Data Augmentation for Risk Assessment in the Auto Industry

adult-automotive-blur-13861 (1)
       Over the past 3 years, auto insurance companies introduced a new way to charge drivers by tracking their driving data rather than using traditional statistics as driving history, instruction, location and age. Instead of having the users filling out the online form and providing a quote, the insurance company collects data on the drivers and come up with a quote based on that actual recent driving history.

 A few examples of insurance companies that are providing this offering to their customers are:

          Progressive offers its customers discounts based on their driving through its “Snapshot” program. From the Progressive website: “Snapshot is a program that personalizes your rate based on your actual driving. It's technically called usage-based insurance. That means you pay based on how and how much you drive instead of just traditional factors… your rate may increase with high-risk driving. But you're in control of what you pay for car insurance, and most drivers earn a discount. In fact, Snapshot rewards the average driver with a $130 discount.”

         Root Insurance has a similar approach. The drivers can download their app onto the smartphone, turn on the location tracking and upload a picture of their license. Nonetheless, they don’t need to log trips or open the app while driving.  In fact, the app runs quietly in the background, passively collecting data about the customer’s driving skills.

        Unlike Progressive, Root doesn’t offer discounts over time. Instead, after a two-week initial trial period, it provides an insurance premium calculated based on that trial period, and then the user no longer needs to keep the app running (even though to build a profile of the user, the app continuously sits in the background collecting data).

        In May 2019, Tesla CEO, Elon Musk, announced that Tesla’s intention to enter a new industry: auto insurance. One of the largest complaints of many Tesla owners is the high cost of insurance. Even if not in activity yet, Tesla Insurance Services, Inc has been incorporated since June 2017, showing that Tesla was planning this move for some time.

       Thanks to the innovative and data-driven nature of the automotive company, Tesla has direct access to driver’s data and therefore it can evaluate the risk profile.

       If a customer wants to purchase a Tesla insurance, the cost that they will end up paying will depend on their driving habits (reckless driving habits will result in higher insurance rates).

Adoption of risk-based assessment programs

      According to Root Insurance, its customers can get rates up to 52% lower than their previous insurers, averaging savings of $1,187 per year on car insurance policies compared to their previous rates with other providers (this average includes policies with more than one driver and a younger user base that tends to be charged higher rates).

      Progressive’s Snapshot program has handed out over $700 million in discounts, according to the company, with individual drivers saving an average of $130 a year.

       AAA is also offering this type of program, reporting that 27% of AAA policyholders are using the program in the states where it’s currently offered, with a 4% month-over-month participation growth.

What do the customers think?

       In 2016, the Pew Research Center conducted a study on how Americans approach privacy, asking if they were willing to allow insurers monitoring their driving habits and location, in exchange for a discount. Many Americans deemed the tradeoff unacceptable (45%), only 37% of respondents would be comfortable with it. The remaining said it would depend on the circumstances.

      The persistent data monitoring of Root Insurance has led to some complaints. In fact, users noticed that the app was still recording data when:

  1. Users run on a treadmill
  2. The airplane is taxiing before takeoff and after landing
  3. The users ride in an Uber / Lyft
  4. The users are in the passenger seat when someone else is driving

     Nonetheless, these are instances that should be an easy fix for the company, and we are positive that, in the future, we will see much more of this data augmentation for risk-based assessments in the auto insurance industry.


 READ MORE: User Research Methods for Automotive User Experience, Internet of Things: Is This Big Data?, Best Mobile UX Experiences in Auto Insurance, What is Datafication?

More by this Author

Comments

Add Comment