How mobility data helps reduce distracted driving
The automotive insurance industry is heading towards dramatic change. Telematics is enabling both increased accuracy of existing business as well as the transition to new forms of mobility. By strategically leveraging the right data, insurers can go on the offense and not be victims of it. Insurers must become increasingly responsive to the changing needs of their customers and create insurance solutions that take the changing landscape into account, enabling customers to experience mobility options seamlessly and efficiently.
At Arity, we believe that capitalizing on this opportunity today starts with the mobile device. It’s a catalyst for many of the changes happening in transportation – most notably the gig economy and the increase in distracted driving. But it’s also a relevant and reliable tool that can offer insights to help understand true driving risk and enable a better transportation system for everyone.
At the Auto Insurance Report National Conference last month, our President Gary Hallgren shared the importance of having an accurate understanding of distracted driving, and how telematics with associated loss data must be leveraged to truly account for and help reduce this risky behavior.
Didn’t get a chance to attend? No worries, we’ve packaged up some of our favorite knowledge nuggets from Gary’s presentation below. Check out these key takeaways:
Use the right data from the right sources to build the best models.
Early adopters of credit data to price insurance gained significant competitive advantages. Combining credit data with insurance claim data, they created very predictive models that allowed them to out segment the competition. Today, a similar opportunity is presenting itself for companies to crest the next wave of competitiveness: mobile telematics data. For the past decade or so, we’ve learned that using driving behavior data-based models are an even better predictor of loss, but only when the right data elements are applied correctly. So, what does that mean exactly?
Telematics data from onboard diagnostic (OBD-II) devices launched the revolution of usage-based insurance. This type of data has proven to be the source of truth for verifying mileage, detecting speeding, hard-braking, aggressive turning, and the time of day that a vehicle is driven. While this is useful, device-based data doesn’t provide the full picture of who is driving the vehicle and what they’re doing behind the wheel, preventing insurers from truly understanding how risky a driver is.
Data from consumers’ mobile devices capture meaningful driving behavior for each individual driver offering more predictive inputs. For example, mobile devices allow you to contextualize what’s happening during a trip and captures detailed behaviors like if the driver is sending texts at 65 mph or checking a navigation app at a red light. We’ve determined to incorporate these insights into models are more predictive than verified mileage or credit history, but only when matched with claims data. For any model to accurately predict losses, you need to know if an accident occurred, if a claim was filed and the cost of the claim.
Relying primarily on OBD data to build a mobile score can result in significant mispricing. Even good normalization can’t correct for differences. For a driving risk score to be most predictive, the modeled data should be based on the type of sensor in use. And finally, any model built on an analysis of driving behaviors must account for traditional rating factors that are already in play – such as age, geography, years of driving experience, and type of car – to avoid double counting the impact of these correlated variables. Too often we see insurers investing a lot to collect the right data, but end up pricing less accurately because the right measures to avoid double counting have not been taken.
It’s time to get smarter about distracted driving.
Distracted driving is a widely known issue, but just how much this behavior impacts an insurer’s bottom line has historically been unknown. At Arity, we continue to analyze distracted driving behaviors, implementing naturalistic test settings and leveraging video data as a reliable source of truth to increase our level of accuracy and acumen on this behavior to give insurers best chance at accounting for it accurately in their pricing.
It’s estimated that since 2011, distracted driving has cost insurers $9 billion and Arity research correlating mobile sensor data with associated insurance claims data determined the most distracted drivers have a loss cost more than 1.5 times higher that of the least distracted drivers, on average.
While we all agree that any phone usage or distraction when behind the wheel is dangerous, when trying to predict for insurance losses, our analysis shows that not all distracted driving behaviors are equally risky. For example, we can now distinguish opening GPS at a red light from texting while driving and texting while driving will likely yield a different insurance loss than checking directions at a red light.
And finally, mobile telematics pricing, especially that accounts for distracted driving, not only holds the promise of improving profitability, it presents an opportunity for insurers to help improve safety on the roads. The mobile phone is the one tool that not only captures this risky behavior but gives a direct feedback channel to empower consumers to control their pricing and develop safer driving habits for the benefit of themselves and everyone around them.
The data you need to activate change already exists.
So, if you’re serious about pricing more accurately and tackling distracted driving, then it’s time to get serious about incorporating the right data into your models. Adding the mobile phone to your toolkit is the key to determining a better prediction of insurance loss that you can act on strategically and smartly. The mobile phone is the first step to getting the driving and loss data necessary to more accurately account for distracted driving, while also enabling a more active role in coaching consumers to be safer drivers.
But accumulating and analyzing this level of high frequency, granular data is challenging and takes time. If you haven’t already, there are opportunities to quickly catch up and gain a complete understanding of the risk on your book. So where to start?
Over the past decade, Arity has made significant investments in accumulating data (as of May 1, nearly 130 billion miles of driving data and 16 million active connections) and data science expertise to help insurers with the heavy lifting. In addition to the insurance programs we’ve powered over the past decade – such as Drivewise and DriveSense – we’re partnering with third-party applications to leverage the driving behavior data your customers are already sharing to deliver insights and data models you can use immediately to account for distracted driving and loss.
Auto Insurance Report Conference was informative and offered a great experience to connect with others in the business about many things, including how to solve for Distracted Driving – a big passion for all of us at Arity. For an industry that may sometimes be seen as slow-moving, exciting changes are front and center for the automotive industry, and especially those focused on how insurers can grow successfully and leverage smart data to ensure that the future of transportation is safe and reliable.
Looking for a place to start on the road to understanding how to account for distracted driving and make transportation safer? Check out our blog on steps insurers can take to solve for distracted driving.