Prompt
Why are so many subscription payments failing and what dunning setup should I use?
Latest observation
Oct 15, 2026
OpenAI APIWeb search: off
A lot of subscription payments fail for a few common reasons, and the “right” dunning setup is usually one that’s automatic, multi-step, and tuned to the reason for failure.
Why subscription payments fail so often
1) Card issues
- Expired card
- Insufficient funds
- Card reported lost/stolen
- Card limit reached
- Bank blocks the charge as suspicious
2) Payment method changes
- Customer gets a new card
- Bank reissues the card
- Account funding source changes
- Stored payment method is no longer valid
3) Authentication / bank declines
- 3D Secure or bank verification required
- Fraud rules trip
- AVS/CVV mismatch
- Issuer simply declines without a clear customer-facing reason
4) Technical / processing issues
- Network timeouts
- Temporary gateway or processor errors
- Retry happening at a bad time
- Currency/country mismatch or configuration issues
5) Subscription lifecycle problems
- Free trial ends and the payment method isn’t ready
- Renewal happens on a date the customer forgot about
- Annual renewals fail after a long gap because the card changed
What dunning setup you should use
A good default dunning system should do 4 things:
1) Retry automatically, but not too aggressively
Use a retry schedule like:
- Immediate retry: 0–15 minutes after failure
- Second retry: 1 day later
- Third retry: 3 days later
- Fourth retry: 5–7 days later
- Final retry: 10–14 days later
Why this works:
- Some failures are temporary.
- Banking networks often succeed on a later attempt.
- Spacing retries avoids looking like fraud and reduces wasted attempts.
If your processor provides decline codes, adjust:
- Retry more for soft declines
- Don’t retry much for hard declines like lost/stolen cards
2) Send customer notifications
Use a sequence like:
- Email 1: immediately after failure
- Email 2: 2–3 days later
- Email 3: before account suspension
- Final email: when service is about to pause/cancel
Each email should include:
- Clear failure reason if known
- Updated payment link
- Amount due
- Deadline
- What happens if payment isn’t updated
3) Let customers fix payment details easily
Make sure they can:
- Update card details in one click
- Change payment method without contacting support
- Pay a past-due invoice
- See retry date and status
This is often the biggest improvement to recovery rates.
4) Use logic based on decline type
Classify declines into:
- Soft declines: temporary issues, retry several times
- Hard declines: card closed, invalid number, lost/stolen, do not keep hammering
- Authentication issues: may need customer action
- Insufficient funds: retry, but time retries around payday if possible
Recommended practical dunning setup
If you want a simple, effective default:
Retries
- Retry 1: immediately
- Retry 2: after 1 day
- Retry 3: after 3 days
- Retry 4: after 7 days
- Retry 5: after 14 days
Notifications
- Email after each failed attempt, but don’t spam
- One SMS/push notification if you have high-value subscriptions
- Final warning 1–2 days before access is restricted
Account access
- For B2C/SaaS:
- Grace period of 3–7 days
- Then suspend non-critical features
- Consider full lockout only after multiple failed retries
- For low-cost subscriptions:
- Shorter grace period is often fine
- For high-value B2B:
- Longer grace period and maybe human follow-up
Final handling
- After final retry window, mark account past due
- Keep a self-serve reactivation flow
- Optionally attempt again if the customer updates card details
Best practices to reduce failures overall
- Enable network tokenization / account updater if available
- Support Apple Pay / Google Pay / ACH / SEPA, depending on market
- Collect cards using optimized checkout and SCA/3DS flows
- Prompt users to keep payment methods current
- Retry at local business hours or likely payday times when possible
- Monitor decline codes by issuer and region
- Segment by subscription value and customer type
A good rule of thumb
If you don’t have strong evidence yet, start with:
- 4–5 retries over 14 days
- Email notifications on every failure
- Self-serve payment update flow
- Different handling for hard vs soft declines
- A 3–7 day grace period before suspending service
If you want, I can also give you:
- a dunning email sequence,
- a retry schedule by decline code,
- or a recommended setup for Stripe / Paddle / Recurly / Chargebee.