Waymo Is Now Completing Over 250,000 Paid Robotaxi Rides Per Week and Losing Money on Every One

Waymo has crossed a milestone that would have seemed optimistic even three years ago: the company is now completing more than 250,000 paid robotaxi rides every week across its operating cities, which currently include San Francisco, Los Angeles, Phoenix, and Austin. That is more than 1 million rides per month from fully autonomous vehicles with no safety driver in the car.

The operational scale is genuinely impressive. Waymo's fleet has driven over 50 million miles autonomously on public roads, and its safety record, while not without incident, compares favourably with human drivers in equivalent urban environments. The technology, by most measures, works. But working technology and viable business are different things, and Waymo's unit economics remain the most interesting and least-discussed part of its story.

The Cost Problem

Waymo does not publish detailed financials, but Alphabet's quarterly reports give some indication of the scale of investment. Waymo remains firmly in the loss-making category despite its impressive ride volumes. The core issue is that autonomous vehicle technology is extraordinarily expensive to deploy per vehicle. Each Waymo car carries sensor suites, compute hardware, and software licensing costs that add up to a figure substantially higher than a conventional vehicle or even a well-equipped human-driven ride-hail car.

Industry analysts who have modelled Waymo's unit economics estimate that at current cost structures, the company would need to charge somewhere between 3 and 5 times current Uber or Lyft rates to break even per ride. Waymo currently prices its rides at broadly similar rates to those platforms, which means it is subsidising every ride with investor capital, specifically Alphabet's.

The bull case for Waymo is that hardware costs will decline as the technology matures and scales, that software and sensor costs follow roughly similar curves to other tech hardware categories, and that at sufficient fleet scale the economics flip decisively in its favour because it does not pay driver wages. The bear case is that the safety and regulatory requirements for autonomous vehicles create a cost floor that does not have a straightforward downward trajectory, and that the window to achieve scale before well-funded competitors catch up is narrower than it looks.

What 250,000 Rides Per Week Actually Tells Us

The ride volume number is meaningful in two specific ways. First, it generates real-world data at a scale no competitor currently matches. Every ride adds to Waymo's training dataset, informs its safety models, and helps it handle the rare, difficult edge cases that autonomous systems struggle with most. That data advantage compounds over time.

Second, it builds the operational muscle, fleet management, remote monitoring, vehicle maintenance, customer experience, and regulatory relationships, that turning a robotaxi service into a reliable business requires. These are not purely technical problems. They are logistics and operations problems, and 250,000 rides per week is genuine operational experience.

For comparison, Cruise, which was Waymo's most credible US competitor, suspended operations in 2023 following a serious incident and has not returned to full service. Tesla's robotaxi ambitions remain largely in the announcement phase. BYD and other Chinese players have significant deployments in China but face regulatory barriers in Western markets. Waymo's operational lead is real.

The Longer Game

Waymo's parent company Alphabet has the financial resources to sustain losses for an extended period, which is both its greatest advantage and the thing that makes its competitive position difficult for others to replicate. No venture-backed startup can absorb losses at Waymo's scale for as long as Alphabet can.

The genuine question is whether the economics tip into viability before Alphabet's patience or strategic priorities change. At 250,000 rides per week, Waymo is further along the path to answering that question than any other autonomous vehicle company in the world. That is not nothing, even if the business model remains unproven at the margins.