Companies around the world are investing many billions to win the self-driving race, and from time to time the payout is described as being worth trillions. Tesla’s $1T market cap is often attributed to hope they will be dominant in this space. At the same time, others question if there’s much of a business there at all. As I was one of the main original advocates for this robotaxi vision, it’s time to examine it again. But it’s not about ride-hail.
Put most simply, the value proposition is to create a self-driving “Mobility on Demand” service which is good enough that a decent fraction of people decide they can partially or fully give up some or all car ownership. Today, the world spends around 5 trillion dollars annually on ground transportation expenses, so it’s entirely possible to see trillions of value for the companies that revolutionize that. A company that sells rides and shipping as a service is selling the entire value chain, not just the vehicle.
This is not the only way to make money from self-driving of course, and many companies have different targets. In fact, today all we see are ride-hail companies that are more like traditional taxis–higher priced rides on demand which don’t easily replace owning a vehicle. Uber brings in $44B/year, which is far from chicken feed, but a small fraction of the big picture. We also see companies trying to make self-driving cars to be owned by consumers, hoping to create the most must-have car feature and prosper in the vehicle market. Last week saw the launch of the “Tensor” in that area, and Tesla hopes to make all their cars retroactively capable. Waymo and Toyota have also announced plans to build such cars. There are many other applications in delivery, industry, agriculture, mining and more. Since today the typical car owner spends $7,000 to $12,000 per year on car ownership, taking over that is highly lucrative, but hard to do with a taxi service that’s much more costly per mile.
Not Just The Vehicle Industry
Most of the personal transportation industry today revolves around car ownership. There are taxis and there’s public transportation but they are small in comparison. The brass ring is here. People love car ownership, and spend outrageously on it, it’s often the 2nd largest personal expense after housing. It’s not economical–alternatives like transit can cost much less, at least when subsidized, but people will pay for this luxury.
Only a fraction of the population will be willing to replace car ownership for quite some time, but it’s much easier to convince people to give up 3rd and 2nd cars in a household, and some will give up all of them. New generations are generally more amenable to big shifts like that. Only 1/3rd of households have only one car, so 2/3rds are ripe for car replacement at the right price.
If you replace car ownership for somebody, you take over much more spending than just vehicle purchase. You become the replacement for the oil/electric companies, car loan banks, dealerships, cleaning, parking, insurance, service, parts, tires and more; all of these are large industries. You subcontract many of these things but you are in charge, you own and bill the customer, and control where the money flows, and you make sure the best part flows to you.
Tesla has pioneered some of this. Tesla has some customers for whom they provide the car, the service, the charging or solar on the roof, the insurance, the service and the financing, all in the traditional car industry way.
You need to offer a great service that almost fully replaces the private car. That means an urban robotaxi is just the start, you must also please customers going on ski trips and other road trips, towing boats to the lake, taking kids to soccer and every other reason people own a car. That’s pretty hard, so you won’t get everybody, and you’ll have to offer sweeteners like lower costs and lower hassles to make up for what you can’t deliver. Fortunately you can offer things like not having to drive or park, not needing a garage, no ownership hassles, access to every different kind of vehicle on demand and several other factors.
You also have to provide great mobility on demand. One’s personal car is always there for you immediately. There’s never a wait time (but sometimes it’s parked further away.) There’s never a surge. You pay for that exclusive access, but people today are willing to pay. And there’s no really great replacement for the ability to keep your “stuff” in your private car and have it always there with you. There must also be a solution for everywhere the robotaxis won’t go, such as human driven service, a robotaxi that takes you to a seamless transfer to a you-drive rental car, or a car delivery service that brings a car for those trips out of town.
As tough as great mobility services are, in many cities of the world, many people give up on car ownership. Even the American city of New York. It’s much harder to obtain that in places like suburbs, where car ownership is made very easy, thanks to garages in every house and free parking at every business. Even though people may not be thrilled with what that did to urban design, it would take many decades to reverse. To win the battle of mobility service vs. personal robocar ownership, the main tool a service can have is price. Of course, a company can and will be a player in both the car replacement and the car business, so they may not need to “win” that battle.
The self-driving markets
These leaves the following core markets for self-driving technology, in rough order of market size:
- Car replacement (subscription)
- Feature on consumer cars (first luxury, then standard)
- Shipping (long haul, middle-mile, last mile, sidewalk)
- Specialty vehicles/transit
- Ride-hail/taxi
- Off-road/Campus (agriculture, industry, mining, airports, etc.)
Being Lower Cost
The all-in cost of car ownership today is estimated to be around 60 cents/mile plus parking, though it can be much less if you car is old, more if new and expensive. $2-3/mile ride-hail services can’t compete with that if the robocar owner no longer has to drive or park. Worse, may car owners act like their car only costs the incremental cost of operating it, which can appear be under 5 cents/mile in an EV. That’s an illusion, but it’s how people think.
The “cost of goods sold” for a low-end robotaxi can get very low, perhaps under 20 cents/mile. (It won’t get there for some time until R&D costs reduce and things happen at large scale.) Even though, it will require new, creative pricing models to compete with how people perceive the cost of a private car. Flat rate pricing, at least for some aspects of the car, will be a must.
In addition to low cost, the new markets will rely on one key difference between robots and humans. Robots don’t mind waiting, nor do they mind working 24/7. There are things robots will do that human drivers never would, and those open up or make economical new classes of services. People have had chauffeurs and taxis since before there were cars, the new revenue opportunities come from services that were not possible before
Until then the robocar industry will be limited to the “small” markets of ride-hail and luxury car add-ons and subscriptions. Car replacement may succeed or fail in different markets, though if it fails, there’s still plenty of money to make, just not as many trillions.