Is Waymo About to Dethrone Uber?

images/Is Waymo About to Dethrone Uber.webp

Creator: Maxinomics · Published: 2024-11-01 · Length: 9:30 · Watch on YouTube

The $25 billion Uber pays it's drivers each year is moving to a new set of companies that have solved autonomous rideshare and are deploying it around the US at an increasingly rapid pace. (video description)

1) Detailed outline

[0:00] The rise of driverless rides

Phil asks what happens when Uber no longer has to coax drivers onto the road. He calls it a $25 billion a year question, because that's how much of the fares riders paid Uber and Lyft in 2023 was passed on to drivers.
Driverless rideshare, specifically Google's Waymo, is here and starting to scale. Last January, Waymo gave fewer than 300 driverless rides a day, all in Phoenix. Now it gives about 15,000 a day across four cities, with Atlanta coming online in early 2025.
With new cities and another 1,000 cars under construction, Phil expects more than 40,000 rides a day by this time next year.

[0:27] Waymo's technology and growth

Phil greets a Waymo as it pulls up ("Hello, Waymo." "Hi, Phil.").
The 1,000 new cars come out of 15 years of research that began in 2009 as the Google self-driving car project. It went through several hardware generations on a few vehicle models before settling on the electric Jaguar I-Pace.
Each car carries radar, lidar, 29 cameras, and other sensors, including ones that hear emergency-vehicle sirens.

[1:22] The rider experience

Phil says the hype isn't a gimmick: riders are switching to a better product.
A Waymo removes the usual uncertainties: what car will show up, whether the driver will cancel, what music will be on, whether it's too hot or too cold, and whether there's room for your legs and bags. It's also safe.
It's private. The ride is entirely yours, so you can say anything to anyone at any volume and look however you like. A rider in a clip says she's coming home from her boyfriend's house in Barbie pajamas with no driver to judge her.

[2:06] The $25 billion opportunity

The technology works and will improve, so the roughly $25 billion paid to drivers will move to companies that run or build driverless vehicles.
The estimate comes from the City of Chicago's complete last 12 months of rideshare data, covering every trip in the third-largest US city: 136 million rides with pickup and drop-off locations, fare, tip, mileage, and duration. It's about 40 GB and took 15 hours to load into a database.
Some findings: 26% of riders tip their drivers, and the average tip is $3.
Chicago is representative enough of a major US city that Phil scales it up. Drivers in the 30 largest US cities account for about $25 billion a year of the Uber fares riders pay.

[2:53] Uber and Waymo's partnership

Uber stands right in the path of that money. In a surprise move last month, it signed an exclusive agreement with Waymo: in Austin and Atlanta, you can order a Waymo only through the Uber app.
Uber has historically been asset-light. It runs an app and tries to own no cars or maintenance shops, just offices full of developers. The bet was that matching riders with drivers and taking a cut was more profitable than owning cars, and drivers could worry about the vehicles. That has now changed, for good reason.
The unit economics: riders pay Uber about $3.20 a mile, about $2.20 goes to the driver and their vehicle, and Uber keeps about $1. Uber has no reason to lower that price even if its costs fall, so whatever it can shave off the $2.20, it keeps.
The high-end estimate of a Waymo's all-in operating cost is about $1.50 a mile, covering insurance, depreciation, licensing, maintenance, and energy. Uber could immediately pocket the roughly 70-cent difference.
That's enticing enough that the deal has Uber managing physical things: whole fleets of thousands of driverless cars, plus the cleaning, maintenance, storage, and parking lots for charging them.

