Fraud & Chargeback Risk Score
Assess transaction fraud risk and chargeback probability. Enter values for instant results with step-by-step formulas.
Worked Examples
Example 1: E-commerce Order Risk Assessment
Problem: An e-commerce store receives a $450 order: new customer, expedited shipping, partial AVS match, new device, US IP but shipping to different state. Calculate fraud risk.
Solution: Risk Factor Calculation:\n\n1. Order Value ($450 - medium-high)\n Score: +15\n\n2. Customer Tenure (new)\n Score: +20\n\n3. AVS Match (partial - ZIP only)\n Score: +15\n\n4. Device Fingerprint (new)\n Score: +10\n\n5. Shipping (expedited)\n Score: +10\n\n6. IP/Shipping Mismatch\n Score: +10\n\n7. Velocity (assume normal = 3)\n Score: +24\n\nTotal Risk Score: 104 β capped at 100\nRisk Level: HIGH\n\nChargeback Probability: ~15%\nExpected Loss: $450 Γ 15% = $67.50\n\nRecommendation: MANUAL REVIEW\n\nRisk Mitigation Steps:\n1. Call customer to verify order details\n2. Request 3D Secure authentication\n3. Verify phone number matches card file\n4. Check email domain (free email = higher risk)\n5. Consider holding shipment 24-48 hours\n\nDecision Matrix:\n- If customer verifies: Sh
Result: Risk Score: 100/100 (HIGH) | 15% chargeback probability | Manual review + verification required
Example 2: Subscription Service Fraud Check
Problem: A SaaS company sees a $199/month subscription signup: established customer (2 years), known device, full AVS match, but 3 failed payment attempts before success. Assess risk.
Solution: Risk Factor Breakdown:\n\n1. Order Value ($199 - medium)\n Score: +15\n\n2. Customer Tenure (established - 2 years)\n Score: +0 β Excellent\n\n3. AVS Match (full)\n Score: +0 β Excellent\n\n4. Device (known)\n Score: +0 β Excellent\n\n5. IP Risk (low - consistent location)\n Score: +6\n\n6. Failed Attempts (3 failures)\n This is unusual. Could indicate:\n - Card limit issues (legitimate)\n - Stolen card testing (fraud)\n - Technical problems (legitimate)\n Score: +20 (elevated concern)\n\nTotal Risk Score: 41\nRisk Level: MEDIUM\n\nContext Analysis:\n- 2-year customer = strong trust signal\n- Known device = account takeover less likely\n- Full AVS = card holder controls card\n- Failed attempts = could be legitimate limit issue\n\nNet Assessment:\nThe customer history st
Result: Risk Score: 41 (MEDIUM) | Customer history mitigates risk | Auto-approve with monitoring
Example 3: High-Value Electronics Purchase
Problem: An electronics retailer receives a $2,500 laptop order: new customer, shipping to freight forwarder address, overnight shipping, VPN-masked IP, new device. Full AVS match.
Solution: Risk Factor Calculation:\n\n1. Order Value ($2,500 - very high)\n Score: +25\n\n2. Customer Tenure (new)\n Score: +20\n\n3. Shipping Address (freight forwarder)\n This is a MAJOR red flag\n Score: +30 (additional penalty)\n\n4. Shipping Speed (overnight)\n Score: +20\n\n5. IP (VPN detected)\n Score: +25\n\n6. Device (new + VM indicators)\n Score: +30\n\n7. AVS (full match)\n Score: +0\n Note: AVS match doesn't offset other signals\n\nTotal Risk Score: 150 β capped at 100\nRisk Level: CRITICAL HIGH\n\nFraud Probability: 40-60%\nExpected Loss: $2,500 Γ 50% = $1,250\n\nThis has ALL the hallmarks of card-not-present fraud:\nβ High-value electronics (resalable)\nβ New customer (no history)\nβ Freight forwarder (international reshipping)\nβ Rush shipping (get goods before charg
Result: Risk Score: 100 (CRITICAL) | 50%+ fraud probability | DECLINE - classic fraud pattern
Frequently Asked Questions
What is a chargeback and why does it matter?
A chargeback occurs when a customer disputes a transaction with their bank, reversing the payment. Merchants lose the sale amount plus fees ($15-100). High chargeback rates (>1%) can result in processor penalties, higher fees, or account termination.
How is fraud risk score calculated?
Fraud scores combine multiple signals: transaction velocity, address verification (AVS), card verification (CVV), device fingerprinting, IP geolocation, customer history, and behavioral patterns. Machine learning models weight these factors based on historical fraud patterns.
What is friendly fraud?
Friendly fraud occurs when legitimate customers dispute valid chargesβclaiming non-delivery, dissatisfaction, or unauthorized use by family members. It's harder to prevent than traditional fraud and requires good documentation for representment.
What chargeback rate triggers processor penalties?
Most processors flag accounts at 0.5% chargeback rate and apply penalties at 1%. Visa and Mastercard monitoring programs activate at 0.9-1%. Exceeding thresholds leads to higher fees, reserves, or termination. Keep rate below 0.5% for safety.
How do device fingerprints help detect fraud?
Device fingerprinting creates unique identifiers from browser, OS, screen resolution, fonts, and other attributes. Known devices from good customers are low risk. New devices with suspicious attributes (VM, VPN, fraud-associated) trigger higher risk scores.
Should I decline all high-risk transactions?
No. Aggressive fraud prevention creates false positivesβdeclining legitimate customers who then shop elsewhere. Balance fraud loss against customer friction. Consider step-up verification (3DS, phone) for medium-risk rather than blanket declines.