Tuesday, August 24, 2004
Drive less, pay less
So goes the slogan from Progressive's new auto insurance program called TripSense(SM).
Here's how the program, currently in pilot test in Minnesota, works: drivers are asked to plug a small device into their car that records when they drive, how far and how fast. At the end of six months, they can upload the data that's been collected to Progressive and see if they qualify for a usage discount on their next premium. Likely discounts range from 5 - 25%. If drivers don't want to, they don't have to upload anything.
The company has been trying for years to figure out a way to predict risk (and therefore calculate rates) better. From 1998 to 2001, Progressive tried to make a GPS based monitoring system work, but as cell networks meeting E-911 requirements are learning now, implementing a solution with comprehensive coverage isn't easy.
There is a lot of potential good news in what Progressive has going this time. First, the where, when and how fast data is logistical information directly relevant to risk. Unlike zip code profiling (adjusting your rates based on the number of claims from other drivers in your area - and potentially unfair to low risk drivers in high risk neighborhoods), this data takes individual driving needs into account. Second, drivers can see all the data that is collected AND control whether or not the company sees it. If a driver doesn't want the monitoring or doesn't think the results do him any good, he doesn't have to submit data. Of course, if a driver really doesn't want monitoring, he can go to another carrier. Progressive realizes that and sensitivity to the feeling of intrusion has been obviously taken into account.
One note of caution in this otherwise rosy consumer-infotech tale:
according to their press release, "TripSensor also collects information about rapid acceleration and braking that is not currently used to calculate the discount." In other words, Progressive wants to know if there is a correlation between how often you hit the gas or brake and the likelihood you'll be in an accident. Profiling driving behavior like this may get a bit dicey and, as with all information collection that offers a benefit, we would do well to watch to make sure that the fine risk/reward balance struck by Progressive today stays in place tomorrow.
Here's how the program, currently in pilot test in Minnesota, works: drivers are asked to plug a small device into their car that records when they drive, how far and how fast. At the end of six months, they can upload the data that's been collected to Progressive and see if they qualify for a usage discount on their next premium. Likely discounts range from 5 - 25%. If drivers don't want to, they don't have to upload anything.
The company has been trying for years to figure out a way to predict risk (and therefore calculate rates) better. From 1998 to 2001, Progressive tried to make a GPS based monitoring system work, but as cell networks meeting E-911 requirements are learning now, implementing a solution with comprehensive coverage isn't easy.
There is a lot of potential good news in what Progressive has going this time. First, the where, when and how fast data is logistical information directly relevant to risk. Unlike zip code profiling (adjusting your rates based on the number of claims from other drivers in your area - and potentially unfair to low risk drivers in high risk neighborhoods), this data takes individual driving needs into account. Second, drivers can see all the data that is collected AND control whether or not the company sees it. If a driver doesn't want the monitoring or doesn't think the results do him any good, he doesn't have to submit data. Of course, if a driver really doesn't want monitoring, he can go to another carrier. Progressive realizes that and sensitivity to the feeling of intrusion has been obviously taken into account.
One note of caution in this otherwise rosy consumer-infotech tale:
according to their press release, "TripSensor also collects information about rapid acceleration and braking that is not currently used to calculate the discount." In other words, Progressive wants to know if there is a correlation between how often you hit the gas or brake and the likelihood you'll be in an accident. Profiling driving behavior like this may get a bit dicey and, as with all information collection that offers a benefit, we would do well to watch to make sure that the fine risk/reward balance struck by Progressive today stays in place tomorrow.
# posted by AmandaWelsh @ 11:15 AM 
