Adyen Fraud Detection: How RevenueProtect Works Behind the Scenes

Adyen’s RevenueProtect is the fraud detection engine that ConvesioPay merchants benefit from with every transaction. It’s the same fraud infrastructure used by enterprise retailers globally — running real-time risk scoring, behavioral analysis, and adaptive fraud rules on every payment. This guide explains how RevenueProtect works and how it protects ConvesioPay merchants from fraud without generating excessive false positives that hurt conversion.

RevenueProtect Architecture

RevenueProtect operates as a pre-authorization fraud layer — evaluating each transaction before it’s submitted for authorization to the card networks. The scoring model combines:

• Global network intelligence: Transaction patterns across Adyen’s entire merchant network, enabling cross-merchant fraud signal sharing
• Per-merchant behavioral baselines: Each merchant’s own transaction history establishes what “normal” looks like for that business — unusual deviations trigger elevated scoring
• Device fingerprinting: Browser and device signals collected during checkout — device age, browser configuration, installed plugins, and behavioral signals
• Velocity and clustering analysis: Multiple transactions from the same device, IP range, or BIN sequence within a short window
• Data element consistency: Does the billing postal code match the card BIN country? Is the email domain consistent with other transaction patterns?

The 2–4 AM Fraud Pattern

One of the most consistently observed fraud patterns in ecommerce is the 2–4 AM transaction spike — a period when fraudsters test stolen card credentials at high velocity, anticipating that automated systems have lower response rates during off-hours. RevenueProtect’s velocity rules specifically monitor this pattern, applying elevated risk scores to unusual overnight transaction clusters. ConvesioPay merchants have their transactions protected by this rule set without any configuration.

Custom Risk Rules

RevenueProtect supports merchant-configurable risk rules that allow businesses to add their own fraud logic on top of the base model:

• Block or review transactions from specific BIN ranges with high fraud history
• Apply elevated scrutiny to first-time buyers above a specified amount
• Flag transactions where shipping and billing addresses don’t match by country
• Limit transaction velocity per card or email within a rolling time window

ConvesioPay account managers help merchants configure custom rules appropriate to their specific fraud profile.

Balancing Fraud Prevention and Conversion

The most important balance in fraud detection is minimizing false positives — legitimate transactions incorrectly blocked as fraud. A system that blocks 50% of fraud but also blocks 5% of legitimate transactions is far more costly than one that blocks 40% of fraud with 0.1% false positives. RevenueProtect’s machine learning models are tuned to minimize false positives, recognizing that blocking a legitimate customer is a real cost to the merchant.