It sounds like a middle-school math problem, but as we continue to refine our telematics program, it could become one more way we help members drive more safely – and save money – with the USAA SafePilot™ app.
Brian Y., USAA's auto product development director, says the launch of SafePilot 2.0 this year has allowed us to capture more data than ever. The program's expansion took the SafePilot app from just four states to 38 and has already reached billions of miles collected.
“We're starting to see how people are actually driving," Brian says. “The next piece is to then start associating different behaviors to the data we have."
The data the team has so far allows them insight into members' driving habits, such as miles driven and which days of the week they drive most often. For example, some of the data they have received shows phone use declined over 10% from 2020 to 2021, while speeding increased more than 5%.“It's data we've never been able to get before," Brian says. “We just need to decide how to use it."
The SafePilot app uses three indicators to determine a driver's score: distracted driving, harsh braking and annualized hours driven.
“Your score may increase or decrease because of the amount you're driving – even if you don't have any infractions," says Connor H., USAA's P&C product management, assistant manager.
Using math much harder than the middle school example above, the back end of the app weighs the three indicators, but because distracted driving takes many forms, a driver's score would be affected differently by texting versus taking a phone call. For example, says Connor, handheld phone calls result in a score about one-third lower than hands free phone calls.
However, the goal isn't for our members to be perfect drivers, it's to take simple steps to improve driving habits.
“It's not all or nothing," says Connor. “We're not expecting the average driver to be perfect. We're aiming for safer drivers."
The duo also has an answer for those who argue that they've been dinged for allowing another driver in the car to use their device.
Future innovations could mean the app could connect with your car's systems to check whether your passenger side airbag has been disabled, which indicates no one is in the seat. If the airbag is not disabled, the app could conclude someone in the passengers' seat is actually using the phone.
“We've not gotten to that level of data, but that's an example of where it could go," Brian says.
Partners on the Advanced Analytics and Machine Learning team are working to take our current data and add in contextual data, such as local speed limits, road types, weather data and more, which could provide more personalized auto insurance policy rates.
“Understanding the context of your driving behavior and what that might mean to your insurance – that's what we are working toward," Brian says.