The Complete Guide to Payment Recovery for SaaS

Payment recovery (also called dunning management) is the practice of retrying failed credit-card charges, communicating with affected customers, and applying save offers — all to recover revenue that would otherwise become involuntary churn.

For a SaaS business, this matters more than most operators realize. 5-12% of recurring charges fail in any given month. Without active recovery, roughly two-thirds of those customers become permanently churned — not because they wanted to leave, but because their card expired, hit insufficient funds, or got flagged by their bank's fraud system.

The recovered revenue is pure margin. You've already done the acquisition work. You've built the product they're using. All you need is to fix the plumbing on the way out.

Voluntary churn: the customer actively cancels. They're telling you something — about your product, pricing, or their own changing needs. The right response is cancellation flows, retention surveys, and product/pricing iteration.

Involuntary churn: the customer stops paying without intending to. Their card expired. They moved and the AVS check failed. Their bank's fraud system flagged a charge. The customer may not even know it happened.

These need entirely different tactics. Read our deep dive on the difference for the full breakdown.

Every failed charge comes with a decline code. Most teams ignore them and retry blindly. That's a mistake.

Decline codes split into six categories: soft (insufficient funds, processing errors — retryable), hard (expired card, do_not_try_again — need new card), fraud (don't retry — risks chargebacks), input (wrong CVC/ZIP — re-prompt), issuer (call_issuer — bank wants confirmation), and limits (velocity exceeded — wait window then retry).

Treating them all the same wastes retries on hard declines (which will keep failing) and risks chargebacks on fraud codes. Treating them differently lifts recovery from ~30% to 50%+.

See the complete decline-code reference →

Stripe Smart Retries default to 4 attempts spread across ~7 days using machine-learned timing. That's a fine baseline.

Better cadence by category:

• Soft declines: retry day 1, day 3, day 7. Pause if still failing — don't spam.
• Insufficient funds: day 3, day 7 — funds typically reload mid-month.
• Expired cards: don't retry blindly. Email customer with update-card link first.
• Fraud / hard declines: don't retry at all. Email customer for new payment method.

The numbers: best retry schedule for failed payments covers the data behind these recommendations.

The single highest-ROI improvement to dunning is adding a personalized email between retries. Stripe doesn't email customers when payments fail — and most failed payments resolve faster when the customer knows about them.

A working sequence:

1. Day 0: friendly notification of failure with one-click update-card link.
2. Day 3: gentle nudge mentioning the specific decline reason.
3. Day 7: final notice + save offer (discount or pause).
4. Day 14+: cancel; win-back email at day 30.

Tone matters more than copy. Friendly recovers 2x better than urgent for first failure. Urgency is for attempts 2-3 once the deadline is real.

12 copy-paste email templates per scenario →

When retries and emails fail by attempt 3, save offers are the difference between recovery and churn. Three types work in different contexts:

• Discount (e.g. 20% off next 3 months) — best for B2B SaaS, courses, agencies. Frames it as a deal, not a beg.
• Pause (e.g. 60 days no charge, full access preserved) — best for membership sites, fitness apps, seasonal-use products.
• Downgrade (cheaper tier) — only if you have a genuine cheaper plan. Most SaaS pricing is too tight for downgrades to feel like a save.

Test which works best for your segment. See industry-specific save-offer guidance →

The case for any dunning tool is one of the easiest in SaaS.

• $100K MRR × 7% failure rate = $7K/mo of failed charges
• Stripe-only recovers ~30% = $2.1K/mo. You lose $4.9K/mo to involuntary churn.
• Adding emails + save offers lifts recovery to 50% = $3.5K/mo. Net gain: $1.4K/mo over Stripe-only.
• Rechurn costs $49/mo. ROI: 28x.

Run your own numbers in the free ROI calculator.

Build if dunning is your core competency or you have unusual requirements. Cost: 3-6 weeks of engineering plus ongoing maintenance.

Buy if you're a SaaS focused on your product. Even Stripe Smart Retries (free) is better than nothing. A dedicated tool that adds emails + save offers + per-decline-code logic typically pays for itself with 1-2 saved customers/month.

Major tools differ a lot in pricing and depth. See full comparisons across all major dunning tools →

1. Week 1: Run a free Stripe audit (read-only) to measure your current failure mix and recoverable MRR.
2. Week 2: Activate dunning automation — Stripe Smart Retries minimum, Rechurn or competitor for full recovery.
3. Week 3: Configure save offers. Discount for B2B, pause for B2C is a good starting heuristic.
4. Week 4: Review the recovery rate. Iterate tone + cadence per failure code.

Most teams see meaningful recovery within the first 30 days. The longer you wait, the more revenue silently drains.