To determine policy premiums, auto insurers have traditionally used factors such as driving record, vehicle usage and insurance history. This is in addition to outside variables like customer location (collision/theft rates in the area), age and gender (teen males tend to have higher rates), and non-driving elements like credit score.
Recently, however, some insurance companies have begun to harness the internet of things to monitor real-time driving habits. By employing telematics devices, which plug in to the vehicle’s on-board diagnostic port, insurers are provided with analytics about a policyholder’s specific behind-the-wheel behaviors. So, instead of paying for collision coverage even when the vehicle is parked in the garage, customers would only pay for it when the car’s being driven.
This is called usage-based insurance, and Business Insider Intelligence estimates that over 50 million U.S. drivers will have tried UBI by 2020. By leveraging such refined data, premiums can be closely tailored around a driver’s level of risk. For safe drivers, especially, the prediction is that car insurance rates will be lowered.
But what we’ve seen with telematics is only the beginning. According to an SMA research survey, 74% of insurance executives believe IoT will disrupt the industry by 2020. It’s therefore no surprise that many insurers have begun to invest in IoT research and implementation. When autonomous cars hit our roads en masse, the business model and underwriting practices of auto insurance companies are likely to be changed forever.
Embracing IoT to evaluate claims
The National Insurance Crime Bureau reports that customers shell out an extra $200 to $300 per year on premiums to counteract the cost of fraudulent claims. While top-tier insurers do rely on rigorous investigative methods and analytics, the insurance industry still estimates that at least 10% of property-casualty claims could be fraudulent.
However, the UBI model already exhibits promise as a fraud-detection system. For instance, a common form of fraud is exaggerating vehicle damage after, say, a minor fender-bender in the hopes of getting a higher claim payout. But by cross-referencing that claim with real-world data showing the time, speed, location and position of the vehicles in an accident, insurers can better gauge how its severity stacks up against the nature of the claim.
And when vehicles are regularly communicating with each other and highway infrastructure, the rich tapestry of information therefrom can be leveraged for even more predictive fraud exposure analytics, thus, saving consumers hundreds of dollars every year.