Pump the Breaks: Progressive Insurance using IoT to Snitch on your Bad Driving Habits
A few weeks ago, when I borrowed by sister’s car, she warned me to not make any short stops. Her insurance provider was tracking her driving, and frequently hitting the breaks too hard would disqualify her from potential premium discounts.
In 2015, Progressive Insurance, the fourth-largest U.S. auto insurer, rolled out a new voluntary “Pay as You Drive” program which gives drivers the chance to lower their premiums based on analysis of their driving habits. Drivers who opt-in to the program receive a device to plug into their car. Progressive monitors driving and assigns customized insurance rates based on driving behavior. Safe drivers will receive a discount, while more reckless drivers could see an increase in their premium. Of those who choose to plug the device, Progressive estimates that “nearly eight out of 10 drivers save” with average saving of 10-15%.1
The Progressive Snapshot device is a small module that is plugged into your car’s On-Board Diagnostics Type 2 (OBD-II) port. Like a black box on an airplane, the OBD-II allows automotive technicians to diagnose vehicle activities. Inside this 2”x3” device is a memory chip, cellular radio, and GPS radio that can track location. The Snapshot continuously streams data wirelessly, via AT&T’s network, to Progressive while you drive. Progressive records the number of miles driven, time of day, and how often the driver makes sudden stops.
While car insurance requirements vary by state, drivers must always prove ability to pay damages in the event of an accident. The most common type of insurance is liability insurance, which covers damages when you are at fault.2 Providers calculate a driver’s likelihood of making a claim and base the premium on the perceived level of risk. Providers historically bucketed drivers in broad actuarial categories to supplement limited driving records. Certain factors statistically increase your chances of getting into an accident, and providers use these factors to determine your premiums.2 Traditional factors include age, gender, occupation, vehicle’s make and model, and credit history. The Internet of Things (IoT) has created new opportunities for providers to better estimate risk based on driving and price accordingly. Because good drivers will shop around for the lowest rate, this program results in increased premiums (by attracting more drivers), reduced loss ratios (by attracting better drivers) and improved margins. The program has also encouraged safer driving behavior.
Despite the potential for savings, only 25% of customers have opted-in. Some customers choose not to participate because of the lack of privacy. Progressive discloses that they currently track vehicle speed, but not location. Smartphones have made Americans more comfortable with location-tracking, so the company “is once again researching GPS use, including in some of its Snapshot devices.”3 Other customers opt out because it gives Progressive the ability to use data against them in a future claim. If a driver was speeding at the time of an accident, the data could adversely impact his claim or future premiums. Additionally, security remains a big concern for IoT-enabled devices.
Turning this data into usable insights will require larger and more sophisticated IT departments, investments in data warehousing, and data analytics teams to interpret the data. Going forward, Progressive will need to think about what additional data to gather and how to best use it. More detailed monitoring may be imminent. Allstate was awarded a patent last year for cameras to record “potential sources of driver distraction within the vehicle” such as pets and phone usage, as well as sensors for alcohol and loud noises.3 Overlaying GPS location tracking with a map of speed limits would allow providers to know when the driver is speeding. While this data could help Progressive make premium adjustments, should it be used by authorities to issue traffic tickets? Currently, Progressive will turn over data with court orders.
I believe insurance providers will eventually require all drivers to participate in usage-based insurance. Progressive partnered with General Motors to install a built-in device in all new vehicles. Customers can decide if they want to share the data with Progressive. A mobile version of the Snapshot system is also in the works. This raises an important question of who owns the data. If it is the driver, they should be able to “shop” their scores to other providers, which could threaten Progressive’s competitive advantage built by offering the Snapshot program. Progressive is not the only auto insurer using IoT; Allstate has developed a smartphone application to monitor driving habits, and Nationwide has a similar program.
IoT has helped Progressive improve its operating model, but a larger threat to its business model is on the horizon. The impending reality of driverless cars may make auto insurance providers obsolete, as individual drivers will no longer require coverage. Progressive will need to evaluate how it pivots from personal car insurance to product-liability and coverage bought by ride-sharing businesses.
