Intelligent Insurer – Baden-Baden Daily
Response by Justin Davies, Vice President, Region Head – EMEA
How is robotics likely to change the insurance industry?
Justin: Robotic process automation (RPA) has the potential to spark insurance transformation in a number of ways. First, it can free up resources for more valuable and strategic tasks. Second, it increases throughput and reduces turnaround time in operations. Third, it helps to improve overall accuracy across many manually-driven functions of the insurance lifecycle. RPA can be interpreted as an effective “virtual resource” in that it performs repetitive, rule-driven tasks, directed by software and existing applications. The technology is extremely versatile; Robots can work unattended for back office operations without requiring human input. Alternatively, they can work in a mode that requires human intervention and prompting at designated intervals.
As a use case, Xceedance recently completed an in-house RPA proof of concept with impressive and successful results for an insurance client. RPA was applied to policy processing tasks, and the initiative significantly cut the time to issue policies and significantly increased accuracy — so that insureds quickly receive correct and comprehensive policy forms. There was a marked improvement in productivity and reliability. Among the findings of the RPA initiative, the time to issue a policy was cut in half, there was 100 percent accuracy in formatting, and it took 90 percent less time to validate the policy information.
What effects could robotics have on re/insurers?
Justin: Robotics can have a significant impact on the workforce for both insurers and reinsurers. According to a recent study, approximately 80 % of insurers within the P&C domain are considering RPA integration. Because of this wide adoption, there is growing fear that robotic technology could ultimately replace humans. This is especially true of the underwriting function which was recently ranked fourth on the Center for Business and Economic Research’s report (CBER) of the top 10 list of occupations that can be automated. And while RPA likely will change the makeup of tomorrow’s workforce, human intelligence and intuition in risk evaluation and management will always remain as a steadfast and essential part of the industry. Insurance is driven by an understanding and appreciation of risk in conjunction with special conditions, contingencies, and a viewpoint to expect the unexpected. Only human experience, logic and foresight can address these factors. Rather than working in competition with robots, the workforce of tomorrow will likely be one in which humans and robots work in tandem. By taking over menial, back office tasks, robots can free up more time for humans to take more complex tasks like making machines and software even smarter. For underwriters as an example, this human-machine collaboration increases the possibilities for evaluating individual risks with greater precision and efficiency whilst also improving productivity and average revenue per underwriter.
How is artificial intelligence likely to change the insurance product offering?
Justin: Within the P&C domain, AI has the potential to dramatically transform processes for data retrieval and analysis while improving client interactions.
One way AI can change product offerings is related to client experience. According to a recent study by Celent, 37 % of policyholders surveyed said they use smart technology to communicate with insurance companies and preferred this method over human interaction. Another 21 % who have not used smart technologies indicate an interest in trying it. A growing trend is the use of chatbots or robotic assistants which are programmed to quickly answer client questions about policies, payments, and other important service issues. In comparison to conversations between human agents and policyholders which must be done one-on-one and which can vary in their level of quality, chatbots can interface with multiple clients at the same time and provide quick and accurate responses, thereby reducing wait times and ensuring consistent levels of detail and quality in the service process. Although the programming is in its initial stages, one key feature of chatbots is that they can learn through interaction. So, the more chatbots interface with insureds, the more they will be able to pick up on new information and improve the client experience over time.
For auto insurers (and increasingly boat insurers) the use of telematics or black box insurance is a form of AI, used to calculate premium based on the policyholder’s current driving/usage record and preferences – rather than on how individuals in similar categories have historically performed. Access to such data not only facilitates a more efficient and accurate claims process; it can also be an incentive for improvements in driver habits. And telematics can increase safety and security by triggering medical help, if needed, during an accident and protecting vehicles from burglary.
How is the IoT likely to change the workings of traditional insurance?
Justin: The rapidly unfolding Fourth Industrial Revolution (4IR) can be characterized as a blurring of physical and biological environments to create autonomous systems and networks with vast implications. 4IR is introducing a world where machines talk to machines (M2M) to continuously refine operations. Early examples of this include Siri or Alexa, self-driving cars, and voice recognition. As a result, IoT has created a tsunami of new data, resulting in greater accessibility, precision, and individualization. Throughout the industry, insurers are working to rationalize and integrate this new information.
One area that is especially influenced by the IoT is underwriting. At present, the industry is still working with the same questions devised in the 1970s to assess risk. For example, to price homeowners property risk, insurers typically ask standard questions such as how many bedrooms at the property, building material and roof type, and past claims history. Answers to those questions plus risk scores derived from the postcode, create a risk profile, and the client is charged a certain tier and amount of premium. But insurance companies are now looking to obtain answers to those questions not by directly asking the client but by automating the process. More and more, insurers are collecting relevant customer and asset data from a growing number of data sources and platforms, and through social media and other online activities which track and process customer interactions and behaviors. With all of this new information, enough detail is emerging to enable insurers to price most individual risks with relative precision, be it in personal or commercial lines, rather than to broadly price across categories of risk. And while, industrywide, individual risk pricing may not be fully possible today at scale, the massive and growing data flow from IoT, coupled with continued innovation in intelligent technologies, can result in advanced machines becoming capable of individualized risk assessment and pricing in the not too distant future.