[4:23] Designing the ride experience

The 70 cents a mile is only the start. Working with one big company like Google, Uber can design the entire door-to-door experience.
Phil's analogy is air travel. Half of it is the flight, and the other half is the airport: lines to check bags, worrying about making your gate, bad food, uncomfortable chairs. About the only upside is that a 9 a.m. beer is acceptable.
If Delta could build airports just for its passengers, they'd look very different: elegant drop-off, a security line that isn't like checking into prison (if there's one at all), great chairs and lighting, and food like Chick-fil-A, Sweetgreen, kebabs, and pizza. It would win loyal customers and could sell them things travelers want.
Uber controls only half the experience today. With millions of drivers and vehicles it can't unify what happens once you're in the car, and that's a big opportunity.
With one partner like Waymo, Uber could change the cabin layout, the feel, what you see, and what you can buy. Seats needn't all face forward, and there could be food and drinks, like a KIND bar, a Diet Coke, or a hot coffee on the way to the airport.
Uber could also offer a cheaper subscription with in-car ads turned on. Google is an advertising company, and a captive rider would appeal to local businesses near pickup and drop-off points.

[6:20] Logistics and challenges

First they need cars. Only about 700 Waymos are operating now, and riders say waits are too long because there aren't enough to cover the cities they serve.
To size the problem, Phil asks for Uber's busiest day of the year in Chicago. It's New Year's Eve (the slowest is Christmas). The peak hour that night had 28,359 rides averaging 23 minutes, and back-of-napkin math says about 20,000 cars could cover them within the city.
A typical weekly peak, like a Saturday night, has about 17,000 rides an hour and needs roughly half that, about 10,000 cars.
At about $120,000 per car with its custom hardware, covering one major city costs billions. Covering the 30 largest cities would take at least 300,000 Waymos and tens of billions of dollars.
That cost, plus what will be slow city-by-city regulatory approval, is why Uber is set to stay at the center of rideshare.
The 1,000 Waymos being built at a Magna factory (Phil says in Canada) will join the 700 on the road across the five cities where Waymo operates. They'll absorb part of current demand, and human drivers will fill the rest.

That's why Waymo is partnering with Uber instead of building its own network. Riders want one app, not several, and Uber has almost three-quarters of the rideshare market, so it can offer options at many price points while the driverless fleet grows.
Waymo also gets Uber's routing algorithms, which Phil calls second to none at predicting where riders want pickups and drop-offs, and Uber's dynamic pricing. Waymo needs its cars routed efficiently and priced competitively against every ride requested on Uber.
That matters more as Waymo grows, because Uber is likely paying Waymo a per-mile fee for each ride, and rideshare racks up a huge number of miles each year.
If Waymo made 15 cents a mile and covered half of all rideshare rides in the 30 largest cities, its profit after vehicle and hardware costs would be about $4 billion a year. The auto-captions have Phil comparing that to "20%" of Google's current profit, but that figure doesn't match Alphabet's actual earnings, so treat the comparison loosely.

[8:55] Competition and next steps

Footnote one: there is competition. Phil says Tesla FSD (Full Self-Driving) is very good despite what the internet says. Zoox and Nuro are also in the race. But Waymo leads, because it has paying customers, an Uber integration, and real rides.
Footnote two: all five Waymo cities are Sun Belt cities with mild weather. To judge how fast robotaxis will spread, watch where Waymo's next city is. If it has snow and ice, like Boston, New York, or Philadelphia, that says a lot about Waymo's confidence in its software.

2) Things mentioned

Economics, markets, and unit costs

Driver pay pool: about $25 billion a year paid to rideshare drivers in the 30 largest US cities (2023), extrapolated from Chicago data.
Uber fare split: riders pay about $3.20 a mile, about $2.20 goes to the driver and vehicle, and Uber keeps about $1.
Waymo operating cost: at most about $1.50 a mile all-in (insurance, depreciation, licensing, maintenance, energy).
Margin opportunity: about 70 cents a mile that Uber could keep by replacing drivers.
Pricing power: Uber has no reason to lower fares when its costs fall.
Asset-light vs. asset-heavy: Uber's shift from an app-only marketplace to managing physical fleets.
Vehicle cost: about $120,000 per Waymo with custom hardware.
Fleet sizing: about 20,000 cars for Chicago's New Year's Eve peak hour (28,359 rides), about 10,000 for a typical Saturday-night peak (about 17,000 rides), and at least 300,000 for the 30 largest cities.
Hypothetical Waymo profit: about $4 billion a year at 15 cents a mile and 50% share of rideshare rides in the 30 largest cities.
Market share: Uber has almost three-quarters of the US rideshare market.
Tipping: 26% of Chicago riders tip, and the average tip is $3.
Dynamic pricing and routing: Uber assets Waymo gains through the partnership.
New revenue ideas: in-car food and drinks, ad-supported cheaper subscriptions, and ads from local businesses near pickup and drop-off points.

