Zurich’s account fraud team aims to create positive customer experiences by processing claims as quickly as possible. At the same time, they must prevent insurance claims fraud from slipping through the cracks. This means they have to work fast.
Fraudulent insurance claims can come in many forms. A fraudster might file a claim for a car accident that did not actually happen. Or they might claim damage from a hail storm when no storm actually occurred. Fraudsters may work individually or in organized networks.
To help process large volumes of claims without compromising on fraud prevention, Zurich Insurance Switzerland were early adopters of automated fraud detection.
Automation helps by flagging the potentially suspicious insurance claims cases among the mass of claims Zurich receives. But while automated detection is an asset in sorting through claims to identify the ones requiring the attention of the fraud team, it also produces a mass of alerts. The new challenge for the SIU team in Switzerland became deciding which cases truly appeared to be fraud, requiring further investigation, and which were merely false positives.
Case triage involved looking at data that had been collected manually in different source systems. It simply wasn’t efficient, and their existing tools didn’t help them see the big picture since there was no way to see the connections between the data.
The 25 members of the SIU also found themselves in a race against time. “Sometimes we have only minutes to react to a claim,” says Paul Kühne, Head of the Special Investigation Unit at Zurich Insurance Switzerland. Customers expect their claims to be processed without delay after submitting them online, so the team needs to make the right decision as quickly as possible. Having an efficient way to weed out false positives would also save time for investigators to spend on truly suspicious insurance claims cases.
The stakes are high on making the right call, since preserving a good customer experience is paramount for Zurich. “It’s hard to make an investigation against a customer. It’s hard to explain to honest customers why we are asking so many questions and why we need more time to process their claim,” says Paul Kühne.
The account fraud team needed a solution to add to their existing toolkit to help them see the big picture and make better informed decisions on their alerts while accelerating the decision-making process.