A little after 9 a.m. every weekday, crypto entrepreneur Alex Masmej orders a Waymo to shuttle him from his home in Hayes Valley to a coworking space in the Presidio.
“It’s a delightful part of the day,” said Masmej, who clambers into the plush Jaguar SUV, reclines the passenger seat, and blasts earsplitting dance music for his 15-minute commute. “You regain your freedom.”
Masmej is part of a growing cohort of San Franciscans who have essentially quit using Uber — barring exorbitantly long Waymo wait times or a required trip to the airport. (Waymo does not currently have a permit to operate rides there.)
His Waymo account recently ticked past 620 rides, including more than 1,600 miles traveled and 9,000 minutes spent in the vehicles. Masmej believes San Francisco is well on its way to becoming a total robotaxi town.
Since Waymo opened its service to the general public last summer, San Franciscans have watched the robotaxi’s signature silhouette become a regular part of life. As more than 300 cars in its fleet trundle up the steep incline of Telegraph Hill and along the boulevards of Outer Sunset, Waymo appears to be on an inevitable road to domination.
“The change in San Francisco has been dramatic,” Lyft CEO David Risher wrote last month in a blog post. “I can’t go two blocks without seeing a self-driving car.”
Just as San Francisco functioned as ground zero for the rideshare revolution in the 2010s, the quick adoption of robotaxis gives the impression that the city is a bellwether for another full-scale infrastructure transformation — one that won’t end well for the harbinger of the last paradigm shift.
“Uber’s product is getting eaten by Waymo,” said a former Uber product executive who left after the company went public in 2019. “The thing Uber did to everyone else, Waymo is doing to them now.”
In other words, Uber — whose global ascendancy wrecked the taxi industry — is toast. Right?
That’s what Travis Kalanick, Uber’s cofounder and former CEO, predicted would be the case if the company didn’t win the robotaxi race. In a 2016 interview, Kalanick said that if Uber didn’t have an autonomous vehicle before its competitors, “then the future passes us by basically, in a very expeditious and efficient way.”
“The entity that’s in first, then rolls out a ride-sharing network that is far cheaper or far higher-quality than Uber’s, then Uber is no longer a thing,” added Kalanick.
Kalanick’s predictions are not quite how industry experts see things playing out, however. Thanks to its big head start, Uber’s ubiquity means it may be the network that Waymo can’t live without. The flexible, low-overhead model in which drivers own their own cars makes it more attractive as a partner than as a competitor, especially at times of peak demand.
But that doesn’t mean Uber is safe, because beyond Waymo looms another threat to its business model — in the form of Elon Musk.
The master network
Uber once had ambitions to be the star of the robotaxi revolution. The company spent billions to outmaneuver its bitter rival Google and its Waymo self-driving car unit. Google sued in response, alleging that Waymo cofounder Anthony Levandowski stole trade secrets before selling his startup to Uber, leading to a $245 million settlement and his criminal conviction.
Uber waved the white flag on developing autonomous technology in-house in 2020, when it sold off its research unit, citing its high expense.
Still, Uber is the dominant rideshare player in the U.S., and globally. In 2024, Uber had $9.8 billion in net income — a four times increase from the previous year — while Waymo, which is owned by Google’s parent company Alphabet, has guzzled more than $11 billion from investors, with no profits in sight.
Uber is operating at a scale so large that at one point in December, there were 1 million trips happening through its platform simultaneously.
“That’s more than the entire population of San Francisco using our platform — at the same time,” CEO Dara Khosrowshahi wrote on X. Put another way, Uber makes 1,000 times the trips Waymo makes every week.
But data released in December appeared to show Waymo rapidly chipping away at the market shares of Uber and Lyft. When Waymo launched in 2023, Uber and Lyft had San Francisco market shares of 66% and 34%, respectively, according to YipitData, which uses customer receipts for its estimates. Fifteen months later, at the end of 2024, Waymo was at 22% — the same as Lyft — with Uber at 55%.
Uber, Lyft, and independent rideshare industry analysts dispute these percentages, which factor in only the rides within Waymo’s SF operating zone, thus excluding highways and trips to and from the airport.