As for my sister, she is removing the device. Her premium savings amounted to $0.
 Progressive News Release. https://www.progressive.com/newsroom/article/2015/may/safer-drivers-pay-less-for-car-insurance-with-snapshot/
 Department of Motor Vehicles. “How Car Insurance Works,” http://www.dmv.org/insurance/how-car-insurance-works.php
 Scism, Leslie. “Car Insurers Find Tracking Devices Are a Tough Sell” Jan. 10, 2016,
 Progressive News Release. https://www.progressive.com/newsroom/article/2015/may/safer-drivers-pay-less-for-car-insurance-with-snapshot/
 Meola, Andrew. “How the Internet of Things will affect security & privacy” Aug. 24, 2016, http://www.businessinsider.com/internet-of-things-security-privacy-2016-8
 O’Connor, Amy. “GM’s OnStar Ready to Offer Progressive’s Usage-Based Insurance” February 4, 2015, http://www.insurancejournal.com/news/national/2015/02/04/356387.htm
 Reiss, Robert. “5 Ways The IoT Will Transform The Insurance Industry,” http://www.forbes.com/sites/robertreiss/2016/02/01/5-ways-the-iot-will-transform-the-insurance-industry/2/#39de42dd7d9c
Student comments on Pump the Breaks: Progressive Insurance using IoT to Snitch on your Bad Driving Habits
Alicia, this is a very interesting post!
You mentioned that Progressive estimates that nearly 80% of the drivers save on average 10-15%. However, it seems odd that just 25% of the customers have opted-in. I believe that, with the increasing use of IoT and personal data gathered by large corporations, people are more and more concerned with the use of their data, and thus regulators are reacting accordingly. Reforms to data protection rules have recently taken place in the EU (http://ec.europa.eu/justice/data-protection/reform/index_en.htm) and I’m sure this is also the case for the U.S. Hence, I disagree with your statement about insurance providers eventually requiring all drivers to participate in usage-based insurance. For this to happen, I believe many regulations should be changed, which seems highly difficult in this moment of data sensitivity. Many people spend several hours a week driving their cars and I’m sure they don’t want to be constantly monitored. Neither does your sister nor do I 😉
Very interesting and thought provoking post, Alicia, especially in light of the cases this week and the concept of dynamic pricing! I think Cristina raises a very important point about privacy. I am wondering how you view the tradeoff between less privacy and more efficient pricing (ie. matching premiums to better reflect the actual risk). Furthermore, given that society and individuals will all likely have different feelings on the privacy issue, how do you think consumers who value privacy will react to what will most certainly feel like a penalty to them (since they won’t be able to benefit from the discounted rates because they don’t want to share their data)?
For me personally, I am very in favor of this for a few reasons. Firstly, I don’t think good drivers should have to subsidize worse drivers. Driving behaviors can be improved and if consumers want to pay less, they should work to be safer drivers (it is not the same thing as pre-existing conditions in healthcare, for example). Moreover, by incentivizing better driving through discounts, the insurance companies are encouraging safer driving for everyone, which is a huge net benefit to society.
I think where is breaks down (or at least, not perfect yet) is in some of the ways Progressive defines good drivers. First, lower mileage means less risk of accidents, but this penalizes those who have to drive further to work. Second, driving at night makes you a riskier driver. Therefore, someone who is a good driver but has to commute far for a night shift might be hesitant to sign up for Snapshot. Ultimately, as the “safe” drivers sign up and get the premium reductions, those who are bad drivers or don’t sign up will have to pick up the excess premiums. Currently, the only measure of driving safety is hard breaking. However, Progressive and other insurers are looking at additional monitoring. But it gets back to the tradeoff that you and Christina mentioned- how much do we want our insurers to know?
