Improving Chargeback Protection Acceptance Rates

📈 Maximizing Chargeback Protection

To get the most out of Approvely's chargeback protection, it’s critical to pass data that clearly ties each transaction to an individual user.

Our system depends on identity signals—such as email, name, device behavior, and historical activity—to evaluate the legitimacy of each transaction and reduce chargeback risk.

📘


🚨 999 Error Code Explained

A 999 error means our chargeback protection provider reviewed the session data but declined to approve the transaction.

This is not based on a single factor—AI models evaluate thousands of real-time signals.

These errors are most common:

  • Early in an integration
  • When key user data is missing or low quality

To improve approvals:

  • Use the Events API to send key user events:
    • Registration
    • Login success/failure
    • KYC verification

These signals help the system build a user risk profile before their first transaction.


🤖 Why This Data Matters

The provider’s models evaluate:

  • Average transaction size
  • Behavioral patterns in your vertical
  • Card and identity history

To optimize approval rates, the model should be trained on your specific users.

Before going live, we recommend sharing:

  • ✅ Historical transaction data
  • ✅ List of past successful customers

This allows the model to:

  • Recognize returning buyers
  • Learn expected behavior
  • More accurately flag risky users
⚠️

Without this data, the system defaults to caution, which may reduce initial approvals.


✅ Best Practices by Customer Type

If You Have Card Transaction History

FieldDescription
paymentAmountTotal transaction amount
paymentMethod"card" (or other method)
last4Last 4 digits of the card
binBank Identification Number
pan (optional)Full card number (if securely stored)
firstNameCardholder’s first name
lastNameCardholder’s last name
emailEmail used at time of transaction

Helps the model match identity to historical card behavior.


If You Have User & Payment Data (No Card Info)

FieldDescription
firstNameUser’s first name
lastNameUser’s last name
emailUser’s email address
paymentAmountTransaction amount
paymentMethode.g., "wallet", "bank"

Enables modeling of expected users and flows.


If You Have No Transaction History

FieldDescription
firstNameLikely transactor’s first name
lastNameLikely transactor’s last name
emailEmail address

Start building buyer profiles early.
If available, share lists of:

  • Returning buyers
  • Whitelisted/safe users
  • Marketing or newsletter recipients

🧾 Summary: What Data Points to Send

Your SituationRequired Data Fields
Have card transaction historypaymentAmount, paymentMethod, last4, bin, pan (opt), firstName, lastName, email
No card, but payment historypaymentAmount, paymentMethod, firstName, lastName, email
No history availablefirstName, lastName, email