Applying Behavioral Science to Insurance | Peter Hovard

What is behavioral science and what excites you the most about this field?

Behavioral science aims to understand how people really think, make decisions, and act. Typically, we assume people will or should behave in certain ways: being honest and accurate on insurance applications, looking after their health, buying the right insurance or annuities for them, and so on. Unfortunately, how we’d like or expect people to behave is often different from how they really do. Often these incorrect assumptions lead to less efficient, effective, or even ethical products and processes.

Behavioral science brings together insights from disciplines such as psychology, economics, neuroscience, consumer behavior, and others, and uses scientific research methods to systematically understand, predict, and change human behavior.

It is an applied science. What is most exciting about it is that by understanding what influences behavior, we can design simple solutions that make a real difference to the organizations we work with and ultimately the people they serve.

Behavioral science touches nearly every part of insurance, from customer application design to in-force retention. In which areas could it have the greatest impact?

RGA has behavioral solutions for many aspects of the insurance journey, which all make an impact. One area I am particularly excited about at the moment is understanding the behavioral economics of lapses and improving customer retention. This is an area where we can introduce some simple ideas and make a real difference. For example, we helped one client redesign their quote renewal communications, which improved comprehension by 60% and policy renewal by 48%.

One way we think about our impact is likening our work to that of either a mechanic or mechanical engineer. A mechanic is responsible for fixing things that have already been designed and built, whereas a mechanical engineer is responsible for the initial design that determines how the product or structure functions. We are involved in the insurance version of both, but the more we can play the role of the engineer, the more we can ensure a realistic view of how customers will behave is baked into products, processes, and communications.

AI continues to expand its presence. What role do you see it playing with behavioral science?

This is an area we are researching at the moment. I think there are two key areas where AI and behavioral science meet.

First is how AI may impact key insurance behaviors. For example, we have studied customer disclosure behaviors closely over the past eight years. We have found that customers are more likely to disclose sensitive information via non-human channels. In a recent study, participants told us that they would be more comfortable disclosing mental health conditions to a chatbot than a human adviser. We think this is because of the strong emotions involved. But how AI-enabled insurance journeys may change customer behaviors is an important area for us to continue to research and understand.

Second, how can behavioral science help make our interactions with AI more efficient, effective, and ethical? For example, there are behavioral challenges that will need to be overcome to fully realize the potential of AI tools – such as adoption of the tools in the first place – ensuring that the right amount of trust is put in the outputs so that it leads to good decision making, and being aware of biases and inefficiencies (such as poor prompt writing) in how people use the tools that might lead to poor results or poor decisions.

Watch this space for our upcoming research on human-AI interaction in insurance.