Car Insurance: When Google Shrugged

The global market for auto insurance is expected by some to surpass $1T in five years. The U.S. market alone already represents a $300B opportunity.

The market is driven by nearly $6B of total annual television, radio, and digital advertising. The mature, low-margin business is uniquely characterized as highly fragmented (with more than 50 providers) and highly regulated (by 50 state insurance commissions).

The ad-driven nature of the business has made it attractive to startups and tech companies particularly as attention has turned to new underwriting models based on vehicle data – so-called usage-based insurance of the sort offered by Progressive in the U.S. It was likely for those reasons that Google dipped a toe into the market nearly five years ago with Compare.com.

That foray by Google didn’t last long. After spinning up an insurance shopping Website with multiple carriers and a back end provided by Elephant, part of the Admiral Group, Google balked and backed out.

It’s important to consider that decision now that car makers are entering the car insurance game including Toyota (Toyota Insurance Management Solutions in partnership with National General, Nationwide, Safeco, and Travelers), Tesla (in partnership with State National Insurance), and General Motors (with OnStar Insurance in partnership with American Family Insurance). These companies, like Google, see an opportunity to deliver a new customer experience and to, of course, make money. But will these new industry participants have the stomach for a high cost, low margin, high churn, and heavily regulated business? 

There’s a reason car insurance has attracted so many service providers. Car insurance is widely regarded as a loss leader business intended to on-board customers for the purposes of selling other lines of insurance – mainly renter’s or home- owner’s insurance.

Car insurance around the world is a mature business. There are rarely any significant shifts in market share or profitability as the rules of the game of underwriting are well understood and rooted mainly in assessing risk from driving history and credit scores and some other factors selected to avoid any bias (gender, ethnicity, location of residence) threatening protected insured classes.

The emergence of car data-driven underwriting and the onset of artificial intelligence to streamline claims has opened the door for startups and innovative market players to gain advantage in this otherwise mature market. Insurers are generally perceived as slow to adopt new technologies and the industry’s regulatory structure further retards industry change.

Nevertheless, car companies have sought to either offer car insurance directly, in the case of Toyota, Tesla, and GM, or to create insurance marketplaces. Many believe strongly that usage-based insurance – the now-10-year-old effort to base insurance on directly measured driving behavior – remains the future of underwriting.

I am personally sympathetic to the vision of connected insurance. I am also realistic. At the end of the day, I have found the cheapest insurance to be readily available from insurers that are not interested in my driving data. Progressive may want to know what time of day I am driving, how much I am driving, and how frequently I brake or accelerate harshly – other UBI players are looking solely at miles driven – but I’d honestly rather not have my insurance company evaluating my driving.

Call me crazy but I think my views are pretty widespread. Few analysts expect usage-based insurance to be embraced by more than maybe 10 percent of any given population.

Google’s reversal, four years ago, of its auto insurance market entry is remarkable given the widespread use of Googlemaps and Waze on smartphones. One could have imagined a smartphone-based usage-based insurance front end on an insurance marketplace where service providers could bid for those drivers with the best scores.

The prospect of Google capturing all of that search activity and adding that nexus to its already dominant search market presence must have been an appealing prospect. The whole point of usage-based insurance is to enable innovative or upstart players in the market to steal away the lowest risk (i.e. “best”) drivers from existing insurance carriers.

Usage-based insurance even promises the prospect of a stickier relationship with the customer, cutting down on churn by tying their policy to their driving data. The 50 state insurance commissioners that must approve the underwriting model, however, was a key fly in the ointment. Google would not be able to rapidly scale the platform and therefore opted out. The margins didn’t look that great either.

The reality is that the mature car insurance business is increasingly characterized by churn. Buyers know they must check prevailing rates annually and jump from carrier to carrier to ensure they are paying the least. This is why it is problematic for car makers to get into the game.

There are three primary motivations for car companies to enter the car insurance business:

1.    Claims management – to capture vehicle repair and replacement opportunities and retain customers with a streamlined and integrated process
2.    Cost management – to offer preferential rates in order to keep third-party insurers honest;
3.    Revenue & Retention – to offer car insurance as part of an all-purpose subscription service package

Many auto makers have sought to offer their customers an insurance marketplace where consumers can select their carrier of choice and where the auto maker might integrate their vehicle data for usage-based insurance quotes. This approach is attractive to the car maker because the car maker can take a cut of the insurance business and charge a bounty for insurance leads.

What has yet to be seen in any of these marketplaces is some kind of system of rate comparisons and automated rate shopping. The marketplace approach is doomed to fail because many of the market leaders refuse to participate and because there is too much churn – and auto makers typically are not working in the customer’s interest, which would include helping the consumer jump from carrier to carrier chasing the lowest available rate.

Car makers like GM and Toyota which are simply trying to grab a share of the insurance business are doomed to fail. It will be nearly impossible for these companies to compete on the open market and dealers – that the car makers will have to rely on to pitch their captive programs – are increasingly offering their own insurance products.

Tesla is the exception. Tesla typically has longer-term interests in mind and has wider flexibility in its pricing and marketing strategy. Tesla is also powerfully motivated to provide more competitively priced insurance to counter rates that may rise as the company continues to promote clever though sometimes confusing semi-automated driving technology. Tesla may even be interested in the renter’s or home owner’s insurance business.

More importantly for Tesla, the company completely controls its distribution channel and customer relationships. Tesla doesn’t have to worry about its priorities becoming entangled with conflicting dealer objectives that play out in the F&I office when the customer is signing on the dotted line.

GM, Tesla, and Toyota are all steadily bringing their insurance offerings to a growing number of U.S. markets. GM says its insurance offering is available in six states and expects to be nationwide within a year. Tesla plans to issue policies through Redpoint County Mutual Insurance in Texas, Midvale Indemnity in Illinois, and Homesite Insurance in Washington. TIMS, Toyota’s offering, is available in 38 states.

Among the discounts Tesla is reportedly planning to offer is an Autopilot discount based on the autonomy level of the vehicle. Such an offer shows how Tesla can leverage a captive car insurance product to sell an optional $10K full self-driving upgrade.

By comparison, Toyota and GM are exploring traditional 10% or 20% usage-based insurance discounts, provided the customers are inclined to share their driving data. Again, this is a non-starter. I not only don’t want my insurance company to know how I drive, I’d rather my car maker and dealer were ignorant of this as well.

Google was wise to steer clear of the auto insurance business. GM and Toyota are destined to pointlessly spin their wheels with these efforts. Tesla alone has a compelling value proposition capable of attracting customers and transforming car insurance as we know it.