The evolution of credit hire fraud detection
Intelligence Analyst, Ste Holmes, takes a look at the ever improving tech used by Keoghs for credit hire fraud identification.
The olden days
Keoghs Intelligence Services first started to review credit hire fraud (CHF) referrals back in the antiquated days of 2012. At that early stage, if a credit hire case handler had concerns in respect of a claim, it would be referred to the CHF team and the relevant details would be screened against our intelligence databases, whilst validation searches would also be conducted against the Credit Hire Organisation (CHO). At the team’s inception, such manual referrals could effectively deal with the volumes being transferred, which proved handy given it was the only real method available at the time.
The need for systematisation
As the number of credit hire claims handled by Keoghs increased, the volume of credit hire fraud identified also grew at a fast pace. As the team gained more experience, our handling and investigation skills developed to the point where the traditional method of manual identification needed to improve. It soon became apparent that the vast majority of referrals were being retained, with effective handling of these cases heavily reliant upon the early identification of the risk of credit hire fraud, even if the actual evidence of fraud was not yet apparent.
These risks are identified through indicators, the most common being the involvement of a particular CHO. We therefore began to conduct ‘mop up’ exercises, reviewing MI reports to ensure the identification of all claims featuring companies which had provided a negative experience. Whilst this enabled the team to step up detection rates, it also proved time consuming for both intelligence and CHF handlers, meaning an alternative solution was required.
The response to this problem required the facilitation of mass data analysis along with the identification of high risk claims within extensive data sets. The result has represented a sea change in credit hire claims handling for the Keoghs team.
Underpinning the solution is the Keoghs Intelligence Database which holds data relating to all claims handled by Keoghs arising out of concerns in respect of fraud. Central to this is our Advanced Data Analytics product (ADA), into which large amounts of data can be fed and matched against the claims data within the database. ADA then scores the results and flags the high risk claims.
With the credit hire fraud team’s primary concern being the identity of the CHO, we have compiled a list of over 300 organisations which present a potential fraud risk. ADA has the ability to match data against this list, flagging any relevant claims for review by a handler. Importantly the system also has the ability to identify duplicate and overlapping hire, adding to our arsenal in respect of identification.
Keoghs’ credit hire team has continued to expand and now handles in the region of 40,000 claims per year. All of these are screened through ADA and enrich our database of hire vehicles and hire periods which can be used to identify any conflicts.
ADA has now been operational within our CHF team for two and a half years, sitting alongside our manual and automated referral processes to complete a comprehensive identification model:
- ADA – flags claims featuring high risk companies
- Automated referrals - gives the team visibility of all new non-GTA payment packs received on a daily basis
- Manual referrals – claims where case handlers have identified concerns
This robust model enables us to validate suspect claims proactively and efficiently in order to set effective claims handling strategies from the outset of each claim.
With ADA as a crucial element in this process, we recently ran a review for the second half of 2016 which showed that two out of three claims flagged by ADA as high risk have been transferred to our Pre-Litigated Credit Hire Fraud team for permanent handling. This team has now consistently achieved average savings on the amount of hire claimed in the region of 90% year on year since 2014.
In January 2017, successful case reports prepared for clients showed that over 60% of the successes delivered were identified by ADA as high risk. Those claims included duplicate hire as well as claims featuring new to market CHOs which will now be flagged by ADA as high risk if we see any further claims.
There is no doubt that the development of an intelligence database specifically tailored to the credit hire market is a crucial tool. When combined with a product which can flag a variety of risk identifiers, the results allow us to step up the fight against this ever increasing area of fraud.
The CHF team continues to work closely with the ADA development team to ensure that ADA meets our needs, and its ability to identify duplicate and overlapping hire is a direct result of this close working relationship.
There is also scope in the future for ADA to assess numerous other important aspects of the claim when assessing its risk level, including value and key dates.
As ADA has been developed in-house by Keoghs, we now have a platform upon which we can continuously improve and expand and adapt our fraud identification processes, allowing Keoghs to stay ahead of the game.