Hunter is a credit application fraud prevention tool developed by Experian. It is used by more than 100 clients worldwide and has saved its users over £1 billion to date.

Hunter looks for inconsistencies in application form data provided and then checks the current information against any previous applications, plus known suspect and known fraud data. The checks are performed through the use of sophisticated algorithms, controlled by the end users. For every match that is triggered, a risk rule code is returned. Referral teams are then able to use these triggers as a guide in their decision making. A natural progression from the rule-based trigger solution is to prioritise the matches by the use of scoring.

Experian Decision Analytics, which also developed the scores used in Detect, has developed an innovative scoring solution that enables Hunter users to both optimise their referral work lists and automatically accept low risk applicants.

The methodology is based on the grading of the Hunter rules into 10 risk-based groups. The count of the number of rule hits in each group then moves forward into a regression-based scoring model. The resulting Hunter score can then help to:-

• sort work queues by risk of fraud
• work high-risk cases first that have the largest loss potential
• automate decisions for low probability of fraud scores
• allocate suitably skilled staff to work cases according to their pre-defined levels of risk.

Compared to the standard rule-based approach, the score allows referrals to be reduced by 15%, whilst still identifying 95% of total application frauds.

The score is scheduled to be integrated into the Hunter system by Q3 07.

Adrian Paine - Analytics Consultant, Experian Decision Analytics

Contact us for further discussions about this article

View this article in German

This article originally featured in e-news Analysis and scoring update 2007. Click here to request a copy.

Top