Can behavioral science become our newest tool in customer-centric product development?
Interview with John Thiel, Gen Re behavioral science subject matter expert, by Ewa Widenka
We often talk about customers and their importance in product development. Recently I have been engaged in a conversation on behavioral science in innovation and a eureka moment hit me – behavioral science principles and theories can be great tools to help us understand our customers better.
Many retailers already use the expertise in this field to serve their customers and create better interactions with them. In insurance, it is still a topic that does not get a lot of attention, although pioneers like Lemonade seem to lead the way with using the advice and understanding that behavioral science offers. In a recent article, by their Chief Behavioral Officer, Dan Ariely, they discuss the issues of trust, or rather lack of it, that underpins the relationship between customers and insurers in their industry. The lack of trust and behavior that comes from it is something that we observe repeatedly when talking to insurance customers who often perceive the insurance purchases in a negative way due to it being a necessity created by law (like need to have auto/car insurance for example). It is quite interesting to see how Lemonade is trying to rebuild the trust with their customers through activities like electing a charity of the customer’s choice upon policy purchase. The charity will then be able to receive a share if company profits, which would encourage people to not submit fraudulent claims so as to not harm the cause of their choice.
These observations lead me to believe there must be many more phenomena that innovation in insurance could use to create a better relationship with the customer and place the customer in the center of the new product development. To deeper explore the topic, we spoke to John Thiel, who is a Casualty Facultative Underwriter at Gen Re and has recently completed his graduate studies in Behavioral & Decision Sciences at the University of Pennsylvania.
EW: What are a few key behavioral science phenomena that you think impact insurance broadly as an industry?
JT: Not too long after joining Gen Re, I became aware of white papers the company published discussing the potential impact of cognitive bias on underwriting performance, as well as research Gen Re has performed on improving life insurance disclosure rates by minimizing choice overload during the application process. This made me realize that there is quite a bit of potential to leverage behavioral science across the insurance value chain.
As an underwriter, I’m particularly interested in the ways we make judgments about risk. For instance, does reading about a large claim for one applicant impact my perception of the next one I’m evaluating? What if the next risk is one within the same industry? What if I’m underwriting the renewal of the same risk? The availability heuristic leads us to overestimate the likelihood of an event occurring (i.e., a loss) based on how easy that event comes to mind.
I can see behavioral science also playing a role when adjusting claims (adjusters realizing whether they are anchoring to a settlement demand; leveraging trust and reciprocity to improve the overall claim experience and reduce fraud). Lemonade is a good example here, as they trust their policyholders in making their own claim. When the claim is below a certain threshold, it is only quickly vetted by anti-fraud AI and settled immediately.
When it comes to behavioral science and innovation, I feel that, broadly, innovation implies a new and different way of doing things. So, incorporating behavioral science concepts into insurer workflow and risk analysis framework can still be considered innovative. Personal lines carriers may be better suited to marry behavioral science and digital innovation gave the prevalence of buying home and auto insurance online. The larger, established carriers are also well-equipped to test the impact of different ideas before scaling them.
EW: How do you think insurance is different from other industries in this regard?
JT: Insurance is tricky because of the uphill battle of convincing some customers that our products are important, even though insurance isn’t really tangible, and it’s difficult to predict whether one will “use” the product after it’s purchased. Flood insurance policyholder retention in the U.S. is a good example. Those who haven’t experienced a flood over a long stretch of time may think the relatively smaller upfront cost of insurance may no longer be worth it, despite the much larger benefit of reimbursement after a future loss. Thus, policies lapse, even if coverage was mandated. On the other hand, a recent flood loss tends to make an insurance purchase feel more worthwhile, since customers’ perception of future flood risk often increases. So, what can result is this evolving sentiment towards the insurance product over time based on whether you think a loss is going to happen. This helps explain why the median length of carrying a flood insurance policy was found to be 2-4 years. I think finding a way to reinforce the importance of insurance over the long run presents an opportunity to leverage behavioral insights.
Comparing again to the retail sector for instance, the probability of gratification after buying a certain good can be certain and immediate. Generally, when you go to the grocery store, there’s little question as to whether you will consume what you set out to purchase, nor do you have to wait months or years to do so.
EW: How can we impact some of the negative biases that are slightly more built-in in people’s natures in terms of innovation? For example, it’s always easier to find reasons to ‘kill’ an idea than to champion it further – any tips and tricks?
JT: I think it depends on who you ask; there are certainly a lot of proponents of innovative thinking. But, for those opposing innovation, it may be due to uncertainty and lack of know-how. We are much less likely to invest if we’re uncertain, so, in trying to persuade an on-the-fence decision-maker, clearly articulating the project’s benefits and the teams’ expertise could help. Being able to provide examples of other successes outside the firm, i.e. relating to the opinions and actions of relevant others, might also make a difference.
There is also an opposite behavior connected to innovation which can be true as well, when teams often prolong projects past the point at which the work is deemed no longer useful, leading to budget overrun. This is an example of the sunk cost effect – essentially the tendency to get your money’s worth once they are invested.
I’ve learned that we can all benefit from behavioral science, whether it’s in the C-suite or on the front line. We’re all human.
As we can see, behavioral science can explain many of insurance’s day-to-day activities, and perhaps holds more ideas on how we can all work in a better and more efficient and unbiased way, than one could think initially. When it comes to our original question of whether behavioral science can become a tool in customer-centric product development, I think we can safely say that it does hold a promise of better understanding the customer. I am definitely looking forward to applying some of the thinking into my next customer interaction.
For those who want to find out more about some of the behavioral science theories and principles, take a look at this resource.
John Thiel is a Senior Underwriting Specialist for Gen Re’s Casualty Facultative department, based in Philadelphia. He underwrites both individual certificate and program business for various clients throughout the New York Metro and Mid-Atlantic regions. He joined Gen Re in 2014 after roles in both Underwriting and Claims with Liberty Mutual Insurance Company.
John earned his master’s degree in Behavioral and Decision Sciences from the University of Pennsylvania. He holds undergraduate degrees in Psychology & Spanish from Lafayette College.
 Michel-Kerjan, E., Forges, S. L., & Kunreuther, H. (2012). Policy Tenure Under the U.S. National Flood Insurance Program (NFIP). Risk Analysis, 32 (4), 644-658. http://dx.doi.org/10.1111/j.1539-6924.2011.01671.x