Thanks for sharing this post, Alicia! I encourage you to check out my post about Lytx – which is a company that has also created driver-monitoring technologies, principally for the commercial trucking industry (see here: https://d3.harvard.edu/platform-rctom/submission/lytx-l%c9%aat%c9%aaks-n-1-a-company-saving-lives-on-the-road-digitally/)
I’m curious about your (and your sister’s) experiences with the Progressive Snapshot. Did you find that the “buzz” when you had to brake was useful feedback? Did Progressive give you any kind of follow-up or “report card” about your driving?
Lytx’s product is a two-way video recorder that gets mounted on the windshield of a truck. Unlike the OBD-II products used by the P&C insurers, the Lytx service links into an accelerometer, all of the other sensors on the truck, and also records video. All of these sources of information give their algorithms the ability to determine more about a driver’s behavior (are they actually accelerating too quickly? or is the accelerometer picking up a false signal? etc.) and the driver’s context on the road (braking quickly is a good thing if there is a pedestrian that enters a cross-walk!). The Company uses all of this information to identify risky driving — but also to provide coaching and feedback to drivers. If everyone on the road is getting better at driving, then perhaps we can reduce the number of accidents, injuries, and fatalities. There were over 35,000 motor vehicle fatalities in the U.S. last year!
I wonder whether Progressive – and other insurers – can do more than just use accelerometer data to set an insurance “discount.” To me, there’s an imperative on those collecting driving behavior data to also provide ongoing coaching and feedback.
Yes, the insurers do provide a “scorecard” of sorts, and let you check online how many hard breaks you have racked up. I’m not sure what safe driving coaching they provide but I would expect them to have resources dedicated on their website. Personally, we both found the buzz extremely frustrating, especially in contexts like you mentioned. Driving in the city, you have a lot more occasions where a hard break could be a behavior that is brought on because you are a good driver. I know several providers are moving towards more comprehensive monitoring to get a better picture of what is going on.
Alicia, thanks for sharing! I think the issue of paying a “privacy premium” is a very important one. As many good drivers will “take the bait” and submit to monitoring, it will gradually make it more expensive for the average consumer to maintain the status quo.
More important, however, is your point on Progressive’s need to adjust to a changing insurance terrain. I applaud Progressive for flexing its innovation muscles to challenge the traditional insurer model. As the ratio of driverless cars on the road increases, all drivers will inevitably face increasing premiums to maintain privacy and control over their cars. What is uncertain is what role Progressive and other insurance companies will play here. What do you think the most successful insurance companies will do to adapt? You mentioned in your article that insurance companies should pivot to become product insurers rather than driver insurers – do you think the market for this will be sufficiently large, given the high bar that driverless cars will have to clear to be put on the road?
I agree there is a very high bar to clear for driverless cars, but I think it is inevitable. Perhaps it takes 10+ years, but it is a very real threat to auto insurers’ business model. Perhaps best positioned will be insurers who sell property and liability policies to automakers and suppliers of the equipment. “Deloitte forecasts approximately $200 billion in personal-car-insurance premiums to hold steady for seven or eight years, then slide to about $40 billion by 2040. It projects about $100 billion of this $200 billion could migrate to product-liability insurance and coverage bought by ride-sharing businesses.”
Thank you for this interesting post, Alicia! Having followed the insurance industry closely for the last few years, it has amazed me how many fantastic opportunities big data will bring to this industry. On the property/casualty side of the business, insurance companies could obtains much cleaner data sets of customer behavior which would immensely benefit their actuaries in correctly pricing individual risks, especially if they are behavior related, such as car insurance. On the life/health insurance side, insurance companies could help customers adopt healthier lifestyles and thus save costs in the long run. Generali, an Italian insurer, has recently launched a health insurance product (‘Generali Vitality’) which lets customers opt into a program which makes recommendations regarding a healthy lifestyle. If customers follow the advice, they can earn rewards and thus lower their health insurance costs. It remains to be seen, however, how much customers will end up valuing product components, such as privacy or simply the ability to make their own decisions, over cost savings. How much are we willing to sacrifice in terms of lifestyle in order to save money on health insurance?