GM's Wrong Way Insurance Plan

Word arrived last week that General Motors was returning to the car insurance business, after a decade-long sojourn post-bankruptcy, in the form of a tie up with Homesite Insurance Group, an affiliate of American Family, and a service launch for GM employees in Arizona – land of quick-draw insurance startups. This is a bad idea and a waste of time and money – but why?

Let’s look at the potential strategic rationales for GM to get into car insurance:

• To capture premium revenue and take advantage of superior vehicle system knowledge to offer more profitable underwriting at lower rates.
• To better manage insurance claims and thereby capture vehicle repair or replacement revenue opportunities via franchise dealers – and ensure use of genuine GM parts and superior calibration of safety systems.
• To offer usage-based insurance and create a consumer opt-in program for gathering driving and driver behavior data – critical to developing the next generation of usage-based insurance offers and to the development of semi-autonomous and fully autonomous vehicles.
• To offer a linked-up OnStar/insurance usage-based insurance program whereby GM owners that keep their safety systems turned on (a major drawback to these systems once users become annoyed with alerts and warnings) will be rewarded with insurance discounts.
• To offer competitive rates for GM vehicles thereby bringing down rates from competing insurance companies for GM makes.
• To deliver a transformative user experience sufficiently unique and streamlined as to give GM a leverage-able market advantage.
• Leverage new car sales to build a portfolio of insured vehicles and disrupt the mature car insurance business.
• Create a range of novel car insurance offers capable of supporting subscription, rental, ride hailing, car sharing, or other mobility-related episodic vehicle use cases.

Now let’s look at GM’s stated objectives – as seen in interviews given by Andrew Rose, who has joined GM as President of OnStar Insurance and Vice President of Global Innovation, and others:

From the Wall Street Journal:

• “Customers who sign up agree to have their driving habits tracked, and those who obey the speed limit, avoid sudden stops and practice other good-driving behavior will be rewarded with cheaper rates, GM said.”
• “Who knows more about your vehicle than the people who manufactured it?” – said Rose.
• “GM believes the data it can collect directly from its vehicles will deliver more-precise information than insurance companies can now access.”
• “To start, the company will set rates much like a typical insurer, relying on traditional factors such as ZIP Code or the amount of driving the policyholder does, Mr. Rose said.”
• “Eventually, GM will more heavily weigh driving behavior and other variables in determining rates.”
• “Eventually, it could even calculate the amount of times safety features such as emergency braking are deployed, he said.”
• “Another example: GM could track whether tires are inflated properly, which improves stopping distance and reduces the risk of crashes, Mr. Rose said.”
• “Running an insurance division should help GM prepare for a day when it may offer a commercial driverless-car service, Mr. Rose said.”

And from Techcrunch:

• “Our goal is really to create greater transparency and greater control for our customers in influencing what they pay for insurance and their total cost of ownership on the vehicles,” said Russell Page, GM’s head of business intelligence in a recent interview.

To be clear, none of the reasons expressed by Rose or Page makes any sense. There are existing usage-based insurance programs in place suitable to satisfy the tiny minority of drivers (about 10%) that are interested in this sort of coverage – normally either really good drivers who have self-selected into such programs or young drivers willing to do anything (including sharing their driving data) to get a better rate.

Neither Rose, in his public comments, nor GM in its press materials has described the elements or details of a unique, compelling, disruptive, or transformative insurance value proposition. In fact, the new program which is only available to GM employees in Arizona while it awaits regulatory approval in that state, will need to pass regulatory muster in 49 other U.S. states before becoming a national offer. This will take many months.

It is most likely that the real motivation lies in the voice over from a Youtube video describing the new service which also describes the broaching of an OnStar heresy: “In the future, if you ever have an accident, we intend to make an instant connection to your dealer providing speed of light damage assessments, loaner car arrangements and more without hassles.”

OnStar Insurance video: https://www.youtube.com/watch?v=HllxgyXHwPE

While connecting drivers of crashed cars to dealers is a so-far unachieved Holy Grail of claims management for car makers, there are reasons that no car company has crossed this line. Consumer distrust of insurance companies is only surpassed by their mistrust of dealers. 

Presumably this connection will be made by OnStar operators, which points to an evolutionary connected car trend impacting call centers – the emergence of the direct connection from crashed cars to emergency responders. Ford Motor Company’s 911 Assist function – an alternative to OnStar – operates without the support of a call center manned by human beings trained to interact with drivers in a post-crash situation.

SiriusXM announced its own direct access 911 service offered in partnership with RapidSOS for existing and future connected car customers. GM appears to be doubling down on its expensive call centers by adding claims management to their post-crash responsibilities. One has to ask: Is THAT a good idea? If I am bleeding and in shock after an airbag-deploying crash, do I really want to talk to my dealer?

There are other challenges including the fact that the path to market will either require the participation and support of dealers or be offered directly to customers through digital means, or maybe GM could opt for a hybrid dealer/digital delivery system. Either way, introducing insurance into the vehicle sales process is an iffy proposition given the state of current automobile retailing with its time-consuming paperwork, credit checks, test drives, and shopping, and negotiation.

GM is fundamentally a B2B operator selling cars to dealers. Is the company really equipped to manage or evolve the insurance user experience? Judging by the recent failure of the company's Maven car sharing operation it is safe to say that user experience is not a strength.

Perhaps GM is seeking to add insurance into a single-price, bundled, vehicle subscription plan – requiring OnStar Insurance. That’s a possibility. It’s worth bearing in mind that most new vehicle purchases involving an existing vehicle owner require a simple swapping of license plates and insurance. Changing the insurance company to the maker of the car at point of sales - the dealership - may be asking too much of the average dealer.

