Resources

Payment Failure Rate: What's Normal, What's Not

Your payment failure rate is the share of subscription payments that do not go through. It sounds like a billing-platform statistic, but operators feel it as something else: subscribers who wanted to stay, disappearing because a card expired, funds were short on the wrong day, or a bank declined a charge it would have approved a day later.

This guide shows how to calculate your own rate correctly, read it by segment instead of as one blended number, and what actually moves it. One warning up front: there is no single universal "normal" percentage, and any benchmark quoted without a source, a population, and a definition should be ignored. Our failed payment recovery benchmarks page collects the published figures we do rely on, with sources attached.

How to calculate your payment failure rate

Payment failure rate = payments that failed on their first attempt ÷ total payments attempted in the period × 100.

Every word of that formula needs a decision behind it:

• First attempt only. Count a payment as failed when its initial charge fails. If you count every retry as a separate failure, a single troubled subscription can appear as four or five failures and your rate stops describing how many payments ran into trouble.
• Total payments attempted means every initial charge you tried to collect in the period — renewals and first charges together, unless you state otherwise. Keep the denominator consistent month to month or the trend is meaningless.
• The period must be named. A failure rate is always "for March" or "for the last 30 days." A floating percentage with no period cannot be compared to anything, including your own past.

Worked example

Example — hypothetical numbers, for illustration only. In April, a subscription business attempts 10,000 initial payments. Of those, 840 fail on the first attempt. The rest succeed immediately.

Payment failure rate = 840 ÷ 10,000 × 100 = 8.4% for April.

Note what that number does not say. It does not say 840 subscribers churned. Many of those payments will be collected on a retry or after the subscriber updates their card. The failure rate measures how often the first charge breaks. How many breakages become permanent losses is a separate question — the recovery question — and mixing the two is the most common diagnosis error in this area.

Count subscribers as a second view

The formula above counts payments. Run a companion count in subscribers: distinct subscribers with at least one failed payment, divided by subscribers billed in the period. The two views usually move together. When they diverge — the payment rate rises while the subscriber rate stays flat — a small group of accounts is failing repeatedly, pointing at account-level problems rather than a platform-wide one.

Failure is not churn: the recovery layer sits between them

A failed payment starts a clock, not a cancellation. Between failure and churn sit retries and card-update requests, and a large share of failed payments are collected there. According to Churnkey's published analysis of Stripe data, Stripe Smart Retries alone recover about 51% of failed payments (Churnkey analysis of Stripe data). In other words, roughly half of a first-attempt failure number, on that data, never becomes a loss at all — before any additional dunning work is layered on.

That gap is why you should never react to the failure rate alone. A rising failure rate with a strong recovery rate is a warning light. A stable failure rate with a weak recovery rate is the actual fire. Report the two together, always, and estimate the money at stake with a consistent method — our recovery calculator is built for exactly that arithmetic.

Segment 1: by decline type

The blended rate tells you something is happening. The decline mix tells you what. Group first-attempt failures by the decline category your processor returns, and calculate each category's share of all failures.

The largest category, in large-scale data, is not close. Across 5.4 million failed payments, insufficient funds was the largest decline category at 40.6% (Churnkey analysis of Stripe data). If your own mix looks anything like that shape, a large part of your failure rate is a timing and funds problem, not a card-details problem — and the remedies differ. Funds-related soft declines respond to when you retry. Expired and replaced cards do not respond to retries at all; they respond to a clear, early request for new card details. Hard declines, such as a card reported lost or stolen, should never be retried.

Watch the mix for movement, not just the level. A sudden jump in one category — authentication failures after a bank changes its rules, or declines concentrated in one card network — is diagnosable in a way a one-point rise in the blended rate never is.

Segment 2: new payments vs renewals

Always split first charges from renewal charges. They fail for different reasons and implicate different teams.

• First-charge failures happen when a trial converts or a new subscriber first pays. A high rate here points upstream: card capture problems, weak validation at signup, pricing-page surprises, or low-intent signups whose payment method was never solid.
• Renewal failures happen to subscribers who have paid successfully before — often many times. A high rate here points at card expiry, changed details, and funds timing across a large installed base. These subscribers demonstrably wanted the product, which is why renewal failures are usually the most recoverable and the most expensive to ignore.

A business can have a healthy blended rate while one of these segments quietly deteriorates, because the other segment is larger and dilutes it. Calculate both, every period, on the same definitions.