An Uber spokesperson said year-over-year gross bookings in San Francisco have been consistent for several quarters. Lyft’s Risher said on X in December that San Francisco continues to be a growing market for rideshare. Within Waymo’s operating zone, Risher said, Lyft’s market share remained consistent in 2023 and 2024, according to a validated external data source.
Waymo declined to comment on the YipitData numbers.
“You can’t simply model San Francisco and the spotty data that exists for it and assume that it magically rolls out throughout the U.S. and that Uber’s roadkill,” said Scott Devitt, an equity analyst at Wedbush Securities.
The consensus is largely that the myopic view from tech-obsessed San Francisco is not representative of the big picture.
“People are looking at San Francisco, and they’re extrapolating it across the globe,” said Mark Giarelli, an equity analyst at Morningstar who is keeping a close eye on the experiments Waymo is conducting in U.S. cities. “That’s probably not the correct way to look at it.”
Instead of just a high-tech taxi service, Uber should be understood as a sophisticated system that can predict demand to boost utilization and profit from its fleet, he said. Uber is a “master network” able to move people and products around in the most efficient way possible, Giarelli said, while Waymo is essentially self-driving technology bolted onto a consumer app.
All that makes Uber “an ideal partner” for an autonomous car service in the early stages of building out its consumer base, Giarelli said.
Waymo has recognized this too. As the robotaxi company has probed entry into various U.S. markets, Uber has been a major beneficiary. Although Waymos are available exclusively through the Waymo One app in San Francisco and Los Angeles, they are only available through Uber in Austin and, soon, Atlanta. In Phoenix, the market where Waymo first launched publicly, users can order the robotaxis through both the Waymo One and the Uber apps.
Were Waymo to forego the benefits of cooperation and seek to compete with Uber head-to-head everywhere, it would face two major disadvantages in the matters of how much it costs to add cars to its network and who bears those costs.
Estimates peg the cost of a single Waymo, a Jaguar iPace outfitted with an array of high-tech sensors and cameras, at as much as $200,000. Waymo owns all the vehicles in its fleet itself, regardless of whether they’re out on the street ferrying passengers or sitting idle in a company-leased parking lot.
Uber, meanwhile, requires its drivers to own or lease their own vehicles, typically Toyota Priuses or other economy cars, and it only pays drivers when they’re carrying passengers.
When demand for rides exceeds supply, Waymo’s model offers only one answer: buy more cars. But demand is anything but static; it spikes during commute times, weekend nights, and special events, then falls off sharply at other times. Were Waymo to build a big enough fleet to satisfy demand at those times, it would sit unused for the rest of the day. The other choice is to build for the average level of demand; that would mean turning away some 40% of riders at peak times, Giarelli estimates.
Uber doesn’t have that problem, because its flexible two-sided-market model lets it continuously match supply to demand, with cars that aren’t actively bringing in revenue effectively off its books.
“Say there’s a Taylor Swift concert that spikes demand by 50%,” said a former Uber product manager who worked at the company between 2021 and 2023. “You can’t wave a magic wand and bring more robotaxis onto the street that don’t already exist. But you can incentivize drivers with more money and get them out on the road.”
Direct competition appears to be intense in San Francisco, where the apps coexist on users’ phones. But the rideshare experiment most analysts are betting on for the long term is an Austin- and Atlanta-style “Waymo on Uber” partnership.
In business speak, this is a “platform strategy,” meaning Uber functions as the underlying platform that supports many companies and ride options. This model allows robotaxi companies to benefit from Uber’s massive customer base and logistics network, while Uber reduces dependence on human drivers and lowers overall costs. In Atlanta and Austin, for example, Uber will provide Waymo with fleet management services, such as cleaning and charging, in exchange for a 10%-20% cut of revenue per ride, according to analyst estimates.
Uber has 14 partnerships with autonomous vehicle companies, including Waymo, Aurora (which bought Uber’s self-driving unit in 2020 for $400 million), and food delivery services like Avride and Serve. Lyft has adopted a similar approach, though analysts are skeptical about how successful its partnerships will be.
“If Uber plays their cards right, it will continue to be the go-anywhere, get-anything app,” said Chris Saad, who was a product executive at the company from 2015 to 2017. “Your muscle memory will be to reach for Uber.”
That’s unless the ultimate wild card pulls off a coup.
Uber vs. Elon
If one person can dethrone Uber, it’s Musk. He has long staked the future of Tesla on self-driving cars and posited a “glorious future” in which parking lots are replaced by parks.
Musk is coming directly for Waymo and Uber’s lunch. At Tesla’s “cybercab” launch last year, he said the steering-wheel-less vehicle will cost less than $30,000 and be available before 2027. Regardless of the ambitious timeline (Musk has repeatedly made predictions about autonomous vehicles that have not come to pass) — Tesla developing a low-cost, scalable car would be a serious game changer. Unlike Waymo, which uses costly LiDAR for depth perception and remote human operators for tricky situations — both of which are expensive and hard to scale — Tesla is building cars that rely solely on cameras and are fully autonomous under all conditions.
This future also threatens Uber. Musk has said that Tesla owners will eventually be able to rent out their self-driving vehicles and generate income as part of a Tesla robotaxi network. While questions remain about insurance and legal liability, Tesla could potentially have a huge fleet of cars able to come online when demand surges.
“Tesla is the 800-pound gorilla,” said Giarelli. “What they want to do is decentralize what Uber does.”
Devitt stressed that this future is entirely uncertain, and it’s currently the “hypothesis phase” of the robotaxi revolution. “We don’t know if Waymo’s unit economics work, if Tesla’s technology will ever work,” he said. Other unknowns include China’s aggressive approach to building low-cost robotaxis and Amazon’s plans for Zoox, its self-driving unit.
The former product executive who believes Waymo is a threat to Uber said the landscape is in so much flux right now that even “if you took Steve Jobs, resurrected him, and made him head of Uber, would he solve the problem? I don’t know.”
Saad, who lives in San Francisco, regularly overhears people discussing the jousting match between Waymo and Uber but believes the title-card combatants should be switched out.
“It’s a mischaracterization to describe this as Waymo versus Uber,” he said. “It’s the Uber network versus Elon Musk.”
Khosrowshahi appears to agree.
“No one wants to compete against Tesla or Elon, if you can help it. …Their capabilities are pretty extraordinary,” the Uber CEO said last month, adding that rather than compete, the companies should find a way to partner, given that 150,000 Uber drivers already use Teslas.
Waymo declined to comment on Tesla’s plans for autonomous vehicles. A spokesperson said the company is in the early stages of commercializing autonomous vehicle technology and is trying various approaches. By partnering with Uber, the spokesperson said, Waymo can offer the Waymo One experience to users more quickly.
On the roads
Rideshare drivers working around the clock in San Francisco hardly have the time to consider the machinations of billion-dollar companies.
Uber said drivers earned more than $33 per hour (before tips) on average in San Francisco during the last quarter of 2024. Lyft pointed to a pay standard it introduced last year stating that drivers will earn 70% or more of rider payments each week, after external fees like taxes and insurance.
“Rideshare is a disaster regardless of Waymo,” said Joseph Augusto, who has been driving for Uber and Lyft for a decade. Augusto said San Francisco has long been a “little Hooverville” of drivers living in their cars because of the large cut Uber and Lyft take from every ride.
Augusto, who drives a Tesla Model X for Uber Black, spends his workdays almost exclusively trekking back and forth to the airport. In contrast to most rides in the city, which are money losers after gas and maintenance, this thin band of business remains profitable.
Augusto’s territory is, for now, protected by a slow-moving regulatory process, and he doesn’t feel particularly threatened by Waymos. Frankly, he believes robotaxis aren’t aggressive enough to navigate the hellish airport environment.
“If I’m stuck there, guess what? I roll down my window and yell, ‘Hey, can I get out of there?’”
Unless Waymos are programmed to be less passive (asshole robotaxi, anyone?) — or the airport opens up routes specifically for autonomous vehicles — Augusto’s job isn’t going anywhere.
“They aren’t a threat to me,” he said. “Yet.”
Last week, the company received a green light to start mapping SFO.