Prediction: Customer Satisfaction.
In commercial insurance, the arrival of predictive modeling—a process that applies rigorous statistical techniques to large bodies of data—heralds a new era of risk management. By lessening the reliance on basic intuition and conventional wisdom, predictive modeling is already helping to manage the cost of risk and improve customer satisfaction. Its impact, and that of other analytical tools and approaches recently brought in from outside industries, should only grow in the coming years.
Consider, for example, the assignment of a claim to a claim professional. The old way of doing things, still in use at many companies today, is a rules-based approach that dictates that X type of claim goes to X type of claim professional. Companies, such as Zurich, that utilize the latest predictive modeling tools are able to employ a far more sophisticated methodology, simultaneously looking at dozens of variables associated with the claim before determining the skill level of claim professional needed. For example, different outcomes are likely in a construction worker’s case and an office worker’s case even when the injury is similar. Or, the fact that the injuries did or did not require initial hospitalization, in conjunction with the reporting delay and state of jurisdiction, can give rise to completely different outcomes. The list goes on and on. While some of these differences appear intuitive, the value of predictive modeling is that it allows one to quantify their significance.
Taking the myriad of claims factors into consideration all at once helps to identify which cases truly require handling by a top-level claim professional and which do not. Some companies have been willing to pay for only the most experienced claim professionals, and that option, of course, remains. Still, paying for expertise above and beyond what is required is often an unnecessary expenditure, especially as predictive modeling continues to strengthen the overall safety net.
“Predictive modeling goes to the entire cost of managing the claims operation, as well as to customer satisfaction,” says Tom Barger, Director, Zurich Services Corporation, Zurich in North America Claims. “It should give the insured confidence that their claims are being handled appropriately.”
The benefits of predictive modeling go beyond the potential expense savings from claim handling. Routing a claim to a claim professional with appropriately matched skills can result in fewer transferred files—speeding resolution—and in automatic referrals to fraud investigators, nurse case managers, or litigation specialists where appropriate. Providing claim professionals with a better understanding of what’s intrinsic to the claim enables them to quickly comprehend the issues that need to be prioritized and what tactics are most appropriate.
“If you know that something has a higher likelihood to escalate in severity, you may take a different approach to resolve that claim,” says George Hansen, Chief Financial Officer, Zurich in North America Claims.
One of the potential benefits to insurers making the investment in predictive modeling is reducing the cost of risk transfer, both from a handling and a claim cost standpoint. But, it is truly a win-win as the customer also benefits in several ways, including improved management of their risk portfolio.
“We can share [risk] linkages with customers that will be worth their time, energy and money to influence,” says Steve Hatch, Chief Claims Officer, Zurich in North America. “If the model shows, for example, that tenure of employees is very influential—if the cost of claims goes down with tenure—the customer may wish to do something about turnover, because their new employees tend to have higher-cost claims.”
“What’s really important from a customer standpoint is that we’re making this investment in predictive modeling to help minimize the total cost of risk transfer,” adds Hansen. “The customer benefits in that a lower cost of risk transfer eventually produces lower premium rates. And, for loss-sensitive customers who have a high deductible, it can directly lower their retained risk.”