Decline Intelligence
How LostChurn's Decline Intelligence engine classifies 210+ raw decline codes from any payment processor into actionable recovery strategies.
LostChurn's Decline Intelligence engine automatically classifies raw decline codes from any supported payment processor into standardized categories that drive recovery behavior. Instead of treating every failed payment the same way, Decline Intelligence ensures each failure gets the right response -- whether that's a silent retry, a customer outreach campaign, or an immediate terminal flag.
What Decline Intelligence Does
When a payment fails, your processor returns a raw decline code. The problem is that these codes vary wildly between processors, are often cryptic, and don't tell you what to do about them. Decline Intelligence solves this by:
- Normalizing raw codes from different processors into a common format
- Classifying each code by cause (why it failed) and action (what to do next)
- Assigning a recovery strategy with specific retry parameters, cooldowns, and escalation paths
LostChurn ships with 210+ mapped decline codes across all supported payment processors. The mapping is updated regularly as processors introduce new codes.
The Dual Taxonomy
Decline Intelligence uses two complementary classification layers to fully characterize every failure.
Classification Buckets -- Why It Failed
Classification buckets describe the root cause of the payment failure:
| Bucket | Description | Examples |
|---|---|---|
| Card Error | Problem with the card itself | Expired card, invalid number, incorrect CVC |
| Bank Error | The issuing bank declined the transaction | Insufficient funds, issuer unavailable, do not honor |
| Network | Infrastructure or processing issue | Timeout, rate limit, gateway error |
| Fraud | Suspected or confirmed fraudulent activity | Stolen card, merchant blacklist, fraud flag |
Recovery Categories -- What LostChurn Does
Recovery categories determine the action LostChurn takes:
| Category | Action | Expected Recovery Rate |
|---|---|---|
| Soft Retry | Silent automatic retry at optimized timing | 30-40% |
| Hard Customer | Dunning campaign prompting customer action | 10-20% |
| Terminal | Mark as unrecoverable, stop all retries | 0% |
| Unknown | Treat as Hard Customer (conservative default) | 10-15% |
How They Connect
Each classification bucket maps to one or more recovery categories depending on the specific decline code:
| Classification Bucket | Typical Recovery Categories |
|---|---|
| Card Error | Hard Customer (update card), Terminal (card deactivated) |
| Bank Error | Soft Retry (temporary), Hard Customer (contact bank) |
| Network | Soft Retry (transient issues resolve on their own) |
| Fraud | Terminal (never retry fraudulent charges) |
How It Works
Step 1: PSP Normalization
Different processors return different codes for the same underlying issue. For example, "insufficient funds" might appear as:
| Processor | Raw Code |
|---|---|
| Stripe | insufficient_funds |
| Braintree | 2001 |
| Adyen | Declined with refusalReasonCode |
| PayPal | INSTRUMENT_DECLINED |
Decline Intelligence normalizes all of these into a single internal representation before classification.
Step 2: Classification
The normalized code is matched against LostChurn's decline code mapping table. Each code receives:
- A classification bucket (Card Error, Bank Error, Network, or Fraud)
- A recovery category (Soft Retry, Hard Customer, Terminal, or Unknown)
- A subcategory for more granular handling (e.g., Balance, Card Expired, SCA / 3D Secure)
Step 3: Strategy Assignment
Based on the classification, Decline Intelligence assigns specific recovery parameters:
- Max retries -- how many times to attempt the charge
- Cooldown period -- minimum time between retry attempts
- Escalation path -- when to move from silent retries to customer outreach
- Terminal flag -- whether to immediately stop recovery
Step 4: Recovery Execution
The assigned strategy is handed off to the recovery engine, which executes retry strategies for Soft Retry codes or triggers campaigns for Hard Customer codes.
If a decline code is not in LostChurn's mapping table, it is classified as Unknown and treated conservatively as Hard Customer. This prevents wasted retries on codes that may not be retryable. Unmapped codes can be reported to support@lostchurn.com for expedited mapping.
Analytics
The Decline Intelligence dashboard provides visibility into your failure patterns:
- Failure distribution by classification bucket -- see whether your failures are primarily Card Errors, Bank Errors, Network issues, or Fraud
- Recovery rates by category -- track how well each recovery strategy performs
- Trending codes -- identify new or increasing decline codes early
- Cross-PSP comparison -- compare failure patterns across different payment processors
Next Steps
- Decline Categories -- full taxonomy of recovery categories, subcategories, and common codes
- Retry Strategies -- how classified codes drive retry timing and behavior
- Decline Code Reference -- complete API reference for all mapped decline codes
- Recovery Metrics -- measure recovery performance by category