First ever driverless Tesla robotaxi spotted in Austin, Waymo rides cost 30-40% more than Uber and Lyft, and multiple Waymos destroyed in LA anti-ICE protests
Great stuff as always. Question to your readers: anyone actually have an informed estimate as to what Waymo actually PAID for the Jags? I understand the MSRP may be $75,000, but there is no way in h*** they paid that much for them. Edmunds.com estimates a retail transaction price of $67,000.... but for one car at a time of course. As Waymo must be THE largest fleet buyer of any model of Jaguar anywhere, they must be below that? $60,000? Anyone want to weigh in? (And yes, 60 is still 2x 30, I know: I am just trying to get closer to the actual numbers.)
On a lighter note, as to the Driveway Detection: Rolls Royce already has a "planar" radar that detects potholes and speed bumps ("Musn't jostle the bubbly, eh Felicity?") so maybe it could detect driveways, too. Of course, the minimum length driveway for a Roller owner is probably a mile...
That's fair although there's always thousands of dollars in taxes and registration etc on top of the msrp. But I think my larger point stands that they are much more expensive than the average uberx. So might be more appropriate to compare waymo pricing to Uber black pricing.
I agree 100%: your point absolutely still stands. I was just trying to get at a more accurate number, which would reinforce credibility of the point even further.
Not sure if data is accurate neither representative.
Their data says that avg price of a Waymo is $20.43 and $/km is $11.22.
That gives as 1.82km/trip, which it is 1.13miles/trip.
But if we look at Waymo’s CPUC official data we have that in march they made 708,180 trips and 2,257,393 miles in period 3(revenue miles). When divided we get that average trip is 3.19 miles.
In the methodology section of Obi's report, it says they just looked at Waymo pricing for SF, so I would expect shorter average trips there vs. LA / Phoenix. CPUC data is all of California.
CPUC data is only from CA so its SF + LA. My estimates(similar as yours) are that in SF has 3x trips than LA, so it would be 531k in SF and 177k in LA.
If we take the avg trip distance from Obi in SF, we get that in SF there were 531k * 1.13 = 600k revenue miles and in LA 1656k and it gives an avg distance of (1656/177) 9.3 miles per trip. It seems unreasonable to be 9x more than SF given that the covered area is pretty similar.
There are several concerns with Obi’s methodology:
• They analyzed only 700 simulated trips from April 2025, and only within a selected zone in San Francisco. This is a tiny, non-representative sample that likely doesn’t reflect Waymo’s actual operating patterns across the city.
• The trips were simulated, not real rides. As a result, the data fails to capture real-world rider demand, pickup/dropoff behavior, routing decisions, or traffic conditions.
• They calculated distances using the Haversine formula, which gives straight-line (as-the-crow-flies) distances. This systematically underestimates real-world driving distances in a city. A realistic adjustment factor would be at least 1.5x to account for actual street networks and routing constraints.
For context:
• NYC average trip: ~3.5 miles, ~$8 per mile
• Chicago average trip: ~7 miles, ~$3.20 per mile
Doesn't make sense to be more than 2x expensive than NYC in a per mile $
Thanks for the details. My gut/experience is that Waymo is definitely more expensive than Uber/Lyft, I have been telling people 10-20%, so 30-40% seems reasonable to me. What do you think?
Looking at the actual Obi data, yes it does depend on your assumptions, when/where you call for trips, etc. But regardless, as long as they compared those trips to identical Uber and Lyft trips, the cost of the trip should be fairly accurate. I think they did a pretty good job overall with the report/data though and Obi has experience doing this so am fairly confident in the data / sample size / etc given the constraints.
You are correct that the estimated miles per trip (also not sure why they did km lol) sounds low, but the same method would have been used to calculate distance for U/L trips, so the miles may be off. But the price which is what I care most about, and ETA second, should be accurate.
ETA is also tricky though since Uber and Lyft take time to match with a rider (since they offer the trips to the drivers who will accept the lowest price basically) where as Waymo is instant. Some times they both re-match you with a closer driver. And then sometimes human drivers cancel/go the wrong way/etc vs a Waymo, who despite higher ETAs, probably has a much better actual ETA vs expected ETA score.
Thank you Harry for featuring my take on Wayve x Uber!
Great stuff as always. Question to your readers: anyone actually have an informed estimate as to what Waymo actually PAID for the Jags? I understand the MSRP may be $75,000, but there is no way in h*** they paid that much for them. Edmunds.com estimates a retail transaction price of $67,000.... but for one car at a time of course. As Waymo must be THE largest fleet buyer of any model of Jaguar anywhere, they must be below that? $60,000? Anyone want to weigh in? (And yes, 60 is still 2x 30, I know: I am just trying to get closer to the actual numbers.)
