“Telematics is a fantastic advance!” says Towers Watson’s Global UBI Analytics Actuary Tony Lovick. Telematics data is revolutionizing the auto insurance industry by creating data driven risk and profitability models and rewarding drivers with better rates because of the data.
But let’s back up a little. Global insurance companies are facing challenges, from low interest rates and economic volatility to increased regulation and technological change. One of the largest actuarial firms in the world, Towers Watson, brings a combination of deep consulting expertise, innovative solutions and advanced financial modeling software to those challenges.
Now, onto the revolution.
Rather than using traditional data sources (age, gender, marital status, etc.) - this is all about sensor data and telematics. A simple explanation is drivers get a device installed on their vehicle, and that device sends sensor data (braking, acceleration, location, and time). That sensor data is combined with weather and traffic congestion data and then the real work begins.
“Using the claims as the response, we build an analytical model. We design a number of different risk factors from the data itself that we think will be predictive of risk, and then we objectively assign the claims as the response and build our models, and that comes up with the relative risk of each different part of driving. So we can divide it into different junctions, different road segments, different types of maneuvers, different situations on the road, and all of those are relevant to risk.” Tony Lovick, Towers Watson
So how does that revolutionize the auto insurance business? Towers Watson can now determine risk with pinpoint accuracy. In one use case they found the insurance carrier was charging inappropriately. Using traditional rating factors, one group of customers was charged $400. But, the top 10% of those drivers were only a $160 claims risk. The bottom 10% were a $1600 claims risk - a difference of 10X. A data driven business uses that information to make decisions about who they want to cover and how much to charge.
“So for the insurer, it's really important that they try and write those risks that are going to cost them 160 dollars' worth of claims-- they can offer a discount from the $400 they were charging before, share some of that benefit with the customer, but still maintain a profitable policy. And then there's the other customers at the other end. They're probably the ones that don't want to select this type of product, and they can find a 400-dollar premium from someone else.” Tony Lovick, Towers Watson
Different insurers are using the data different ways for their customers.
“There's a number of different telematics projects and different insurers in different countries around the world, and there's a number of different propositions which have been tried. So the Progressive proposition is all about price, ‘Come to us, record some data, get a discount.’ It's very simple. The State Farm proposition is all about value-added services, and they have a whole menu of different things that can assist the customer. In a different country, like Italy, for example, fraud is by far the most important thing for them. And so most of the propositions there are designed around recording crash data and then comparing that to the evidence from the drivers and trying to mitigate the chance of a fraudulent claim.” Tony Lovick, Towers Watson
No matter how they incentivize customers, it’s working - because business and data volumes are getting bigger and bigger. That’s why Towers Watson has invested in a robust architecture to analyze all the structured and unstructured data.
“We’re getting a million rows of data per customer per year so that’s billions of rows of data; that’s many orders of magnitude bigger than most data projects in the previous decade. We’re also recruiting a number of different insurers and we’re pooling their data and doing an analysis across all of those. We have to have that flexibility to be able to add new insurers to the project and therefore we need a robust environment that we know will perform as we add more and more clients on.” Tony Lovick, Towers Watson
What’s next for Towers Watson and the telematics technology for automobiles? Mobile apps for smartphones to give drivers the data, so they can make their own decision about safer driving and lower rates.
Thank you, Tony Lovick and Towers Watson for sharing this exciting use case for Teradata and telematics.