Download presentation
Presentation is loading. Please wait.
Published byBrittany Douglas Modified over 8 years ago
2
Are your chargebacks too low? - Manage your profits, not your fraud Akif Khan, PhD Manager, Solutions Architecture
3
Introduction “We have fraud under control. Our chargeback rate is less than 0.1%” Is the goal of a business to have the lowest chargeback rate it possibly can? Is it to maximise profits at all costs, irrespective of chargebacks? Is there a way to do both?
4
Risk Management Pipeline Fraud Claim Management Accept Reject Automated Screening Manual Review Tuning & Management Fraud Claim Management 4 Automated Screening 1 Manual Review 2 Accept Reject 3 ORDER RETAINED REVENUE (PROFIT)
5
Pipeline Dynamics Fraud Claim Management 4 Automated Screening 1 Manual Review 2 Accept Reject 3 ORDER RETAINED REVENUE (PROFIT) Staffing & Scalability Lost Sales Fraud Loss & Administration Review Rate 2005: 35% 2004: 34% 2003: 26% 2002: 20% 2001: 19% 2000: 16% Reject Rate 2005: 3.9% 2004: 5.9% 2003: 4.6% Fraud Rate 2000: 3.6% 2001: 3.2% 2002: 2.9% 2003: 1.7% 2004: 1.8% 2005: 1.6%
6
Automated Screening Question: Which tool is the best to use? AVS CVN/CSC/CVV/CV2 Address checking IP geolocation checking Velocity checking Hotlists Neural network scoring VbV/SecureCode Answer: All of them. Base decisions on a full suite of data across all sales channels.
7
Manual Review Best friend, worst enemy Human judgement Expensive Not scalable Customer experience issues
8
Accept/Reject rates Fraud Model Rejecte d Orders Rejected Fraud Reject Good Rejected Fraud # Detected Fraud Reject Good # Reject Good Not all rejected orders are attempted fraud Can estimate ‘fraud detect rate’ Detected Fraud / (Incurred Fraud + Detected Fraud) Can estimate ‘false positive ratio’ Reject Good / Detected Fraud Can estimate ‘insult rate’ Rejected Good / Total Good
9
Fraud Claim Management Chargebacks tend to understate fraud Chargebacks can account for less than 50% of all fraudulent orders. Many companies issue direct credits to customers in response to fraud claims. When examining fraud loss, must capture losses due to both chargebacks and credit issuance.
10
Calculate Costs How much is fraud actually costing you?
11
Build a Model Model different scenarios Vary chargeback rates, false positive rates, review rates Model the effect these have on profit
12
Profit Optimisation….. …not necessarily chargeback elimination.
13
Summary Determine the profit and loss drivers at all stages of your risk management pipeline. Focusing on chargebacks alone is inadequate. Determine the profit optimisation point for your business, and let this drive your acceptable chargeback rate.
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.