On a lighter note, as to the Driveway Detection: Rolls Royce already has a "planar" radar that detects potholes and speed bumps ("Musn't jostle the bubbly, eh Felicity?") so maybe it could detect driveways, too. Of course, the minimum length driveway for a Roller owner is probably a mile...
That's fair although there's always thousands of dollars in taxes and registration etc on top of the msrp. But I think my larger point stands that they are much more expensive than the average uberx. So might be more appropriate to compare waymo pricing to Uber black pricing.
I agree 100%: your point absolutely still stands. I was just trying to get at a more accurate number, which would reinforce credibility of the point even further.
Reading Obi’s report about Waymo prices.
Not sure if data is accurate neither representative.
Their data says that avg price of a Waymo is $20.43 and $/km is $11.22.
That gives as 1.82km/trip, which it is 1.13miles/trip.
But if we look at Waymo’s CPUC official data we have that in march they made 708,180 trips and 2,257,393 miles in period 3(revenue miles). When divided we get that average trip is 3.19 miles.
That’s why I doubt about Obis data.
what do you think about it?
Thanks!
Where are you getting the # of period 3 miles from?
I see the source for the 708,180 trips in the month of March here - https://lookerstudio.google.com/u/0/reporting/787a90e6-8423-4244-a5d8-2ab1f281c310/page/p_tj01bxgqjd
In the methodology section of Obi's report, it says they just looked at Waymo pricing for SF, so I would expect shorter average trips there vs. LA / Phoenix. CPUC data is all of California.
From CPUC files, the file name is PSG0038152_2025_05_AV_Month-Level_Part0.csv. Here is the link: https://www.cpuc.ca.gov/regulatory-services/licensing/transportation-licensing-and-analysis-branch/autonomous-vehicle-programs/quarterly-reporting
CPUC data is only from CA so its SF + LA. My estimates(similar as yours) are that in SF has 3x trips than LA, so it would be 531k in SF and 177k in LA.
If we take the avg trip distance from Obi in SF, we get that in SF there were 531k * 1.13 = 600k revenue miles and in LA 1656k and it gives an avg distance of (1656/177) 9.3 miles per trip. It seems unreasonable to be 9x more than SF given that the covered area is pretty similar.
There are several concerns with Obi’s methodology:
• They analyzed only 700 simulated trips from April 2025, and only within a selected zone in San Francisco. This is a tiny, non-representative sample that likely doesn’t reflect Waymo’s actual operating patterns across the city.
• The trips were simulated, not real rides. As a result, the data fails to capture real-world rider demand, pickup/dropoff behavior, routing decisions, or traffic conditions.
• They calculated distances using the Haversine formula, which gives straight-line (as-the-crow-flies) distances. This systematically underestimates real-world driving distances in a city. A realistic adjustment factor would be at least 1.5x to account for actual street networks and routing constraints.
For context:
• NYC average trip: ~3.5 miles, ~$8 per mile
• Chicago average trip: ~7 miles, ~$3.20 per mile
Doesn't make sense to be more than 2x expensive than NYC in a per mile $
Thats why i doubt about Obi's report.
What do you think?
Thanks for the details. My gut/experience is that Waymo is definitely more expensive than Uber/Lyft, I have been telling people 10-20%, so 30-40% seems reasonable to me. What do you think?
Looking at the actual Obi data, yes it does depend on your assumptions, when/where you call for trips, etc. But regardless, as long as they compared those trips to identical Uber and Lyft trips, the cost of the trip should be fairly accurate. I think they did a pretty good job overall with the report/data though and Obi has experience doing this so am fairly confident in the data / sample size / etc given the constraints.
You are correct that the estimated miles per trip (also not sure why they did km lol) sounds low, but the same method would have been used to calculate distance for U/L trips, so the miles may be off. But the price which is what I care most about, and ETA second, should be accurate.
ETA is also tricky though since Uber and Lyft take time to match with a rider (since they offer the trips to the drivers who will accept the lowest price basically) where as Waymo is instant. Some times they both re-match you with a closer driver. And then sometimes human drivers cancel/go the wrong way/etc vs a Waymo, who despite higher ETAs, probably has a much better actual ETA vs expected ETA score.
Prices sound pretty high. My studies are no more than $5/6 per mile.