Segment 3: by price point and plan

Failure behaviour is not evenly spread across your price list. Calculate the rate separately for each plan and price tier, and for monthly versus annual billing.

Small, frequent charges fail in a particular pattern: the amount is minor enough that subscribers do not monitor the card behind it, so expired and replaced cards linger until a charge hits them. Large annual charges fail differently: a single big amount is more likely to meet a funds limit or trigger a cautious bank response, and each failure carries a year of revenue. Neither observation gives you a target percentage — it tells you where to look. If one tier's rate climbs while the others hold steady, the cause is likely specific to that tier, not your payment stack as a whole.

So what is "normal" — and what is not?

There is no defensible universal normal, and this page will not invent one. Your rate is shaped by price points, card mix, geography, how much of your base is on its first charge versus its twentieth, and how your processor categorises declines. A number alarming for an established B2B tool can be unremarkable for a low-priced consumer subscription, and vice versa.

What you can judge, reliably, is your rate against itself:

• A sudden spike is not normal. A sharp move in days usually means something changed — a processor issue, a price rise hitting renewals, a bank-side rule change, or a broken card-update flow. Investigate the segment and decline mix before touching anything else.
• A slow climb is not normal either. A gradual rise over quarters often tracks an ageing subscriber base whose cards are expiring faster than they are being refreshed. That is a pre-dunning and card-refresh problem, and it compounds quietly.
• A stable rate with falling recovery is a hidden deterioration. The failures did not change; your ability to fix them did. That shows up in churn a month later.
• One segment far above the rest is a lead, not a footnote. A first-charge, renewal, or single-tier rate well above your other segments deserves its own investigation.

What moves the payment failure rate

• Card lifecycle. Cards expire, get replaced, and get reissued after fraud scares. Every month a slice of your base pays with details that no longer work. Account updater services and early expiry reminders reduce this slice; ignoring it guarantees it grows.
• Funds timing. When the largest decline category is insufficient funds, the day and time a charge lands matters. Retry timing that respects pay cycles recovers payments that a retry an hour later would miss again.
• Retry behaviour. Retrying everything the same way — including hard declines that can never succeed — wastes attempts, can annoy card networks, and does nothing for the rate that matters: failures that become losses. Decline-aware logic treats each category differently.
• Price and plan changes. A price increase shows up first as a renewal failure bump. Model that before the rise, staff for it after, and do not mistake it for a random fluctuation.
• Signup quality and card capture. Typos, missing authentication steps, and payment methods captured carelessly at signup become first-charge failures at conversion. Fixing capture is cheaper than recovering the failure.
• Processor and bank-side changes. Issuer rules and network policies shift without asking you. Segment and decline-mix tracking is how you notice an external change in your own data quickly.

For the practical playbook that follows from this diagnosis — capture, pre-dunning, retries, and messages in order — see our guide on how to reduce failed payments.

The bottom line

Calculate your payment failure rate on first attempts, name the period, and never quote it alone: pair it with your recovery rate, split it by decline type, by new versus renewal, and by price point. Judge it against your own baseline and watch for spikes, slow climbs, and segments breaking away from the rest. The blended number tells you to look. The segments tell you where.

Frequently asked questions

How do you calculate the payment failure rate?

Divide payments that failed on their first attempt by total payments attempted in the same period, then multiply by 100. Count first attempts only, so retries do not inflate the figure, and always name the period the rate covers.

What is a normal payment failure rate for subscriptions?

There is no single normal rate that fits all subscription businesses. Price point, customer type, card mix, geography, and the balance of first charges to renewals all move the figure. Judge your rate against your own history, watch for sudden spikes or slow climbs, and compare segments within your business before using any external figure for orientation.

What is the most common reason payments fail?

In Churnkey's published analysis of 5.4 million failed payments from Stripe data, insufficient funds was the largest single decline category at 40.6 percent. Funds and timing problems make up a larger share of failures than most operators expect, which is why retry timing matters so much.

Does a failed payment mean the subscriber has churned?

No. A failed payment is the start of a recovery window, not a cancellation. According to Churnkey's analysis of Stripe data, Stripe Smart Retries alone recover about 51 percent of failed payments. Only payments still uncollected at the end of the recovery process become involuntary churn.

Should retries count in the payment failure rate?

No. Count each payment once, based on its first attempt. Counting every retry as a new failure lets one troubled subscription appear several times and makes the rate impossible to compare across periods, especially if your retry schedule changes.