On the incredible, shrinking cost of urban transportation.

I call this claim the “Big Ass Assumption” (BAA). The BAA that a critical cost component goes away isn’t as crazy as it sounds. As Mark Suster points out, deflationary economics are often the driving force behind both (i) market expansion and (ii) new use-case creation for older technologies.

What better case study is there for this concept today than the cost of urban transportation — a pervasive cost for any business that wants to touch the physical world?

I believe automotive autonomy will drive transportation costs down precipitously over the next decade for several reasons:

  1. Automobiles will move from merely 5% utilization to something approaching 80–90% when part of a fleet. The amortized cost per mile will fall by at least an order of magnitude from this effect alone.
  2. Drivers will be removed from the equation. An analysis by John McDermott suggests that, for UberX drivers in Los Angeles, the ratio between hourly driver earnings and vehicle costs per mile is 3:2. Without the driver, 60% of the cost of the ride would be removed.
  3. Auto insurance will become cheaper as accidents drop dramatically. Today, a per mile policy with Metromile runs around $0.09/mile including the base fee. More than 90% of auto accidents today are caused by human error, and KPMG estimates that more than 80% of these accidents can be avoided with autonomous technology. It’s conceivable that we could see another order of magnitude reduction of costs here.
  4. As consumers become increasingly more reliant on “transportation as a service,” ownership of cars will shift from individuals to enterprises — the latter of which have massive purchasing leverage with OEMs. The market for computing infrastructure went from individual usage (e.g. a server in your bedroom), to shared ownership (e.g. a server in a co-lo), to full rental (e.g. a server in Amazon Web Services) in a few decades, causing the disruption of companies who make computing hardware. With AWS driving server costs down by 50% every 3 years (or, an order of magnitude every decade), imagine the fate of GM, Ford, and Chrysler when their products are further commoditized and deflated.

These trends should put new research from ARK Invest in context. It shows that Shared Autonomous Vehicles (SAVs) will be cheaper than even walking, when one accounts for the cost of cheap calories to fuel a mile of ambling:

SAVs will be a screaming deal especially for short distances. Public transit may still have a place for longer hauls, but much of the short distance urban commuting we see today will be cheaper and more convenient with SAVs:

Such incredible deflationary economics in urban transportation will create a sea change in the way we design cities. It will also create a sea change in the services that move people or things.

Once the 1st order problem of autonomous urban transit is solved, the biggest boon of all may be the “apps” built on top of this new infrastructure. Think of business models that have been logistically intensive and historically challenge: food delivery, product rental, etc.

I’m excited to see “apps” built on the infrastructure provided by companies like Uber, Didi, and their brethren. Business models that have been economically challenged historically (e.g. on-demand delivery, product rental) may all of the sudden make sense. The BAA of effectively free urban transit should create many new opportunities that entrepreneurs will no doubt pursue in earnest…and many of them will work this time!


  • No comments yet.
  • Add a comment