UBI and driver behaviour: what really works?

Data results show driver feedback within UBI programmes do result in reductions in the behaviours used to assess risk yet there remains of plenty of room to grow, not only in market penetration but also in insight. There’s no one programme that makes sense for all drivers or all insurers and there may be unintended consequences to the feedback.

Why insurers need to get this right

Here’s why this matters so much: insurers are behind the distraction curve. Insurers and their UBI vendors measure speeding, hard braking, cornering and other attributes that are indicators of risky driving but “the industry doesn’t know how to measure distraction,” says Ted Gramer, CEO of TrueMotion, nor the losses resulting from it.

At the same time, traffic fatalities increased from 2015 to 2016, according to the US National Highway Traffic Safety Administration (NHTSA) with at least 10% caused by distraction. Auto insurance claims are also increasing. In just one example, State Farm lost $2.5Bn (£1.85) in 2016, following a loss of $2.1Bn (£1.55Bn) in 2015. Says Gramer, a former Liberty Mutual executive: “Auto insurers wrote policies in 2015 for 2016 and 2017 based on estimated losses that tended to be wrong by a lot. They didn’t have the data that people driving with phones in their faces would cause claims to be some10 percent higher than they expected.” Insurers can’t wait until those policies come up for renewal. Their best bet is to change the risky habits of their current customers.

Promising results

TrueMotion provides a mobile phone app that uses smartphone sensors to provide drivers with overall scores about how much time they spent while driving distracted by phone use. It analysed more than four-and-a-half million trips and found that drivers who were made aware of their risky driving habits reduced them by 20%.

Cambridge Mobile Telematics (CMT) uses phone sensors and an optional wireless tag to measure driving performance and vehicle dynamics. CMT says users of its DriveWell product sees an average reduction of 35% in phone distraction, 20% in hard braking and 20% in at-risk speeding within 30 days of using it.

If 20% reductions don’t seem significant, Hari Balakrishnan, CMT’s founder and CTO, says: “Think about the numbers over a one-year timeframe and you can see it’s significant.”

Octo Telematics’ analysis of one insurer customer’s data found that after four weeks of driving with a telematics device, 55% of drivers were improving on the measures it tracks, including time of day of trips and location of trips, as well as hard braking, etc. After eight weeks, 70% of drivers had improved, according to Nino Tarantino, CEO of Octo North America.

Among insurance customers using The Floow’s FloowCoach system, those who scored below what would be expected of a safe driver during their first 100 miles improved their scores in the last 100 miles of monitoring by 57% overall. The Floow focuses on the bottom 10%, the riskiest based on claims data according to Lisa Dorn, chief education officer. For every 100 policyholders in the worst decile who complete the FloowCoach programme, 18 claims are avoided.

What really matters?

Avoiding claims, instead of hard braking or whatever, should be the goal and measure of a good driver coaching programme, says Geoff Werner, global telematics leader at Willis Towers Watson. Some feedback may simply encourage drivers to swap one risky move for another. For example, one way to avoid hard braking might be to sail through yellow lights instead of stopping for them.

“For the most part, harsh braking is not the thing that causes an accident to happen,” Werner says. “Instead, it’s something a driver does to avoid getting into the accident. The behaviour we need to change is really whatever puts the driver in a situation where they need to avoid that accident.”

Werner sees three things as necessary for effective driver coaching. First is correctly identifying the things, behaviours or decisions that actually cause accidents to happen and the behaviours that should be changed. Second is figuring out the best way to deliver information and/or coaching to the driver. Third is a mix of incentives that will make drivers want to change their behaviours. “It needs to be appropriate to that person,” he adds.

Art of the nudge

The feedback provided by Octo’s systems, both mobile app- and dongle-based, is customised for each insurer. The reward is typically a discount on the premium. “It’s a better sell to their customers,” Tarantino says. “They can just say, ‘We charge you this because you’re risky. If you want to pay, less drive better’.”

Drivers will change substantially for a very small reward and it doesn’t need to be financial. CMT has found that safe-driving contests have strong results even when the total prizes are modest, according to Balakrishnan. “Somewhat surprisingly, we have found that even after financial rewards are removed, many users improve,” he says. He attributes this to a combination of engaging feedback and tips, coaching at the moment of an error and leader boards that let drivers compete with others.

The Floow has found that coaching itself can be rewarding. The company takes a very personal, hands-on approach. Instead of a mobile app or other technology, it uses human coaches who engage with drivers in regular phone calls, although its pilot Reward Services App includes tangible rewards such as Amazon gift cards or a reduction in annual premium.

The programme is initiated with a personal phone call to the driver, Dorn says. This call assesses the person’s openness to coaching. Using methods that have been successful in other areas of health psychology, further coaching conversations aims to get drivers to understand the triggers for speeding or other behaviours. Next, they’ll set goals for improvement.

Dorn explains: “The main reason why I think people engage with the program is because of our coaches. These are very engaging personal conversations with the drivers. They feel cared about, like insurance company is interested in them.”

Rewards need to get less vanilla, TrueMotion’s Gramer says. While there’s upside in improving 75% of drivers by 20%, he believes that fine-tuning incentives will lead to much more improvement. Says Werner: “We need to make sure that what we’re telling people to do are things that truly will drive accidents out of the system. If we can change the right behaviours, it will make this so much more profitable and societally beneficial.”

[Ins.Kuchinskas.2017.08.22]

Connected Vehicles