What are Failed Charges?
Failed Charges is the sum of amounts from failed or open invoices in a period. It represents revenue at risk from payment failures — money you billed but Stripe couldn’t collect.
Payment failures happen for many reasons: expired cards, insufficient funds, bank declines, fraud blocks, and network errors. Most are recoverable with retry logic and card update prompts. The ones that aren’t become involuntary churn — customers who didn’t choose to leave but lost access because their payment failed.
Failed Charges is a leading indicator — a healthy level is below 1–2% of MRR. Rising failure amounts signal a growing gap between billed and collected revenue. Left unaddressed, that gap converts into customer churn and depresses Net Cash Flow.
The Failed Charges formula
| Variable | What it captures |
|---|---|
| Failed Charge Amount | Sum of amounts from invoices with failed or open payment status in the period, excluding invoices that were later paid |
Failed Charge Rate
The rate version normalizes failed charges as a percentage of total billings, making it comparable across company sizes:
Spot revenue at risk before it churns
Connect Stripe and see your failed charges, recovery rate, and at-risk MRR — with automated alerts when failures spike.
Track My Failed ChargesWorked example
March 2026: Your company processes 200 invoices during the month.
| Category | Invoices | Amount |
|---|---|---|
| Paid invoices | 185 | $52,300 |
| Failed → later recovered | 8 | (excluded) |
| Failed → still unpaid | 7 | $2,180 |
| Failed Charges | $2,180 |
At a 4.0% failure rate, $2,180 in revenue is at risk this month. The 8 invoices that failed but were later recovered don’t count — they eventually produced cash. If those 7 remaining invoices aren’t recovered through dunning or card updates, they become involuntary churn.
How it’s computed
North Metric computes Failed Charges using a deduplication pipeline that separates genuine payment failures from retry noise:
The deduplication pipeline
| Variable | What it captures |
|---|---|
| Step 1 | Pull all charges from the period via Stripe's charges API |
| Step 2 | Pull all paid invoices from the same period |
| Step 3 | Map charges to their parent invoices — multiple charge attempts may reference the same invoice |
| Step 4 | Exclude any invoice that appears in the paid invoices set — if it was eventually paid, it's not a failure |
| Step 5 | Sum the remaining unique invoice amounts (÷ 100 from Stripe's cent denomination) → Failed Charge Amount |
| Step 6 | Count the remaining unique invoices → Failed Charge Count |
This approach ensures that a single invoice that generated four retry attempts counts once, and invoices recovered by Stripe’s built-in retry or by dunning emails are excluded entirely.
Common failure reasons
Stripe categorizes failures by decline code. The most common in SaaS: card_declined (insufficient funds or generic bank decline), expired_card (card past its expiry date), authentication_required (3D Secure not completed), and processing_error(temporary network issues). Each has a different recovery strategy — card updates for expiry, retry for network errors, and customer outreach for bank declines.
From failure to churn
Stripe retries failed charges automatically (typically 3–4 times over 1–3 weeks). If all retries fail, the subscription transitions to past_due and eventually canceled. At that point, the failed charge becomes involuntary customer churn. Active dunning — email prompts to update payment methods — can recover 20–40% of failures before they reach cancellation.
Failed Charges vs Churn
Failed Charges and churn are sequential — failures today become churn tomorrow if unrecovered.
| Failed Charges | Customer Churn | |
|---|---|---|
| What it measures | Revenue that was billed but not collected | Customers lost — subscription fully canceled |
| Timing | Leading — appears first when payment methods expire or funds run out | Lagging — appears after all recovery attempts have been exhausted |
| Reversible? | Yes — recoverable through retries, card updates, and dunning | Partially — reactivation possible but requires re-acquisition effort |
| Intervention | Payment recovery: automated retries, update-card emails, smart retry timing | Retention: value demonstration, win-back campaigns, exit interviews |
| Direction | Lower is better | Lower is better |
Failed Charges are the leading signal. They appear days or weeks before the customer is marked as churned.
