Our own experience is that an effective fraud detection program can yield benefits including savings in the range of one to three percent of total claims paid out, depending upon the maturity level of the policy. The best approach for achieving these savings relies upon teamwork—combining better service levels and more efficient claims processing, as well as advanced analytics tools for greater focus in detecting and fighting fraud. Better service levels, in particular, generate positive customer feedback, helping increase customer loyalty and deterring individual episodes of fraud.

As our earlier U.S. research indicated, customers are more likely to commit fraud when they feel that service levels are substandard. Click on the next page to learn about four proven strategies for addressing fraud.

Critical Strategies to Win the Fight

1. Using business rules to detect irregularities. Business rules can allow for the identification of anomalies or irregularities during the processing of claims. Such rules, for example, compare claims based on various types of fraud (individual or organized) and determine whether experts should investigate the fraudulent incidents. Sophisticated analytics tools can also detect such irregularities, saving insurers money, time and effort.
2. Employing predictive modeling. Predictive analytics methods and models can be used to review historical fraudulent claims and identify factor and elements that can help prevent future fraud. The goal is to detect potential fraud as early as possible in the claims process and thus reduce payments made to fraudsters. When predictive modeling is combined with other tools as part of a concerned global effort, it can be highly effective.
3. Undertaking network analysis. Examining the relationship that exists between concerned parties can help specialized investigators detect organized insurance fraud. Technological developments in this area can help investigators link different players. New technology solutions can help identify the extent of the relationship between the investigated parties, and the information and insights gathered can be used to define indicators that point to possible fraudulent activity.
4. Closing the loop and intervening. These fraud detection strategies—including the integration of fraud indicators within the claims handling process—can work in combination to increase the overall effectiveness of fraud prevention. When a risk is identified, immediate action can be taken, including investigation by specialists and interaction between the claims management unit and the insured.

Along with improving customer service and overall claims processing, these are the basic elements in reducing fraud losses. At the next level, insurers can develop a claims analytics platform with a common infrastructure, using a broad range of models, including those addressing injury, subrogation, litigation and large loss claims.

Fraud analytics, along with other tools and methods, can be expanded to address other claims issues. The broad use of analytics and statistical tools—along with improved productivity in managing claims and a reduction in fraudulent claims—can help insurers generate important cost reductions. For example, optimizing the validation of claims process and using more robust analytics to make the process more effective can reduce the overall number of fraudulent claims managed. Similarly, reducing the manual effort needed to analyze data allows for more time spent on reporting that really adds value.

Few insurers would argue that increasing the rate of fraud detection, as well as fraud avoidance and recovery rates, can have a significant impact on profitability. Insurers pursuing an approach that combines fraud detection with improved service and processing capability, however, can realize long-term benefits in terms of customer satisfaction and retention. A concerted, coordinated effort is most likely to yield the best results.