Policy and regulation

City-by-city regulatory approval: a slow process that limits how fast robotaxis can spread.
City of Chicago rideshare open data: a complete public record of trips (fares, tips, locations, mileage, duration).
Weather as a deployment test: all current Waymo cities are in the Sun Belt; Boston, New York, and Philadelphia are the snow-and-ice tests to watch.

Companies and organizations

Waymo: Google/Alphabet's driverless rideshare company, the video's subject.
Google / Alphabet: Waymo's parent, called "an advertising company."
Uber: rideshare leader and Waymo's exclusive app partner in Austin and Atlanta.
Lyft: the other major US rideshare app, part of the $25 billion driver-pay figure.
Magna: the auto manufacturer building the 1,000 new Waymos.
Jaguar: maker of the I-Pace that Waymo uses.
Tesla: FSD as a competitor.
Zoox (Amazon) and Nuro: other autonomous-vehicle competitors.
Delta Air Lines: the hypothetical private-airport analogy.
Chick-fil-A, Sweetgreen, KIND, and Diet Coke (Coca-Cola): examples of food and drink in the airport and in-car analogies.

Products and technology

Google self-driving car project (2009): Waymo's origin.
Jaguar I-Pace: the electric SUV Waymo settled on.
Waymo sensor suite: radar, lidar, 29 cameras, and audio sensors that detect emergency sirens.
Uber app: the only way to order a Waymo in Austin and Atlanta.
Tesla FSD (Full Self-Driving): which Phil calls "very good."

Places

Phoenix: Waymo's first market (fewer than 300 driverless rides a day last January).
Austin and Atlanta: Uber-exclusive Waymo cities, with Atlanta launching in early 2025.
Chicago: the third-largest US city and the source of the 136-million-ride dataset.
Canada: where Phil says the Magna factory is.
Sun Belt cities: where all current Waymo operations are.
Boston, New York, and Philadelphia: snowy cities to watch as signs of Waymo's confidence.

Waymo growth figures cited

Under 300 driverless rides a day in January (Phoenix only), now about 15,000 a day across four cities, with more than 40,000 a day projected a year out.
About 700 Waymos operating now, with 1,000 more being built.

3) Biographies

Phil Andrews (presenter)

Phil Andrews is the producer and on-camera host of Maxinomics ("the guy in the videos and the comments"), a Morning Brew channel that explains economics, business, and geopolitics through stories. Here he loads a year of Chicago rideshare trips into a database to size the driver-pay pool that robotaxis could capture.

Waymo rider (appears in a clip, not named)

A rider in a social-media clip who explains the privacy appeal of driverless rides: she's heading home from her boyfriend's house in Barbie pajamas with no driver to judge her.

Waymo (subject company)

Alphabet's autonomous-driving company. It began in 2009 as the Google self-driving car project and became Waymo in 2016. It runs fully driverless paid rides in several Sun Belt metros using Jaguar I-Pace vehicles fitted with its own sensors.

Uber (subject company)

The largest US rideshare platform, founded in 2009 and long run as an asset-light marketplace that matches riders with independent drivers. The video argues its Waymo partnership marks a move toward managing physical autonomous fleets.

Google self-driving car project (context, not named in the video)

The 2009 Google X effort, led at first by Sebastian Thrun, that became Waymo.

Production credits

Producer Phil Andrews; video editor Greg Jacobs; motion graphics by Seth Laupus; camera operator Shawn O'Connor; director of production services Sam Wolf; thumbnail by Tom Grillo; supervising producer James Atamian; chief content officer Devin Emery.