Human touch is key in claims automation
Success lies in using the technology to augment, not replace, human involvement.
Claims automation is a major trend in the insurance industry, helping to drive down loss ratios and do more with stretched-thin adjudication teams. Companies are exploring new data-driven technologies and focusing on the overall claims management process, from first notice of loss (FNOL) and satellite imagery loss assessment to automated triage. The goal is to distinguish between claims that can be handled automatically and those that require intensive human involvement. Further, there is a strong focus on fraud detection and prevention as insurers fear that fraud losses will increase if automation is done in too naïve a way. Digital customer experiences are often upgraded in tandem with an automation rollout. There are two focuses that I consider crucial when navigating a claims management automation journey:
First: Successful automation requires a laser focus on the customer
Second: Focus on necessary internal cultural changes
Handling claims is a key competence of every insurer, and the industry has built sophisticated processes to make claims handling efficient and consistent. But now, insurers are confronted with new challenges. One is inflation. We have seen a strong impact of social inflation on insurance claims since the 1980s. Insurers have not found a way to measure it, but the Insurance Information Institute recently gave some insights into this phenomenon and how it is spreading to motor insurance. Now insurers are confronted with soaring general inflation. Motor insurance claims in the U.S. are increasing significantly, but the premium rates are not following suit. This phenomenon is also occurring in most European insurance markets.
Alongside claims inflation, insurers face a new reality caused by large claims events. U.S. insurers have learned to deal with risks related to hurricanes, but now new natural catastrophes are adding up to the hurricane risk. The winter storm Uri in February last year, with freezing temperatures as far south as Texas, caused $15 billion in insured claims, reports Munich Re. In mid-December 2021, insurers were confronted with a series of tornados that caused damages of $5.2 billion. Four billion dollars of these losses were insured. Wildfires are happening more often, as well. Last year they caused $4 billion in insured losses, Swiss Re reports, but wildfire claims are often higher.
Consequently, insurers need to deal with large claims events much more often — and this happens in Europe as well. The unprecedented flooding western and central Europe experienced in mid-July 2021 caused $54 billion of damage, $13 billion of which were insured. After this event, Europe has become painfully aware of the protection gap and is trying to fill it.
These large claims events are not only a financial challenge — two Louisiana insurers were put out of business after hurricane Ida — for the industry but also an operational one. Typically, these claims are often large losses, and customers file them in an emotionally stressed situation.
Solution providers and insurtechs drive automation progress
Claims automation is nothing new. U.S. insurers have talked about it for the past 20 years, but companies and technology providers have only recently become serious. Take the Guidewire product Claims Autopilot, which just rolled out in late 2020, seeking to “flip a highly manual claims model to make manual processes the exception,” said Diego Devalle, global head of product and engineering at Guidewire Software. Windshield claims, for example, are usually the first fully automated process. We look to Europe for examples of some more leading-edge automation examples.
- In the Polish insurance market, there is competition among insurers in digital services and many of them are implemented in claims. Sollers Consulting has supported the Polish Talanx subsidiary Warta to implement a decision engine based on the open-source platform Camunda. It automates the process of First Notice of Loss (FNOL), including the loss assessment.
- The German insurer VHV has streamlined its claims handling process in a large, agile transformation project to drive automation. The insurer created a future-ready IT infrastructure based on Guidewire Claimcenter. This enables the company to increase the level of automation incrementally.
- The Norwegian insurance association Finans Norway has developed an app for car damage assessment. There is also talk of a new central claims registry that is discussed politically.
- AXA Switzerland has created its own insurtech to transform claims management in motor insurance; AXA UK has launched a new digital claims solution that delivers straight-through processing for its motor insurance customers.
- Aviva Poland introduced electronic claims filing for hospital-related claims.
Many insurtechs and solution providers support the automation of various functionalities of the claims handling process. Innovation Group, for example, started to organize a claim tracking service for clients of 40 German motor insurers. London-based Sprout.ai partners with Zurich, implementing Automated Policy Checking to accelerate property claims settlement times to less than 24 hours. The Danish insurer Codan has implemented an artificial intelligence (AI)-based claims system provided by Scalable Technologies. The Polish market leader PZU introduced an AI-based solution of Tractable to handle motor insurance claims.
However, most U.S. insurers still need technology that supports automated workflows. Putting a workflow engine on top of a legacy system usually does not work long-term. As we see from our European counterparts, broad partner support is required to truly automate. Windscreen replacement providers, payment processors, fraud detection partners, towing partners, field appraisers and repair shops must be integrated into one digitized process.
Successful automation requires a laser focus on the customer, not on efficiencies or cost reductions
We often see efficiencies as the key business driver in many automation initiatives, but it is crucial to start with the customer in mind and consider how and when human touch is required to drive a high quality of service. Let me give you a personal example.
After our newest was born, we bought a new Mercedes SUV and waited six months to receive it. Within the first weeks, it was backed into at a gas station. After that incident, I was in no mood for automated claim processing and required our brand-new SUV’s claim to be handled with “kid gloves.” When I phoned in the third-party claim notification, the FNOL-representative tried to push me to download their app and do things online. Frustrated, I declined. Most will agree that claims automation was not a fit in such a situation, and the FNOL representative should have chosen a different claims strategy.
Claims are an emotional and stressful time for the consumer. If the FNOL processes fail to identify the customer who needs personalized service, then that customer’s claim experience will be poor. Sixty-three percent of U.S. consumers will sever a relationship with a business due to poor customer service. Sometimes a claim will be the only touchpoint a customer has with an insurer for years, or ever. It will define their view of the insurer. Therefore, it is crucial to find the right balance between automated tools and personalized customer service. Insurers should determine whether the customer needs speed or the highest quality and personalized service.
Chatbots are, in many cases, the insurer’s answer to automation. However, their rollout must be judicious and match use-cases where the customer will deem them useful and not a hindrance. Modern technology with cognitive process automation (which does not require structured data to operate) is helping make bots more appropriate, more human-like and, consequently, more effective. Much of whether an insurer can use cognitive process automation or robotic process automation comes down to if the backend systems can support unstructured data.
Focus on internal necessary cultural changes
Claims is traditionally a rigid process. Introducing flexibility via automation will be seen as a huge positive and even a differentiator for the carrier. By automating claims management, managers hope to reduce manual effort and increase cost efficiency. Much focus is on shortening the average claim time and reducing the number of tasks or actions the adjuster must take.
Alongside the technology for claims automation must be an internal cultural change. Helping adjusters, FNOL representatives and other teams embrace hands-off or low-touch processing is crucial to implementing the right tech stack. Claims adjusters and managers will need training in how to handle the copious amounts of analytics data resulting from automated processes. Most are not trained or not skilled in how to make process changes based upon the results.
Adjusters don’t need to become data scientists, but they should know how to adjust based on AI results. Seasoned adjusters can detect fraud and customer frustration better than technology — however, technology is rapidly advancing in this area. Thus, where their interactions with the customer are positioned in the workflow is critical, and we should never forget the carrier’s customer culture must also support claims automation.
To summarize, when embarking on a claims automation project, or iteratively improving your automation, ensure that the tech stack is capable, that you have a robust integrated partner ecosystem and you have focused on the customer journey first. Consider your internal teams for cultural changes required and any training needed to benefit from the new analytics outputs.
Justin Hall (justin.hall@sollers.eu) is Vice President North America, Sollers Consulting.
Related: