Uber Stock After Q3 2026 Earnings: Robotaxis Do More Trips on Uber's App

Uber's CFO said something on the Q3 2026 call that goes straight at the biggest worry about the stock. Asked about robotaxis, Balaji Krishnamurthy said Waymo and Tesla cars in the markets they launched on their own this year do trips per vehicle per day "in the low teens." Autonomous cars on Uber's network, in markets that mix robots and human drivers, do "high twenties." The bear case on Uber is that robotaxi operators won't need it. I think that gap is the price of going it alone, and it's the best evidence yet that Uber's network carries into autonomy.

You may have seen that Mobility revenue fell 2% while Mobility bookings, the total fares riders paid, rose 18%. That's mostly bookkeeping. Since Q1 2026 Uber has counted only its own cut as revenue in some markets, instead of the whole fare, which shrinks revenue without changing what riders spend. Profit is the better test, and I'll get to it after the robotaxi numbers.

A robotaxi costs about the same whether it carries riders or sits empty, so trips per car per day decide whether it makes money. A human driver logs off when demand is thin. A robot sits on the operator's books. That makes a pool of waiting riders worth more with robots than it was with drivers.

There are two comparisons in that table and they're different sizes. Against Waymo's and Tesla's new solo launches, Uber says its network roughly doubles daily trips. Against mature solo fleets in the US, the edge is about 30%, and Krishnamurthy said that comparison includes San Francisco, a market he called especially good for robotaxis. The smaller number is the one I'd lean on. A 30% edge on a car whose cost is mostly fixed is a lot of margin.

The mechanism is the one that runs the human business. Uber has riders already open on their phones, so it can fill a robot's dead time between trips with demand a single-operator app doesn't have yet. If that holds, the network effect survives the driver leaving. Its job changes from finding drivers to filling cars.

The scale is tiny. Seven hundred driverless cars is nothing next to a business that handled 4.1 billion trips across rides and delivery in Q3. What the quarter gave is the first operating numbers on the question everyone argues about.

Utilization figures come from Uber, about rivals who don't report them. The number I can check is how much of each Mobility booking dollar Uber keeps as segment operating income, its profit measure since January 2026. A network losing riders to robotaxi apps has to pay them to stay, and that shows up as a lower share.

Mobility kept 8.0% of bookings in Q3 2026, against 7.4% a year earlier. I have to be careful with the first step up. It came in Q1 2026, the same quarter the reporting change began, and I couldn't confirm from the filings that the change left profit exactly untouched. The step from Q2 to Q3 is the clean one, and CEO Dara Khosrowshahi gave the reason: "we are taking down rider incentives and we are using those dollars to fund driver incentives."

So this year Uber cut what it spends on riders and kept more per booking dollar, in the same year Waymo and Tesla started running their own apps. Delivery, the same gig network with no robots in it, kept more too, so part of this is Uber-wide discipline rather than anything robotaxi-specific. Krishnamurthy put the Q3 slowdown down mainly to driver supply abroad, with fuel costs from the Middle East war hitting Brazil and India hardest, and said US growth moderated less.

For the shares, the debate is whether Mobility is a business robots will route around. If the utilization edge is real, the robotaxi era raises Uber's value as the place fleets go to get filled, and the margin should keep holding as the fleet grows. That's why I've put Uber next to the two operators and next to DoorDash, which runs the same kind of network in delivery and is here as a comparison, not as evidence. My DoorDash piece and Tesla piece cover their sides.

The strongest objection is that a rising margin and a slowing network look the same in one quarter. Mobility bookings growth slowed in Q3 to 17% in constant currency, from 20% in each of the first two quarters, and Uber chose this year to take money away from riders. If riders are drifting to Waymo where it runs alone, cutting rider spend and booking a higher margin is what a company does when it defends profit instead of share. The margin would rise right up until the volume went.

Waymo has said as much in public. It told Uber in July it plans to launch its own app in Austin and Atlanta in January 2028 (TNW), and on November 4, earnings day, co-CEO Tekedra Mawakana told Bloomberg the Uber experience was a "lottery" for riders who can't tell whether a robot or a human will turn up (Ubergizmo, citing Bloomberg). That goes straight at the hybrid model behind Uber's best number.

The hybrid number has a second weakness. In a mixed market Uber's dispatch decides which trips go to robots. If it sends them the dense, easy rides and leaves the rest to humans, trips per car will look great without telling you much about the network. And the 30% gap against mature fleets is still Uber's comparison of Uber's markets. Neither Waymo nor Tesla has any reason to publish a figure that agrees.

I take all of that seriously, which is why my test pairs the margin with growth. If Uber is harvesting, the margin holds while bookings growth keeps sliding, and I've written that into the claim below. If the network is intact, Uber shouldn't need to buy riders back as Waymo's own app reaches more of its cities. The utilization gap is the claim I'd most like someone outside Uber to confirm, ideally with dispatch mix disclosed.

1. 1February 2027, the Q4 2026 report: Mobility segment operating income as a share of Mobility bookings, last 8.0%. Back below 7.4% and the network is paying to hold riders.
2. 2Mobility bookings growth in constant currency, last 17%. Two more quarters below 15% with the margin flat or rising would look like harvesting.
3. 3Whether Uber gives an updated, numbered utilization comparison, and whether Waymo or Tesla ever publish their own trips per vehicle per day.
4. 4The driverless count against management's 700 by end of 2026 and over 2,000 by end of 2027, and how much of it sits on Uber's balance sheet.