The Potential of Artificial Intelligence
In a viral State Farm commercial, Oscar Nunez, playing the role of a friendly, helpful insurance agent, disparages any insurance company that utilizes artificial intelligence in support of customer service activities.
“Nothing compares to an actual human being helping out whenever you need,” he quips to NBA basketball star Chris Paul. “Plus, humans won’t start crying weird AI tears all over your car.”
It’s a clever dig at other insurance companies that are investing in bots or automated responses in their customer service departments. In a recent PwC survey, “Experience is everything: here’s how to get it right,” it is reported that 82% of U.S. and 74% of non-U.S. consumers want more human interaction now and in the future. So, while the majority agree with Chris Paul, AI and machine learning solutions can positively impact insurance agencies in other business processes.
Where paperwork is concerned, there are a few key areas where AI can boost customer experience in a manner that does not affect that personalized human interaction.
In the insurance industry, there is a constant battle to improve customer experience in order to retain clients and elevate satisfaction ratings. Referencing another State Farm customer-first campaign, the phrase, “Like a good neighbor, State Farm is there,” calls to mind easy-to-contact, helpful agents. Outside of a friendly agent though, claims processing time and new policy applications are major indicators for client satisfaction and policy retention.
By adopting AI-powered tools to support claims processing, insurance agencies can automate claim document ingestion and review. Taking largely manual processes, such as identifying incoming correspondence, indexing metadata and routing claim files to their appropriate review queue and automating it with an intelligent capture tool can accelerate adjudicator access to claim-related documents and customer correspondence. This simple automation has the capacity to improve key areas in the customer success matrix without adding additional headcount for faster document processing.
According to the Insurance Information Institute, fraudulent claims are responsible for approximately 10% of P/C insurance losses and loss adjustment expenses each year. These losses are rolled into policy premiums and passed on, at least in part, to the policyholders. It follows that less fraud means lower policy cost. However, combating fraud is an expensive task that can necessitate the manual review of claims and slower claim processing time. And even with additional headcount, humans can still overlook or miss suspect claims.
Inarguably, data science and AI have made fraudulent claim detection a more easily achievable goal. However, an algorithm that determines whether a claim against a property policy falls outside the threshold of expected damage compensation is only as good as the data on which it is running. Applications that identify outliers and suspicious patterns across large sets of insurance claims require structured data sets in order to analyze information. It’s imperative that the largely (document) form-based claim submission process incorporates intelligent capture in order to feed the analytics engine without a human employee having to manually enter claim data into backend systems.
Crying robots aside, there’s nothing to fear from AI robots in the insurance industry.
Science fiction movies have perpetuated the fear that technology is going to replace human workers. But that fear is largely unfounded.
For the insurance industry, cost savings from expedited back office processes and claims review can be applied to customer service departments and customer experience initiatives. Diminished overhead expenses result in less expensive policies. And, applications that employ AI to automate and expedite processes must be embraced to meet the customer service demands.
Campbell is the director of business development at Ephesoft Inc.