Churn is the trailing outcome. By the time a customer shows in your churn numbers, the payment failure happened weeks ago. Monitoring failed charges catches the problem at the recoverable stage.
Common Failed Charges mistakes
- Counting retry attempts as separate failures. A single invoice that fails three times is one failure, not three. Without deduplication, your failure count and amount are inflated 2–4×. North Metric deduplicates by invoice to avoid this.
- Not excluding recovered invoices. An invoice that failed on attempt 1 but succeeded on attempt 3 is not a failure — it’s a successful collection with retry delay. Counting it inflates the failure rate and hides the actual at-risk amount.
- Treating all failures equally. An expired card is easily recovered with a card-update email. A fraud decline may require a new payment method entirely. A $29 failure and a $2,900 failure have very different urgency. Segment by decline code and amount to prioritize recovery efforts.
- Ignoring seasonality. Card expiry clusters around December–January (holiday reissues) and mid-year (corporate card renewals). A spike in Q1 failures may be calendar-driven, not a product or billing problem.
- Not connecting failures to churn. Every unrecovered failed charge eventually becomes involuntary churn. If your churn rate is rising and you haven’t checked failed charges, you may be solving the wrong problem — retention campaigns won’t help customers who simply have expired cards.
Failed Charge Rate benchmarks
Failed Charge Rate benchmarks are segmented by MRR tier. Lower is better — a rate below 3% means your payment infrastructure is healthy. Early-stage companies tend to have higher failure rates due to smaller customer bases where a few card expirations disproportionately affect the percentage.
| MRR Tier | Range | Top 25% | Median | Bottom 25% |
|---|---|---|---|---|
| Seed | < $10K | 5.0% | 8.0% | 12.0% |
| Early | $10K – $50K | 4.0% | 7.0% | 10.0% |
| Growth | $50K – $100K | 3.0% | 5.0% | 8.0% |
| Scale | $100K – $500K | 2.0% | 4.0% | 7.0% |
| Enterprise | $500K+ | 2.0% | 3.0% | 5.0% |
| Failed Charge Rate benchmarks from 1,400+ Stripe-verified SaaS companies. | ||||
How does your failure rate compare?
Benchmark your Failed Charge Rate against 1,400+ SaaS companies at your MRR stage.
Benchmark My SaaSFrequently asked questions
What is a good Failed Charge Rate for SaaS?
Below 3% is healthy for most SaaS companies. The median across stages ranges from roughly 3% to 8% depending on company size — larger companies with enterprise customers tend to have lower failure rates. Above 10% signals a systemic payment infrastructure problem that needs immediate attention.
What causes payment failures?
The most common causes are expired cards, insufficient funds, bank declines, and incomplete 3D Secure authentication. Expired cards are the easiest to recover — a simple email prompting the customer to update their card resolves most cases. Insufficient funds and generic bank declines require smart retry timing (retry when the customer is likely to have funds, such as after payday).
How does North Metric avoid double-counting retries?
North Metric deduplicates by invoice, not by charge attempt. Multiple failed charge attempts against the same invoice count as one failure. If the invoice is eventually paid (through any retry or manual payment), it’s excluded from the failure count entirely. This gives you the true number of unresolved failures.
Do failed charges affect MRR?
Not directly. MRR measures the subscription amount, not whether it was successfully collected. A $100/month subscription with a failed charge still shows $100 in MRR until the subscription is canceled. Failed charges affect Net Cash Flow (less cash collected) and eventually churn if unrecovered.
What percentage of failed charges can be recovered?
With active dunning (automated retry + card-update emails), SaaS companies typically recover 20–40% of failed charges. Smart retry timing — retrying based on the failure code and historical success patterns rather than fixed intervals — can push recovery rates above 50%. The remaining failures usually convert to involuntary churn within 2–4 weeks.
Ready for metric clarity?
Connect Stripe and let AI build your metrics system. No manual setup, no guesswork.
Get Started