Waymo and DoorDash Bring AV Delivery to Phoenix
Wayve in talks to raise $2bn from Microsoft and Softbank, Waymo to launch in London in 2026, and impact of AVs on Uber driver earnings.
Top Stories of the Week
DoorDash to use Waymo’s robotaxis for delivery in Phoenix (link). The pilot will start with DashMart convenience and grocery deliveries, where customers might soon see a Waymo pull up instead of a human dasher. Once the order arrives, the customer will use the DoorDash app to pop the trunk and grab the order — no driver, no doorstep drop-off. This comes after a previous pilot between Waymo and Uber Eats last year1.
I signed up as a Dasher 9 years ago when the company was still doing in person orientations and I remember a Dasher asking the question, ‘Do we have to take the order to the customer’s door?’ and the GM responded, ‘Dude, it’s in the name!’. Lol. In all seriousness though, it is nice having your order dropped off at your door and more importantly, not having to pay attention to the exact time the order will be dropped off. Especially if you live in a condo complex or a high-rise building. But there’s no need to tip a Waymo, or a sidewalk delivery robot, so I imagine the cost savings may outweigh the added inconvenience for some.
This may not be the best use case for a Waymo vehicle since it’s a $75,000 car and weighs 5,000 pounds. But rideshare demand is at its highest during weekends, rush hours and evening hours, so there may be off peak times where you can increase Waymo’s utilization with this type of partnership - since people need grocery and convenience items delivered 24/7. And remember, even though DoorDash is best known for food delivery, this pilot will be focused on grocery and convenience items to start. From Waymo’s blog post:
We’ll begin by delivering orders from DashMart … with plans to include more local Phoenix merchants and a wider variety of offerings over time.
In theory, combining food delivery and rideshare has always sounded great, but in reality, there are a lot of issues. Lunch is really the only peak food delivery time where rideshare demand is low, and food tends to stink up your car. Not to mention the fact that I know I would pay a lot more for a ride for myself, than I would for my burrito. So I would be surprised if Waymo’s vehicles are ever used for this purpose.
Waymo plans robotaxi launch in London, marking its European debut (link). This will be Waymo’s second international city but first true international launch, since Tokyo is only considered a road trip. The company will start testing with safety drivers behind the wheel before rolling out fully next year, once regulators give the green light. Its London fleet will use Jaguar I-PACE EVs powered by the Waymo Driver system, with Moove handling day-to-day operations and maintenance.
Moove already operates in London as an Uber fleet partner, supplying EV rentals to drivers. But this expansion marks Waymo’s seventh market without Uber — a shift that’s increasingly looking like a deliberate strategy.
London is interesting to me because of the potential pushback from labor supporters. Most taxi and Uber drivers in London are full-time, career drivers, who have invested heavily in their vehicles and business. They don’t take kindly to new supply, especially when that supply is of the robot form.
Announcements/Partnerships/Fundraising
UK’s Wayve in talks with SoftBank and Microsoft for $2bn fundraise (link, no paywall).
Coco Robotics taps UCLA professor Bolei Zhou to lead a new physical AI research lab (link). Related: I had a blast interviewing Bolei Zhou at the ‘AI, Autonomy, and Future of Delivery Mixer’ with Coco Robotics for LA Tech Week (link). Big thanks to everyone who came out and said hi!
Terawatt and Verkada security partnership (link).
Stellantis, China’s Pony AI to Test Autonomous Vans in Europe (link, no paywall).
Starship raises $50M to scale delivery bots (link).
Other Stuff
‘Waymo wait times are consistently better than Robotaxi right now’ (link). Waymo is only available ‘on Uber’ in Austin, so I would imagine that Uber does not want to present longer-than-average Uber wait times — probably around 3 to 8 minutes in most cases. And that’s one of the big benefits of being on Uber and having additional human driver supply. You can ensure it’s a great Waymo experience when the ride matches Waymo’s ideal operating domain.
Check out AVmap.io, a site that tracks autonomous vehicle deployments offering public rides in the United States. You can see where services operate today, which cities are running pilots, and how coverage has expanded since Waymo’s 2017 Phoenix launch (link). Nice work
.New tariff might add another $50k to Waymo Zeekrs (link). Vehicle supply seems to be the real limiting factor to Waymo’s growth at the moment. The Jaguar I-Pace is out of production and the mounting tariffs on Chinese EVs are bad news for the Zeekr. Waymo doesn’t have any other vehicles production ready so I’m not sure what the plan is here.
Regulating Robotaxis (link).
On Sunday mornings, Waymo now pops to #7 in travel in the App Store. This is for an app that’s only available in SF/LA/PHX (ATL/ATX are Uber). (link).
Waymo poaches top Tesla audio engineer (link).
I’m glad that Uber is finally starting to talk about the impact of AVs on drivers but it seems like they don’t want to admit the elephant in the room: more supply has always meant lower hourly earnings for drivers. And saying ‘drivers are busier’ could actually be a bad thing if they’re doing more trips and getting paid less per trip (more trips also means more miles and expenses, which drivers are responsible for). Gridwise data showed that drivers were busier in AV cities, yet they were earning less than the national average, thus confirming this theory. Andrew also states that in Austin, ‘driver earnings per hour are out-pacing the rest of the US’ so that would imply that they are not in Atlanta and Phoenix? Ultimately, I need to see the actual data before taking Uber’s side on this one.
WeRide Founder & CEO Dr. Tony Han delivered the opening keynote at Mobility Live AU on how AVs are redefining transport, and joined a panel exploring why Australia is a promising AV market (link).
Former Waymo director on Waymo long-term licensing ambitions, and small fleet strategy (link).
What else we’re reading/listening to
Alternative Data Weekly by
: Alternative Data Weekly #203 (link). on AVs, Congestion, and the Future of Cities (link). My latest podcast episode with David Zipper, Senior Fellow at the MIT Mobility Initiative, explores how autonomous vehicles could reshape cities — from rising congestion and safety risks to their impact on public transit. We discuss how cheaper AV rides might worsen traffic, what policies could help, and how governments and AV companies can work together to build smarter mobility systems.Autonocast by
, , and Kirsten Korosec: #347: Waymo DDOS Attack, Tesla Door Handles, Wayve, 2026 CES Party Info (link).Driven by Data: A Conversation with Nexar’s Jon Miller (link).
Cool Rides
‘Thoughts after my first Waymo ride: It drives smoother than ~80% of drivers I rode with, the “you know exactly what to expect” aspect (no driver who might or might chat me up, or be on the phone, or be a bad driver) is underrated. Waymo is the best ad for Jaguar’ (link). I liked
’s third point but it is ironic that you can’t even buy a Jaguar I-Pace anymore. It would be smart for an OEM to partner with Waymo just for the brand exposure.‘Tonight I rode in a Waymo. No driver. No small talk. No scanning the situation for safety. Just quiet’ (link).
‘I can safely say that Zoox was one of my favorite parts of my Vegas trip! The music selection is AWESOME too!!’ (link).
‘Just had my mind-blown moment of the year: my first ride in a Waymo’ (link).
AVs/Humans behaving badly
‘The plan? At dusk, 50 people went to San Francisco’s longest dead-end street and all ordered a Waymo at the same time. The world’s first: WAYMO DDOS’ (link). My friend
wants to ban these Waymo riders but I thought this was pretty funny. Who’s side are you on?‘BAD BEHAVIOR: San Francisco residents have started leaving weed pens in Waymos. No respect for the commons’ (link).
‘I don’t think they [Waymos] can read’ (link).
LA Waymo in a hurry (link). Nice move by this Waymo!
Neat Jobs
Staff Machine Learning Engineer, ML Infrastructure (Predictive Planner) at Waymo (link) via Brian Wilt.
Sr. Process Engineer, Production Control, Megafactory at Tesla (link) via Nisarg Prajapati.
Market Research Lead at Waymo (link) via Megan Neese.
Engineering Technician at Wayve (link) via Brandon Diaz.
Senior Manager, HRIS at Zoox (link) via Zoox.
Until next week.
-Harry
Although I don’t recall ever seeing anyone actually get an order delivered from this partnership :)
The DashMart focus for this pilot is smart - groceries and convenience items have more forgiving delivery windows than hot food, which buys time for customers to retrieve from the trunk. Your observation about the $75k vehicle weight vs. delivery economics is interesting, but I wonder if the real value prop is operational learaning at scale. Each delivery teaches the Waymo Driver about different pickup/drop-off scenarios in a lower-stakes environment than passenger transport. Plus, as you mentioned, off-peak utilization could meaningfully improve the unit economics of these expensive assets. The London expansion shows they're confident in the tech, but I agree labor pushback will be significant there.
Thanks for writing this, it clarifies a lot. I'm wonderin if the increased utilization for a 5,000-pound vehicle truly offsets its ecological footpritn for small deliveries. Food for thought.