Dunning Thresholds Worth Writing Down

Q: What's the difference between dunning and retry logic?

Retry logic is the technical mechanism (processor attempts charge again). Dunning is the business strategy around retry count, timing, and messaging. Dunning is the policy; retry is the execution.

Q: Should we pause or cancel after the final dunning attempt?

Pause is better for retention and chargeback prevention. The customer can resume by updating their payment method. Cancel is simpler operationally but loses the customer entirely. Most DTC brands pause for 30 days, then cancel if no action is taken.

Q: How many retries is safe before chargebacks spike?

3-4 retries with 3-7 day spacing is the safe zone for most DTC brands. A 5th retry on day 14 adds 2-4 percentage points of recovery but increases chargeback risk by 0.1-0.2 percentage points. Only worth it if LTV justifies the risk.

Q: What's the most common dunning failure?

Silent failure: the policy exists on paper but isn't configured in the payment processor. Default processor settings often retry only once. Verify your Stripe, Shopify Billing, or Adyen configuration explicitly matches your written policy.

Why Dunning Thresholds Matter

Most failed charges are recoverable. Card declines happen for temporary reasons - insufficient funds, fraud holds, expired cards with auto-renewal on file, or processor timeouts. A customer who fails on day 1 of their billing cycle often has funds available by day 3 or 5.

Without a written dunning strategy, teams either retry too aggressively (burning goodwill, triggering chargebacks) or too passively (accepting churn that could be prevented). The difference between a 1-retry and 4-retry strategy is typically 8-15 percentage points of recovery rate.

Shopify Billing and most payment processors (Stripe, Adyen, PayPal) support configurable dunning. The threshold decision is operational, not technical. It requires knowing your customer cohort, chargeback tolerance, and LTV math.

Core Thresholds to Document

A dunning policy should specify four parameters: retry count, retry intervals, messaging strategy, and end state.

Retry count: 3-4 attempts is standard for DTC. First attempt on day 0 (charge date). Retries on days 3, 5, and 7 recover 70-85% of what's recoverable. A 5th retry on day 14 adds 2-4 percentage points and increases chargeback risk.
Intervals: Exponential spacing (3 days, then 5 days, then 7 days) outperforms linear spacing. Reason: customer cash flow cycles. Paycheck deposits cluster mid-month and around the 1st.
Messaging: First retry is silent (no email). Second retry triggers a soft notification ('We couldn't process your payment - update your card'). Third retry is urgent ('Your subscription will cancel in 2 days'). Fourth retry is final notice.
End state: After final retry fails, either pause the subscription (customer must manually resume) or cancel. Pause reduces chargeback risk and preserves the customer relationship. Cancel is simpler operationally but loses the customer.

Threshold Calculation by Chargeback Tolerance

Chargeback rate is the limiting factor. Visa and Mastercard enforce a 1% chargeback threshold. Exceed it and the processor flags the account. Sustained rates above 1.5% can trigger account suspension or termination.

Each retry attempt increases chargeback risk by ~0.1-0.2 percentage points (varies by processor and customer cohort). A 4-retry strategy with 50% recovery rate typically lands at 0.3-0.5% chargebacks. A 5-retry strategy can push to 0.6-0.8%.

Calculate your tolerance: (LTV of recovered customer) - (chargeback cost + processing fee + support cost) = net value of retry. If LTV is $200 and chargeback cost is $100, the 4th retry is only worth it if recovery probability exceeds 50%. Most DTC brands see 20-30% recovery on the 4th attempt, making it marginal.

Failure Modes That Kill Recovery Rates

Silent failures are the most common. A dunning policy exists on paper but isn't enforced in the payment processor configuration. Stripe, Shopify Billing, and Adyen all require explicit setup. Default behavior is often 1 retry only.

Messaging misalignment: Dunning emails are generic or missing context. Customers don't know which subscription failed or why. Add the product name, billing amount, and a direct link to update payment method. Recovery rate improves 5-10 percentage points with clear messaging.

No pause state: Cancelling immediately after final retry loses the customer entirely. Pausing the subscription (with a 30-day grace period to resume) recovers 3-5% of customers who update their card later.

Retry intervals too tight: Retrying every 24 hours exhausts customer patience and increases chargeback disputes. Spacing retries 3-5 days apart gives customers time to resolve the underlying issue.

No cohort segmentation: High-LTV customers and new customers should have different dunning policies. A $5/month customer may not justify 4 retries. A $500/month customer justifies 5-6.

Implementation Checklist

Before deploying a dunning strategy, validate these operational requirements:

Payment processor configuration: Log into Stripe, Shopify Billing, or Adyen and verify dunning settings match the written policy. Screenshot the configuration.
Messaging templates: Write and test all dunning emails (soft notice, urgent, final). Include product name, amount, due date, and payment update link.
Pause vs. cancel logic: Decide the end state and document it. If pausing, set a grace period (30 days is standard) and a resume workflow.
Monitoring dashboard: Set up alerts for chargeback rate, recovery rate, and dunning email bounce rate. Track weekly.
Cohort rules: If using segmentation, document which customer segments get which dunning policies (e.g., LTV > $300 gets 4 retries, LTV < $50 gets 2).
Chargeback response: Document the process for disputing chargebacks. Most are preventable with clear dunning messaging.

Benchmarks for DTC Subscription Brands

Recovery rate (percentage of failed charges that eventually succeed): 60-75% with a 3-retry strategy. 70-85% with a 4-retry strategy. Diminishing returns beyond 4 retries.

Chargeback rate: 0.3-0.6% is healthy. Above 0.8% signals aggressive dunning or poor messaging. Below 0.1% suggests under-retrying.

Time to recovery: 60% of recoveries happen by day 5. 80% by day 10. Retries after day 10 add minimal value.

Pause-to-resume rate: 15-25% of paused subscriptions resume within 30 days. Varies by product category and customer LTV.

When to Adjust Thresholds

Dunning policy should be reviewed quarterly. Adjust if chargeback rate exceeds 0.7%, if recovery rate drops below 60%, or if customer cohort changes (e.g., new geographic market with different payment behavior).

Test incrementally. If increasing from 3 to 4 retries, roll out to 10% of customers first. Monitor chargeback rate for 2 weeks before full deployment.

Document all changes with dates and rationale. This creates an audit trail and prevents policy drift.

Questions

FAQ

What's the difference between dunning and retry logic?

Retry logic is the technical mechanism (processor attempts charge again). Dunning is the business strategy around retry count, timing, and messaging. Dunning is the policy; retry is the execution.

Should we pause or cancel after the final dunning attempt?

Pause is better for retention and chargeback prevention. The customer can resume by updating their payment method. Cancel is simpler operationally but loses the customer entirely. Most DTC brands pause for 30 days, then cancel if no action is taken.

How many retries is safe before chargebacks spike?

3-4 retries with 3-7 day spacing is the safe zone for most DTC brands. A 5th retry on day 14 adds 2-4 percentage points of recovery but increases chargeback risk by 0.1-0.2 percentage points. Only worth it if LTV justifies the risk.

What's the most common dunning failure?

Silent failure: the policy exists on paper but isn't configured in the payment processor. Default processor settings often retry only once. Verify your Stripe, Shopify Billing, or Adyen configuration explicitly matches your written policy.

Want this on your account?

Thirty minutes. Bring the number that keeps you up.

Book a demo