By Lakshmi Shalini, VP of Risk and Insurance Analytics for CMT

This article first appeared on Carrier Management.

Insurance pricing is an imperfect process.

It’s been gradually refined over countless iterations of product revisions through hundreds of thousands of regulatory filings. Insurers have used different methodologies, data sources and strategies to help them understand how to establish pricing mechanisms that account for variability in the marketplace.

Today, more than ever, major changes in mobility, the supply chain, labor and more are driving variability in the market. Insurers have to react to these changes rapidly, ensuring they can match rate to risk. The insurers who understand these trends, and can react accordingly, have the best opportunity to build a competitive advantage over time.

This is good for consumers. When insurers compete to drive growth, they create a wide variety of prices and coverage ranges. The core element driving these broad shopping ranges is the variability in the data. The more data an insurer has, the better they can structure their pricing to account for the level of risk they are willing to take and increase their coverage. The more imperfect that data is, the more the insurer needs to account for this variability. This can mean the insurer will miss out on major segments of the market because their pricing is not fine-tuned enough to attract those drivers.

So, the more granular an insurer’s data, the more coverage an insurer can provide.

Telematics has been one of the key insurance investments that help identify the core mechanisms that underpin crash risk. The first signal that insurers used en masse was hard braking. Hard braking events turn out to be highly predictive of crash risk when viewed in aggregate. There are many reasons a driver will hard brake. For example, a driver could hard brake because they were following another driver too closely and had to execute an evasive maneuver. Or a driver could hard brake in response to another driver violating the rules of the road. When viewed in aggregate, the insurer has the ability to understand risk more specifically. When an insurer includes claims data, these signals become highly predictive of understanding what leads to crashes and claims.

What’s highly interesting now is understanding the “why” behind the “what.”

In this case, the “what” is a hard braking event measured by a highly sensitive accelerometer. The “why” is much harder, but we can begin to fill the gap by looking at context much more specifically.

Advanced risk variables help us understand this context.

For example, by aggregating telematics speed data across a diverse and large set of users, you can more directly measure contextual speed on a specific road segment. Think of a driver on I-95 just north of New York City. We can measure this driver’s speed and compare it to other drivers’ speeds on this specific stretch of I-95. Understanding the driver’s speed relative to other drivers’ speed on the same road segment gives us more context.

Deriving insights from aggregated data is not a novel concept for insurance carriers. Insurance carriers have used advanced modeling techniques to aggregate demographic or loss data by ZIP codes for granular risk segmentation.

Another example of explaining the “why” behind the “what” of a hard brake is advanced distraction variables. Part of what makes distraction so dangerous is “distraction hangover.” Research from the University of Utah found that it takes your brain 27 seconds to fully refocus on the road after you interact with your phone.

In fact, CMT research shows the 10 percent most distracted drivers are 2.2-times more likely to crash than the 10 percent least distracted drivers. An advanced distraction variable like context switching identifies when a driver was distracted before a hard brake.

These additional levels of granularity for driving behaviors are providing advantages for segmentation lift. One area of new advanced risk variables, context-based risk, has been shown to generate an incremental lift of over 15 percent. As more granular data becomes available with new telematics variables, we will learn more about their segmentation power.

Today, most top carriers in the U.S. have a telematics program that utilizes variables like harsh braking and phone distraction to assess a driver’s risk and develop customized pricing algorithms. They’ve been able to match rate to risk and provide safe drivers with better pricing options.

As the evolution and investment in telematics expands, and pricing models become more sophisticated with the addition of context, insurers will be better positioned to respond to rapid changes in the market and shift strategies based on longer-term trends. Consumers will benefit from expanded shopping options and risk assessments based on contextual data.

Understanding the “why” with more context behind the “what” of driving behaviors—provided by new advanced risk variables—will be a win for insurers and consumers alike.