User-Level Fraud Thresholds for Tender-Switching Transactions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing transaction fraud systems monitor card-level behavior, leading to missed fraudulent activities and unnecessary transaction declines, which negatively impact consumer experience and merchant/financial institution revenue.
Innovation Solution
A system and method that authorize transactions based on a user's tender-switching practices, determining a user-specific fraud threshold score to balance fraud risk against transaction completion, using a cross-issuer fraud detection computing system to analyze transaction history and adjust authorization decisions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If transaction fraud systems monitor card-level behavior, then fraud detection capability is improved, but false declines increase and consumer experience deteriorates
Solution Approach 1:
The patent merges multiple card-level monitoring data into a unified user-level spending profile. By consolidating transaction history across all cards issued to a user, the system creates a comprehensive baseline of legitimate spending patterns, enabling more accurate fraud detection that reduces false declines while maintaining consumer experience.
Solution Approach 2:
The patent transitions from one-dimensional card-level monitoring to multi-dimensional user-level analysis by incorporating additional dimensions such as spending categories, locations, time patterns, and device information. This dimensional expansion enables the system to distinguish between legitimate variations in spending behavior and actual fraud, reducing false positives.
2Reliability
If suspicious transactions are declined without confirmation of fraud, then fraud risk is reduced, but revenue opportunities are lost and consumer experience suffers
Solution Approach 1:
The patent performs preliminary actions by proactively building user spending profiles and establishing baselines of legitimate behavior before fraud occurs. This advance preparation enables the system to evaluate transactions in real-time against established patterns, allowing legitimate transactions to proceed while identifying actual fraud cases, thus maintaining both security and transaction completion rates.
Solution Approach 2:
The patent implements feedback mechanisms where transaction outcomes (approved, declined, flagged) are continuously fed back into the user spending profile to refine and update spending patterns. This dynamic feedback loop allows the system to adapt to changing legitimate spending behaviors while maintaining fraud detection accuracy, reducing unnecessary declines of valid transactions.
3Reliability
If card-level monitoring is used, then individual card security is improved, but overall user spending patterns remain undetected
Solution Approach 1:
The patent merges data from multiple card-level sources into a unified user-level view, preserving individual card security monitoring while simultaneously revealing overall spending patterns. By aggregating transaction data across all cards issued to a user, the system identifies cross-card spending behaviors that would remain invisible at the individual card level.
Solution Approach 2:
The patent creates a multi-functional monitoring system that simultaneously performs individual card security checks and aggregate user spending analysis. The unified user spending profile serves multiple functions: detecting fraud on individual cards, identifying unusual cross-card patterns, and providing comprehensive financial behavior insights, thereby eliminating information loss.
Data Source
AI summary
Systems and methods are disclosed for establishing fraud detection system for authorizing payment for consumer transactions based on a user's tender-switching or transaction abandonment practices. One method includes: receiving transaction history of a user, the transaction history including a first payment vehicle and a second payment vehicle; determining, of the received transaction history, one or more instances of switching from one the first payment vehicle to the second payment vehicle; and determining a user-specific fraud threshold score for the user, based on the determined instances of switching from the first payment vehicle to the second payment vehicle.


