Has the COVID-19 pandemic changed the Insurance industry forever?

Social media is brimming these days with predictions of what the new normal will look like after the coronavirus pandemic recedes, and chief amongst the debates is the topic of whether the insurance agent or broker has been made redundant by the direct-to-consumer model. With a severe restriction on being able to meet brokers face to face, the vast majority of insurance policies taken during the lockdown have been facilitated purely through digital/online channels. Essentially a well designed and developed digital onboarding process has become the defacto norm during the COVID-19 pandemic.  Coronavirus may well be the catalyst to disruption of the insurance industry in the same way the 2008 financial crisis was for banking.  If insurance firms were not prioritizing the acquisition and AI driven underwriting of new customers through digital onboarding channels, they will be now.

In recent years, comments made by the White House claimed that it was highly likely that intelligent automation could relatively easily replace most jobs which command a salary of less than $20 per hour. The rapid evolution of AI over the past 5 years has come about as a general combination of technologies like processing power, data storage and analytics have converged to create the perfect storm of machine learning, cognitive computing and ‘robo-underwriting’. For the average consumer, the results are obvious in our daily lives through the use of search engines and contextual prediction, but increasingly more complex use cases are going mainstream within the insurance industry - like actuarial based data underwriting. Unsurprisingly, the insurance industry is fast becoming an area of lucrative opportunity for Insurtech firms that have no legacy systems to hold them back and are naturally embracing what AI and design thinking can do for the customers experience.

The era of the robo-life agent has arrived and is now empowering insurance carriers with the ability to source and construct custom and personalized life insurance portfolios, monitor their policies in real-time, and develop pure digital products never conceived before. What is already undeniable is that robo-life agents can provide substantially more efficient and superior solutions to those traditionally done by humans and are considerably more cost effective. Consequently, the vast majority of life insurance agents today will, in time, become a memory of the past. Naturally there will be resistance to such a change, in the same way taxi drivers resisted the Uberization model, but as AI competition begins to operate at optimal efficiency levels the advances in cognitive computing will act as a catalyst and make the evolution of the insurance industry a fait accompli.

Amongst the primary impact that artificial intelligence will have on the industry of ‘risk’ will be around the areas of operations improvement and the automating of existing and repetitive front-end interactions (like onboarding a new customer), underwriting and claims processing. Common use cases for this will be in client engagement and improved lead conversion ratios, reducing quote-to-bind and FNOL-to-claim resolution times, and accelerating new digital product launches ahead of competitors. Over time however, artificial intelligence will likely have a more profound impact - it will recognize, evaluate, and underwrite incipient risks and detect new revenue streams. By its very nature, AI will rapidly increase in sophistication and will significantly improve cross-sell and up-sell prospects by converting them to customers, hone risk assessment and risk-based pricing, and elevate the service levels around claims adjustment. The unfortunate outcome of this is that many human roles are likely to become redundant over time – brokers, agent advisors, call center staff, claims representatives and traditional underwriters are slowly disappearing. In their place will be digital-first organizations that have streamlined and automated insurance platforms that deliver superior and personalized experiences to the consumer.

Insurance-focused tech or “InsurTech” is creating new opportunities for insurers facing a barrage of business challenges. Escalating regulatory constraints and an aging workforce are forcing many insurance firms to identify how technology can both enable better differentiation and a more efficient business. One of these disruptive new technologies is cloud-based artificial intelligence. In recent years, firms like Microsoft, Amazon and Google have made available open source cognitive tools and APIs that deliver dramatic operational efficiency using artificial intelligence, machine learning, and cognitive computing. Such tools challenge traditional IT assumptions and provide a step change in how the organization will function in the future.

Underwriting in 2020

The COVID-19 lockdown has highlighted fundamental weaknesses in the traditional underwriting process. People unable to do medical assessments, meet a broker, or provide original documents have been left with few options to do this through alternative digital self-service channels. Automation advances thanks to artificial intelligence could have made dramatic impact on the underwriting process. Life insurance or Term underwriting has typically involved legions of actuaries attempting to evaluate the risk of insuring someone to determine the price of an insurance premium. The job of an underwriter is unenviable and fewer and fewer people are attracted to the position. As such, the underwriting workforce is aging. Additionally, human judgment can play a key role in even a seemingly mathematical algorithm-based determination. Individual assessors have their own predilection to assessing risk, and arguably, insurance companies lose billions of dollars either through inaccurate risk profiling or through onboarding leakage caused by overpricing.

The underwriting process is a perfect example of where artificial intelligence has transformative power. Much of what goes into underwriting has to do with the consumption of data, and largely unstructured data, flowing in from a proliferation of sources from social media to online news. Predictive technologies combined with machine learning can quickly ingest data and harvest accurate conclusions, improving the underwriting process as a whole.

Thynk Digital has identified the following areas of artificial intelligence opportunities in underwriting:

• Creating behavioral and predictive algorithms for risk assessment using decision tree analysis and Bayesian networks.
• Leveraging the power of Natural Language (chatbots), Sensors and IoT proliferation in homes and industry data to create a knowledge base of operational risk intelligence that rapidly matures machine learning.
• Standard automation mining techniques to generate operational efficiencies.
• Deep QA Systems which leverage various techniques to assist underwriters identify common risk attributes.

Claims Processing in 2020

Claims processing has long been a pain-point for both insurers and the insured. Managing claims entails significant manual effort from document processing to spotting potential fraud.

For example, robotic process automation (RPA) and optical character recognition (OCR) can be used in conjunction to automate document processing, to more quickly and accurately scan complex forms. From there, machine learning could be leveraged to determine fraud patterns or repair cost determination. Insurers can cut their claims processing times down from a number of days or weeks to just a matter of minutes. Furthermore, automated processes with inbuilt machine learning are inevitably more precise than humans, meaning that insurers can drive cost efficiencies from the sheer number of denials that result in appeals they may ultimately have to settle.

Thynk Digital has identified the following areas of opportunity in claims processing:

• Chatbot and Natural Language processing to handle Level 1 and 2 claims enquiries for the majority of policy holders.
• Artificial Intelligence using soft robotics to identify bottlenecks and improve efficiencies and ensure full compliance with standard operating policies around the claims process.
• Fraud identification in claims through the use of social network behavior.
• A combination use of intelligent drones and Bayesian networks and decision tree to effectively determine repair costs and automatically categorize the severity of damage to vehicles or homes involved in environmental incidents or accidents and build claims predictive models.

How Thynk Digital can help you.

Although artificial intelligence and machine learning may appear like technologies from the future, the reality is that these mainstream innovations can fundamentally transform the business of insurance. Thynk Digital is a Microsoft Gold Partner and has experts in Azure Cognitive Services. At our Global Innovation Center (Thynklabs), we work with leading global insurers to identify practical applications of artificial intelligence and to support them in building an effective roadmap for digital implementation.