Twenty years ago, drama in the insurance industry was limited mainly to the conflict between the actuarial and marketing departments over the price of products. InsurTech has thrown our industry into a whirlpool of exciting—and sometimes scary—opportunities. There are now far more interesting considerations than price alone. Digital natives have changed expectations. Data and new technologies have created amazing possibilities. As important as technology is to the reinvention of our industry, human considerations are key to the ultimate success of InsurTech.
Actuarial science has always been a blend of technical and social sciences. The future increases in both data volume and variety will require actuaries to have a familiarity with techniques in machine learning, data mining, image processing and predictive analytics. The ability to effectively apply these learnings and outcomes will require familiarity with behavioral economics and ethnography. Based on their in-depth knowledge of the industry, actuaries have a unique opportunity to enable successful integration of technology across all areas of insurance.
Marketing and Distribution in a Big Data World
One of the hottest areas of InsurTech is insurance marketing and distribution. Insurers are faced with increasing amounts of data about their customers or potential customers. Actuaries are perfectly positioned to help interpret this data and identify relevant trends and implications for product design and customer solutions.
One of the biggest challenges for actuaries to tackle is the propensity to buy insurance. Traditionally, propensity-to-buy modeling requires analyzing the purchase behavior and characteristics of people who purchase insurance. The idea is that by better understanding the people who have insurance, actuaries will be better able to target similar people and have a better chance of being successful in their marketing efforts.
But that approach limits the possible scope of target marketing. Try supplementing this by flipping your way of thinking. For example, why not learn more about the people who aren’t buying insurance?
Ethnography, or the science of observing human behavior, is used increasingly in business to develop hypotheses about consumer behavior and motivations. Observation provides deeper insights than surveys, because people don’t always know how to articulate what they want or accurately reflect how they would behave in a real setting. Marketing teams use these insights and hypotheses to build new approaches to reach customers. Propensity-to-buy models that are built from these new insights can help marketing teams target their messages to new populations. Consumer insights also can inform development of new product designs. The combination of consumer behavior studies and analysis of data models can be leveraged to maximize response rates and conversion rates.
In addition to targeting new customers, there are many InsurTech solutions focused around consumer messaging. Some examples include chatbots (virtual advisers), personalized video and natural language generation. One way to maximize the impact of messages is to leverage behavioral economics—the psychology behind why people make the buying decisions they do. This involves testing hypotheses about the context in which the messages are presented and how this shapes human decision-making. As an example, in one test conducted by Swiss Re’s Behavioral Economics team, they saw a 168 percent increase in open rates simply by changing the subject line of the email. This is an extreme example, but it has been consistently shown that small and inexpensive changes in messaging can make a disproportionate difference in outcomes. Behavioral economics is an important consideration in optimizing the impact of technology solutions for marketing efforts.
Underwriting and Data
Increases in availability of information create a plethora of modeling opportunities for actuaries. Predictive analytics is a term widely used to describe the use of new data sources to underwrite cases more quickly. New analytics techniques make it possible to work with increasingly large and varied data sources. Data is derived from sources such as healthy behavior indicators, credit histories, public records and appropriate digital profiles. Health records, now increasingly available digitally, and image analysis, which is advancing rapidly, can now be added to the data mix. The pace of growth in types of available data challenges an actuary’s ability to identify and synthesize trends, and our skills and thinking will need to evolve accordingly.
Certainly, a big part of the human side of data relates to issues of fairness, transparency and data privacy. But it can also be about what consumers want from the buying process. Predictive underwriting usually is considered in the context of simplifying the transaction, but consider the possibilities if we think of it as a buying experience instead of a transaction. Some consumers find the process more credible if it’s rigorous. Because life insurance is an investment in a family’s future, consumers feel more confident when insurers take this concern seriously by providing a process that is both rigorous and secure. Others wish the experience was interactive because they value the ability to share their individual story.
There is a lot of information actuaries and underwriters collect that could provide interesting feedback for consumers. Some new property insurance products, for example, integrate services like warnings when a hailstorm is approaching, or recommendations on how their driving compares to others like them. By investing in understanding what consumers want most from the buying experience, actuaries can design products that use data to deliver greater value.
One of my favorite aspects of the life insurance industry is the alignment of interest between the insurance company and the end consumer. Both want the insured to lead a long, happy life. Many insurance companies are considering ongoing engagement programs revolving around wearable devices or coaching for chronic conditions. In addition to the altruistic benefits, there is some real financial incentive as well.
Consider this simple calculation to explore the financial impact of a behavior change that creates a successful long-term change to an individual’s mortality: Take into account a health intervention that results in a 1 percent mortality improvement for all future durations. The savings in present value of death benefit for an average 45-year-old male duration 5 with a $250,000 universal life insurance (UL) policy would be roughly $80. The savings increases with age, face amount and relative risk, and can vary further based on product design. Policyholder engagement or interventions are also likely to have an impact on persistency. The nature of the financial impact of that change will vary depending on factors such as product type and duration.
Such an analysis can support a business case for investments in technological engagement. The analysis also highlights the importance of achieving the desired behavioral change. Mortality improvement requires long-term improvements in healthy behavior. However, principles of behavioral economics will tell us that incentives built around long-term goals like a longer, healthier life are not as effective as incremental short-term achievements related to everyday life. Actuaries can be involved in interpreting which types of interventions have the greatest impact on behavioral change and, therefore, mortality improvement. Success will depend on the context around how programs are implemented and how they’re communicated to consumers.
As the quantity and type of available data changes, actuaries need to consider not only the implications for product design, but also ethical considerations. Insurance is a product that can only be bought before you need it, and consumers rely on protection coverage to be there when they do need it. For example, if wearables or sensors become prevalent and provide a longitudinal view of the changing risk profile of policyholders, what can or should be done with that information? What are the implications of early detection of disease? Do insurers have any obligation to share insights about a policyholder’s risks? Is it an increased opportunity for service-related products to be developed?
Ultimately, the basic purpose and function of insurance is solid, but the delivery and engagement mechanisms need modernization. It is important for actuaries to keep up with new technologies to support the increasing variety and volume of data. By collaborating across the business, actuaries can help ensure the success of new InsurTech-driven products. Companies with a solid understanding of how people’s daily lives translate into the digital world will have a strong competitive advantage.