Last week, I discussed the results of the Insurance Consumer Fraud Survey 2010. This week, I’ll talk about what insurers can do to detect and deny fraudulent claims.
Improve customer service
The survey showed that poor customer service can result in an increase in fraudulent claims.
Customers who believe that a carrier has their best interests in mind may be less likely to commit fraud. To that end, carriers should take steps to improve customer service.
- Ensure that customers receive correct and relevant information during all sales and service calls. This means ensuring that non-claims personnel are also trained in the claims process.
- Set policyholder expectations of what happens when a claim is filed. Equally important, meet those expectations during claim fulfillment.
- Consistently demonstrate care and concern for the insured.
Whereas improvements to customer service can dissuade individuals from committing one-time fraud, analytics tools are needed to deal with more systematic, organized fraud.
- Set up business rules to detect anomalies. Identify and screen claims for the most common types of fraud, and adopt a set of business rules that flag these instances. Also determine if and when to escalate these instances to special investigation personnel.
- Incorporate predictive modeling. Analyze historical fraudulent claims and identify predictors of fraud to identify possible fraud earlier in the process. When used in conjunction with other analytics tools, this can be a very effective part of a holistic fraud detection system.
- Link analysis. Link analysis has been a very useful tool for special investigation units, and the increase in social media has furthered this. New technology not only links parties but can also identify the strength of relationships to provide input into fraud risk scores.
- Think holistically. By executing these and other analytics tools holistically, carriers can develop a more complete sense of whether or not a claim is fraudulent.
Better ways to detect fraud
In the end, it isn’t just about customer service or technology. Certainly, these practices go a long way toward fraud detection and prevention—but fraud prevention also needs to be aligned with business considerations. Is there a connection between different referral processes and the incidence of fraud? Which analytics are proving to be reliable indicators of fraud?
Accenture has developed a Point of View on this topic. Download Improve Customer Service and Fraud Detection to Deliver High Performance through Claims.
You may also want to read my previous blog series about assembling an analytics toolbox: