Fraudshield Risk Classification
No Risk - always approved conversion as no risk is found
Low Risk - are not rejected by the system and marked as approved but show certain patterns but are not yet enough to reject it on this basis alone.
Medium Risk - stronger pattern that exceed the threshold (defined by Filter Settings). These conversions are flagged as rejected
High Risk - Strong patterns with with clear fraud reasons. (high amount of duplicate IP, duplicate device IDs, Bots etc.)
The risk calculation is additive meaning that when a conversion is a Duplicate IP / VPN and a Bot at the same time it will increase the scoring mark it as high risk. Fraudshield automatically ranks the reasons according to transparency / reproducibility showing the most transparent reason as rejection while taking other metrics / rejection reasons into consideration. Filter settings have a direct impact on the risk scoring.