The future of usage-based insurance (UBI) will look nothing like its past. Soon, the notion of a “plug in” device for a proprietary, carrier-based program—such as Progressive’s Snapshot or Allstate’s Drivewise—may seem as quaint as a flip phone. In fact dongles are already being replaced by smartphone UBI apps.
We are, in short, moving into an era of democratization for UBI. “Driving data bureaus” (similar to credit bureaus) are emerging. These bureaus will develop and sell predictive “portable driving scores” to insurers and others with permissible purpose. Vehicle owners will have control over the capture and sharing of their data, along with review and dispute ability.
This score will become a rating variable at both new business and renewal for the entire auto market. And that will give most insurers what they really want: the opportunity to NOT have a proprietary UBI program... and all the baggage that comes with it.
A hurdle too high
Today, proprietary UBI programs are economically viable only for the top 25 (or so) U.S. insurers. For the rest, developing and administering such programs has proven to be an economic hurdle too high. The amount of driving behavior data and corresponding loss history required for predictive modeling and regulator-controlled rating are also big burdens for insurers and telematics service providers.
This is not to say that insurers outside of the top 25 don’t believe in the predictive power of driving data. They may even hate the way carriers such as Progressive and Allstate have disrupted the market. They just want to see the playing field leveled—just like it did when an industry-specific credit score emerged to combat pricing advantages enjoyed by the early users of credit.
The good news, at least for these insurers, is that the proprietary UBI model is on the verge of giving way. Contextual driving behavior data will become available directly from the car. We will soon see the emergence of a new UBI model, leveraging a portable driving score (similar to an FCRA-governed credit score).
They’ll get the level playing field—and we’ll see the democratization of UBI.
A brief history of the auto insurance market
The first reported car accident in the U.S. occurred in 1891. I can just imagine the only (as in ONLY) two cars on the road running into each other. Actually, that first accident occurred when the car ran into a hitching post. No horses were injured.
There were 4,000 cars on U.S. roads when Travelers sold the first auto policy in 1898. Its insured paid $12.25 for $5,000 in liability coverage. In 1925, Connecticut became the first state to adopt a financial responsibility law where car owners had to prove they could pay for injuries or property damage they caused in a car accident.
Until 1957, Massachusetts was the only state that required auto insurance by law. New York followed in 1956, then North Carolina in 1957. Soon afterward, most states required auto insurance for all drivers. Not surprisingly, the first insurance television commercial aired around that time.
Today’s auto insurance market is pretty efficient. Comparison rates are easy to get. Shoppers shop and tend to stay shoppers. Shopping events are frequent: buying a car, adding a driver, moving, getting married, having an accident, getting the renewal bill, and maybe even seeing a commercial. Incredibly, more than $6B will be spent on advertising by car insurers in 2017.
Pricing as a competitive advantage
Price has always been important to consumers, while pricing has always been important to insurers. And, in a regulated industry—one in which filings can be copied—sophisticated pricing can be a competitive advantage.
In 2004, Paul Mang (then at McKinsey) and Keith Toney (then at InsurQuote) introduced four key pricing dimensions—granularity, dispersion, interactions, and variables—that lead to improved accuracy. Perhaps no single pricing breakthrough did more to confirm their analysis than the introduction of credit. The “haves” and “have-nots” became obvious to shoppers, and adverse selection became rampant. It took almost 10 years for the market to “recover” and for credit to become commonplace.
Auto losses soared by 13% in 2016 due to unfavorable claims experiences, leading insurers to broadly pursue rate increases. And still private auto incurred losses will increase by nearly 7% in 2017 to a new record of approximately $154 billion.
Best’s Special Report, “Despite Top Line Growth, U.S. Nonstandard Auto Results Continue to Deteriorate,” notes that small, regional insurers are hit particularly hard where competition from large, standard, auto writers is constant. The report’s conclusion: insurers without pricing sophistication are going to have a tough uphill battle.
That brings us back to UBI.
Proprietary UBI isn’t for everyone
Let’s be clear, today’s proprietary UBI programs are all about segmentation resulting from sophisticated pricing. The business case only works if the program attracts more new customers who have fewer crashes, and only if those customers renew.
Yes… despite the hype about engagement, coaching, and changing behaviors, proprietary UBI programs are not about creating safer drivers. They’re about finding drivers who are already safer, making them feel as if they’re paying too much, and convincing them into switch.
As stated above, these expensive programs are fairly exclusive to the top insurers. And, penetration is fairly low. Of the 215 million personal auto policies in force in the U.S., fewer than 18 million are UBI policies.
Today’s insurance telematics ecosystem is a function of the relative inaccessibility of driving behavior data, and the reluctance of insurers to get into the hardware business. That’s changing.
The future will be all about the data
The change starts with OEMs, who have all come to see themselves as mobility providers. By 2020, 250 million connected cars will be on the road globally. These cars will generate information on driving behavior, vehicle health, and infotainment.
For the OEMs, mobility is a gold mine—an opportunity to sell data services not only to insurers (for pricing and first notice of loss), but also to dealers, municipalities, road-user charging authorities, retailers, and emergency responders.
At the same time, we’re seeing the emergence of driving data bureaus. In many ways, they’ll work with driving data similarly to how credit bureaus now work with consumer credit data. They’ll collect, normalize, anonymize (when necessary), and model data from OEMs and other third parties. When regulated, data will be FCRA-compliant. When advantageous, models will be filed with insurance regulators. APIs will be created to allow for transactional implementation.
The bureaus will battle to show their driving score is most predictive. All the while, consumers—vehicle owners—will have control of their own portable driving score.
Consumers will be the big winners
For OEMs and data bureaus, success will depend on providing value, choice, and transparency, especially to consumers. For the underpenetrated UBI market, success will depend on scale (having enough data) and ease (the ability for insurers to connect).
As a portable driving score arrives on the scene, insurers outside of the top 25 will get the level playing field they’ve pined for. And, UBI will no longer be the province of just the country’s top insurers.