SUBSCRIBE TO OUR BLOG
What is the fairest, most accurate way to decide on insurance premiums?
Traditionally, the car insurance industry has depended on using static information, including Motor Vehicle Records (MVR). MVR show that person's driving history, including any traffic violations or endorsements on their license.
They add this to other information including the driver's age, address, car, gender, former claims and driving experience.
But as with other sectors, car insurance is being disrupted – and location technology is at the heart of this. For example, Ometry uses data about where someone drives as opposed to where they live, to stop people in US cities with predominantly African American zip codes from paying up to 70% more than those in white neighborhoods.
Information about where you drive could be more useful than how you do it when it comes to insurance, according to kaskgo2go.
Many new models of auto insurance focus on using sensors to monitor driver behavior, with incentives and discounts for safer driving habits.
“We started out collecting raw data, and assuming some behaviors or parameters are effective for predicting the accident rate ratio for insurance companies," said kasko2go CTO Oleg Korol.
“We were trying to assess driver behavior on the road based on acceleration, speeding and braking data that we were collecting from the phone and other devices."
After an extensive study, the issue they found was that these parameters may not be useful to insurance companies at all. The amount of data they would need to extract to come to any meaningful conclusions is prohibitive.
Instead, they focus on the roads the driver is using. kasko2go uses HERE's accurate and detailed traffic data used to create highly accurate accident predictions of where accidents will be, based on the history of those areas.
“Let's say a driver is driving at 7am, as opposed to a driver driving through the same area at nine in the evening," Oleg explained. “Do they take the same amount of risk? And the answer is, in most cases, no. Because in the morning, the risk that the driver takes is much bigger than in the same area at night."
This map shows how risky roads in Zurich are at 8am. Image credit: kasko2go.
kasko2go now receives high-quality and accurate traffic data from over three million kilometers of road in the DACH region [Germany, Austria and Switzerland], updated every five minutes. This data shows the volume of cars passing through a specific road segment at any time of the day.
Combined with existing data sources, including accident data, HERE Real-Time Traffic now forms the cornerstone of kasko2go's entire system.
“It's more accurate," Oleg said. The company hopes this model will benefit insurance companies and their customers alike.
Instead of being driven purely by price, the industry can now depend on data for decision-making, and benefit from a lower loss risk.
That benefits drivers who can get more attractive premiums, and in general it can contribute to safer roads.
The same streets at 4pm look very different. Image credit: kasko2go.
At present, there is not enough data or evidence to show what impact driver behavior has on the risk of accident, Oleg said.
And in the future, the question may no longer even be relevant.
“Looking at future trends, would measuring driver behavior be necessary at all? Will you even have drivers to measure in the future?"
Before that happens, he said insurance companies have no choice but to adapt their business models if they want to remain profitable.
“A lot of insurance companies are losing money on car insurance products," Oleg said. “This kind of product gives the insurance companies a technological edge.
“They can reduce their loss ratio – not by increasing their premiums, but by correctly classifying risk."
Get highly accurate traffic data that decreases congestion, reduces accidents and improves navigation for drivers.