Everyone treats autonomy as the disruption. It isn't. The disruption is whatever's cheapest — and robotaxis only matter on the day they actually win that fight.
The real threat was never "robotaxis" as a category. It's autonomous vehicles that are already owned outright — an autonomous car that's paid off can run for close to the cost of electricity plus a small margin. As long as a robotaxi ride is priced higher than the human-driven alternative, it's close to irrelevant, no matter how impressive the technology is. Riders don't pay for autonomy. They pay for the cheapest way to get somewhere.
Take a dense city like San Francisco. A healthy ride-hailing market, when prices are cheap, runs at roughly 2 trips per inhabitant per week. A million people means roughly 2 million trips a week. At around 100 trips per driver per week, that's about 20,000 cars to run the entire market on ride-hailing.
But a robotaxi fleet doesn't need anywhere near that to start mattering. Even 1,000-1,500 robotaxis can be genuinely useful in a dense city that size — because what riders actually care about is time-to-pickup, and density drives that far more than raw population does. A sprawling, lower-density city like LA needs roughly 10x as many cars to hit the same relevant coverage. There's also no partial rollout the way there is with human drivers: you can't run a robotaxi fleet "Friday and Saturday nights only" the way you might push a weekend incentive campaign — the cars are available around the clock regardless, so covering only peak demand isn't really an option for an operator trying to make the unit economics work with a small fleet.
The intuitive guess is that robotaxis hit low-wage markets first, since fuel is a bigger share of the fare there. The opposite is true. Robotaxis at scale will take a very long time to reach markets like India, large parts of Africa, or Southeast Asia — roads and regulatory environments are far harder to operate autonomous fleets in, and critically, human labor is cheap enough there that robotaxis simply aren't economically competitive yet. The rule runs the other way from what most people assume: the cheaper the labor and the more efficient the existing human-driven market, the longer that market can avoid robotaxi disruption.
The current rollout pattern backs this up. Right now robotaxi operators are launching almost exclusively in expensive US cities, and pricing still isn't much below a human-driven ride there. If the goal were actually to win the market fast, the obvious move would be to give rides away close to free — the fact that isn't happening yet mostly reflects that there simply aren't enough robotaxis on the road to do it at scale, not a lack of ambition.
If still running a challenger operation right now, this wouldn't be a five-year worry. The fleet sizes needed to actually move a market's pricing aren't close in most places, the markets where it happens first are the expensive, low-labor-cost ones — not the ones most challenger operators are fighting in — and the transition, when it comes, will be gradual: combustion-engine price competition first, then EV price competition (far more relevant in lower-wage markets than high-wage ones, since vehicle costs are a bigger share of the fare there), and only later, full autonomy at fleet scale.
A short working call — I'll tell you what's actually worth watching in your market.
Book a working call