Manual Review
When an order is denied, a Fraud Analyst reviews the order manually based on the order’s data points, indications, and open-source intelligence.
The analyst evaluates relevant parameters, such as:
The geo-location
Position related factors
Repetition of data points such as the billing and shipping address, phone number, IP address, and payment method.
Products
Historical data
Buying patterns
After the analysis is complete, the analyst determines if the risky profile is legitimate or a fraudulent attempt. The fraud engine then utilizes the analyst’s decision to enhance the model’s parameters for the Merchant’s future orders.
As the fraud engine screens more transactions, it continuously improves to accurately identify and scrutinize only genuinely suspicious transactions.