?
When every retry in the schedule fails and the invoice stays open past the window, Stripe applies the ending you set, which is often cancellation. The customer usually never saw a notice. A working retry schedule plus dunning emails is what keeps a payment failure from turning into silent churn.