And, in the end, is it worth the effort? What is the goal? Is GM trying to compete with established insurers? Why? What does it have to gain?

Ask an executive operating in the car insurance business and they will say car insurance is a loss leader business intended to attract new customers in order to sell more profitable lines – such as home, renter’s, or life insurance. Car insurance is a mature high cost, low margin business.

In fact, the pressures on insurers to better understand the rising influence and impact of advanced safety systems and the sophisticated sensors that support them is driving up the cost of claims (repairs) even as the number of crashes declines. OnStar is launching its program in the midst of a rapidly expanding adoption of collision avoidance systems and it is possible that the collection of vehicle data might help GM better understand the efficacy of the company’s systems.

But GM doesn’t need to create an insurance operation to collect and analyze this data. GM has, in fact, been looking at the nature and severity of vehicle crashes from the company’s inception – only expanding the effort after the launch of the OnStar connected car service.

It is true that usage-based insurance has had an impact on the industry – introducing a new metric capable of transforming underwriting. Progressive has pioneered usage-based insurance based on the use of an aftermarket data collection device plugged into a customer’s car for a limited period of time to determine the risk level based on miles driven, time of day driven, and harsh braking and acceleration.

Across the industry, though, insurers are racing to understand the impact of advanced safety and automated driving systems on claims. There is a related need to understand the behavior of drivers driving cars equipped with these systems. But this does not appear to be the stated objective of the new insurance effort and, again, GM need not get into the insurance business to gather this data.

Besides, usage-based insurance is fundamentally a radical invasion of privacy. Who really wants their insurance company looking over their shoulder while they drive? No one wants to be TOLD how to drive or have their driving EVALUATED. Some might not mind a mileage-based policy if they, in fact, don’t drive much.

OnStar Insurance says it will offer bundled car, home, and renter’s insurance and may dig deeper into vehicle data for underwriting insights – but why not leave this to the experts? And who exactly will sell the policies and how? 

More perplexing is the fact that GM already offers an insurance marketplace accessible to owners of GM vehicles willing to share their driving data via their OnStar connection. Is GM now going to compete with its own marketplace partners and will those partners have access to the same set of vehicle data as OnStar Insurance or will OnStar Insurance have an unfair advantage? Of course, GM won’t have that same advantage when competing to insure non-GM vehicles which it apparently intends to do.

To launch the program, GM has brought on board two stars from the world of insurance innovation – Andrew Rose of Compare.com and Katie DeGraaf of Allstate’s Arity division. Rose appears to be the “face” of the effort. DeGraaf brings actuarial heft.

Rose’s shift to GM from leading car insurance comparison site operator Compare.com raises a variety of questions. He comes to GM after serving as president of Compare.com, which may have burned through as much as $185M in investor capital, with its last funding round concluding in 2018. Compare.com is known for its digital insurance platform and is at least partially owned by Admiral Group which also owns Elephant Insurance and Confused.com, another financial services comparison site based in the U.K.

Compare.com provides comparisons of car insurance rates from 60 insurers. Four years ago Compare.com and Google parted company after a failed attempt to create a Google-managed car insurance marketplace. Rose led the Google effort, which crashed and burned casting into question Compare.com’s at-the-time tantalizing valuation and nine-figure investment.

Four years later Rose is leading GM’s new misguided (unguided?) foray into car insurance. His success will hinge on how much authority he has been granted regarding hiring, investments, potential partnerships or acquisitions, or integration with existing GM systems, vehicle development strategy, marketing, or its dealer network.

Just getting access to vehicle sensor data will be an expensive internal GM challenge. Woe be to Mr. Rose when he attempts to muscle in on vehicle design or marketing activities.

Backing up Rose is Katie DeGraaf from Allstate’s Arity mobility insurance division. By all accounts DeGraaf is something of a star but she too will face internal skepticism at hide-bound GM. If the going gets rough will Rose and DeGraaf join the conga line of external hires that have flashed like SpaceX satellites across the GM firmament on their way to better opportunities?

Tesla Motors entered the insurance business to better understand and manage the cost, complexity, and inconvenience associated with vehicle claims and repairs. Musk also asserted that his vehicles were safer than most and his entry into the insurance business was intended to keep independent insurers honest in their underwriting.

Toyota Motor Company is the largest shareholder in MS&AD Insurance Group Holdings which owns Aioi Nissei Dowa Insurance which, itself, acquired the U.K.’s Box Insurance Group (including InsuretheBox and DriveLikeaGirl) five years ago. Toyota offers car insurance with an emphasis on usage-based programs in multiple markets around the world under a variety of brands in partnership with these operators.

The interest in vehicle data and the desire to apply insights derived from that data into a competitive position in the car insurance business and capture vehicle repair and replacement revenue is a real business proposition, but will GM’s customers embrace this value proposition or be turned off. Using vehicle data to better understand the interaction between humans, vehicles, and safety systems is a loftier objective and one that may save lives – but doesn’t require creating an insurance division. It’s difficult to say whether GM is being cynical, foolhardy, or dishonest.

GM can achieve most of its corporate objectives of capturing sales and service leads, retaining customers, and better calibrating safety systems without selling car insurance. The entire exercise appears to be pointless and misguided – or it could introduce a closed loop relationship between GM and its customers and dealers. The track record thus far is unimpressive. To be successful GM will need to be more transparent and will also need an entirely new relationship with its customers and dealers.

GM’s public-facing stance in the era of CEO Mary Barra has been one of fighting off ignition switch litigants, embracing Trump Administration de-regulation, and resisting Takata airbag recalls. GM is going to have to change its tune if it wants consumers to sing along with OnStar Insurance.