Fraud Detection Benchmarking Report
How does your fraud performance stack up against the best in the industry? Feedzai is publishing the industry’s first data-driven benchmarking report that shows exactly what top-tier fraud prevention looks like and what it takes to get there.
Download Now
Find out where you sit, what the best are achieving, and what sets them apart.
- Discover what Best-in-Class institutions are achieving
- Understand the two metrics that define fraud program performance and how to improve both simultaneously
- Get a clear, tier-by-tier breakdown of what separates top performers
- Identify specific gaps that are holding your institution back
- Build a practical roadmap for moving up a tier
Why benchmarking matters more than ever
Fraud loss is one of the most visible metrics on any executive dashboard, but knowing your own numbers tells you very little without context. Are you performing in line with your peers? Are you ahead of the market or in the middle of the pack? And critically, what are the best institutions actually achieving? We’ve built an industry first like-for-like comparison that gives banks concrete targets to aim for rather than vague aspirations.
The metrics that matter
The report focuses on two primary KPIs, evaluated at a fixed intervention rate to ensure a fair comparison across institutions:
Value Detection Rate (VDR)
the proportion of total attempted fraud value successfully detected and prevented.
False Positive Rate (FPR)
the ratio of legitimate transactions incorrectly flagged per confirmed fraud case. Industry analysis shows that false declines and unnecessary alerts can cost institutions nearly three times more than the fraud itself, through lost revenue, customer complaints, and churn.
FAQs
What is fraud detection benchmarking?
Fraud detection benchmarking is an industry-level comparison of how financial institutions perform against their peers on key fraud metrics. By anonymizing and aggregating data across institutions, it reveals where a bank stands relative to the market and what the best performers are actually achieving, so institutions have concrete targets to aim for rather than vague aspirations.
What metrics are used to measure fraud detection performance?
Two primary metrics form the foundation of fraud benchmarking. The first is Value Detection Rate (VDR), the proportion of total attempted fraud value that is successfully detected. The second is False Positive Rate (FPR), the ratio of legitimate transactions incorrectly flagged per confirmed fraud case. Both are evaluated at a fixed 0.1% intervention rate to ensure a fair, like-for-like comparison across institutions.
Why is benchmarking important for fraud prevention?
Fraud loss is a highly visible metric, but knowing your own numbers tells you very little without context. Benchmarking shows where you sit relative to peers, what top-tier institutions are genuinely achieving, and critically, where your specific gaps are. That clarity turns a vague sense that things could be better into a precise understanding of what better looks like and what it takes to get there.
How often should fraud detection systems be evaluated?
System evaluations should occur continuously as fraud is not a static problem. Attack patterns shift, new typologies emerge, and model performance degrades over time as the threat landscape changes. Top-performing institutions don’t treat evaluation as a periodic exercise.