Editor's Note: This article is part of a larger feature piece on underwriting analytics; see
As telematics-powered usage-based insurance gains steam, underwriters are learning a lot about their customer base with new data that wasn’t previously available.
“We’ve done our own studies, and we’ve found, not surprising, that safe drivers were overpriced based on traditional rating variables,” says Dave Pratt, GM of usage-based insurance for Progressive Insurance. “Under traditional insurance models, pricing is based on proxies for risk. You hear people complain: insurance companies create rates; it’s a black box; it’s not fair; they have no control; they don’t understand how rates are done; they’re based on some finan ial-performance or age metrics. Yet usage-based insurance is the promise that, if you’re a safe driver, based on how you drive, you’re going to be treated more fairly and given a more accurate price. It’s very empowering as a consumer.”
One of the obstacles for insurers heading down the path of telematics is that the data is categorically different than other types of insurance data, Pratt explains. “You’re not talking about actuarial data; you’re talking about engineering data: things like accelerometer data and GPS data and speed. You’ve got these data sets coming in and they’re massive. And so they require new skill sets.”
Other challenges include the amount of noise in the data, which needs to be fil ered out to produce meaningful results, and regulatory. “Regulators love the fact that you’re measuring behavior as opposed to the other proxies in the market. At the same time, they don’t like ‘black boxes,’ so there’s a lot of explanation that goes out as to how the UBI program works. It can be complicated and it varies by state,” Pratt says.
On the upside, in addition to better and more accurate pricing, Pratt says the retention rate for Snapshot customers is higher because the insurer knows more about individual customers, which creates new opportunities to interact with them.
“In traditional insurance, you had a relationship when the insurance contract is signed, and when a claim happens. And you might have a relationship when the renewal comes up. That’s it in many cases.” However, vehicle and behavior data can create new opportunities to interact with customers beyond offering discounts, immediate negative feedback for hard braking, for example, to more pro-active alerts for such issues as alternator problems, which the company recently began detecting in the volumes of customer data.
