Part 4: Why Mis-Spelling My Address Led to Higher Rates for Collision Coverage
Behaviour Today = Claims Tomorrow
In this article series, FTI Consulting’s Insurance team explore how new behavioural data is generated and early signs of how this data could be the next wave of predictive power for issues like fraud, and maybe even claims costs.
In our article last week, we showed how the use of a completely new type of data, digital behavioural data, can enhance the fraud models of insurers.
This is generally true of any new vector of data with little to any correlation to the conventional data used to build today’s models – and the more novel the data is, the more likely that the data will have predictive power in very unexpected places. The history of credit data predicting motor losses, telematics data predicting non-driving losses, and other novel datasets experience confirms this.
What is the future of digital behavioural data?
We have known from decades of academic research that driving in certain emotional states creates a propensity to dangerous driving and accidents1. Several investigations have shown that emotional state has a significant impact on driving ability, with poor driving more likely to occur at both extreme positive and negative moods2.
We also know that individuals retain habits which were formed at a young age. Normalised for age, income, education, and other factors, people who are more angry, distracted or excitable as young drivers are more likely to exhibit the same traits as older drivers. Therein lies the next frontier of using digital behavioural data.
Sign up for our webinar - Behavioural Analytics – The Next Frontier – on 23 September here.
1: Myounghoon Jeon (2016) Don’t Cry While You’re Driving: Sad Driving Is as Bad as Angry Driving, International Journal of Human–Computer Interaction, 32:10, 777-790, DOI: 10.1080/10447318.2016.1198524
2: Braun, Michael; Schubert, Jonas; Pfleging, Bastian; Alt, Florian. 2019. «Improving Driver Emotions with Affective Strategies.» Multimodal Technologies Interact. 3, no. 1: 21.