Pay as you drive, or pay how you drive?

Welcome to The TechCrunch Exchange, a weekly startups-and-markets newsletter. It’s inspired by the daily TechCrunch+ column where it gets its name. Want it in your inbox every Saturday? Sign up here.

Having talked to many insurtech investors lately, I found myself thinking about usage-based insurance (UBI, which in this case doesn’t refer to universal basic income). On a surface level, this approach makes a lot of sense: For instance, why should drivers pay the same premiums regardless of how many miles they drive? But differentiating users also raises all sorts of questions on what’s fair, and where UBI is heading next. — Anna

Stop paying for others?

“There has been a lot of noise around UBI […] over the past few years. It was supposed to be the next big thing, but it hasn’t really taken off yet,” New Alpha Asset Management associate Clarisse Lam told TechCrunch.

AV8 VC‘s partner Amir Kabir concurred with Lam, noting struggles among startups and legacy insurance providers alike: “Early startups operating the UBI space had a hard time creating meaningful moat,” he said. Meanwhile, he added, “incumbents have been operating in the UBI space for decades and have yet to see major adoption.”

Coincidentally, or perhaps not, one of the insurtechs that was most badly hit by the stock market sell-off was Metromile, which went public in 2021 and saw its valuation decline over 85% before getting acquired by fellow former startup Lemonade. Metromile’s focus was pay-per-mile car insurance, a self-explanatory concept in which drivers get charged less if they drive less.

In its SPAC deck, Metromile made the argument that “auto insurance is unfair to most consumers,” because “65% of drivers subsidize the other 35%” — those who drive more. According to the company, there’s a linear relationship between miles driven and expected losses.

Having enjoyed universal healthcare all my life, Metromile’s reference to unfairness snagged my attention: Isn’t pooling risks the whole point of insurance? And isn’t it inherently fairer for all consumers to pay the same amount into the same pot? Well, maybe not, if increased risks result from bad personal choices — a thorny question when applied to health issues, but a lot less controversial when it comes to driving. After all, nobody wants to subsidize reckless drivers.

Behind the wheel

Miles driven are undoubtedly a reasonably good proxy for automotive risk, but with technology and data, we can do a lot better, and UBI might not take off until it does. This explains why usage-based auto insurance is shifting from pay-as-you-drive (PAYD) to pay-how-you-drive (PHYD) — an acronymy way of saying that behavior matters more than miles.

In a recent blog post on the topic, Insurtech Gateway Australia‘s co-founder Nick Proud mentioned several factors that created tailwinds for UBI, “including the COVID-19 pandemic, technological advancements such as in telematics and smartphones, and growth in connected cars.”

Telematics refers to vehicle tracking, and connected cars obviously make it easier to gather this type of data on how drivers are behaving behind the wheel, which in turn could inform PHYD-based discounts. However, startups such as Ownli and MileAuto are also bringing similar capacities to nonconnected cars, OMERS Ventures principal David Wechsler said.

MileAuto, which offers its own insurance product, noted that a driver’s “base rate and [their] per mile rate are based on standard insurance variables like [their] driving history, location, vehicle type and use, age and gender, years of driving experience, credit history, and coverages and deductibles.”

If you raised an eyebrow at the mention of credit history, you are not alone. Says Wechsler: “There is a lot of anxiety in the market about bias in underwriting where things like credit score are used as a proxy for driving risk. There is a proven correlation between credit score and driving risk — but isn’t how you actually drive an even better proxy for risk?”

For Wechsler, the logical next step is also a difficult one: “The best way to underwrite risk is on actual behavior, not proxies of behavior. But that’s really hard in insurance from a data, technology and regulatory perspective.”

Car manufacturers seem well placed to attack the challenge, especially as some are now moving into insurance themselves.

Automakers are now offering direct insurance, as seen with Tesla and others. This is a meaningful approach as the car itself is a complete connected car, and data is ubiquitous. The emergence of electric vehicles also presents an opportunity to create meaningful insurance options in the UBI and PHYD space given the seamless integration of data and continuous data streams. The more connected the car becomes the more meaningful the insurance will get, and the more meaningful will UBI/PHYD get. — Amir Kabir, AV8

Kabir is an investor in HDVI, which uses technology to rethink commercial trucking insurance, so he is by no way a naysayer about the space. However, he raised a point I have been wondering about, noting that “a bigger question still is if UBI will actually bring insurance premiums down and if driving data will be fairly assessed.”

What happens with UBI will be important to track, because it will give us food for thought on whether these approaches expand beyond driving — for instance, to connect wearables and health premiums, or smart home devices and home insurance. That’s a question for another day, but one to ponder for sure.