Brandon Donnelly
Daily insights for city builders. Published since 2013 by Toronto-based real estate developer Brandon Donnelly.
Brandon Donnelly
Daily insights for city builders. Published since 2013 by Toronto-based real estate developer Brandon Donnelly.
The autonomous vehicle narrative has historically gone something like this: remove the labor component of rides (i.e. drivers) and rides will become significantly cheaper. Then, people won't need or want to own a car anymore. They'll just Uber or Waymo or whatever around.
But as Waymo provides in and around 250,000 paid trips per week in the 4 cities in which it operates, the opposite has proven to be true — at least so far. A recent report by Obi (an app that aggregates real-time ride pricing) has just revealed the following for San Francisco during the period of March 25 to April 25, 2025:
In other words, Waymo is more expensive than Uber and Lyft, especially for shorter distances. Is this right? Well, Waymo may not have to pay drivers, but they do own and operate their own cars. Uber and Lyft do not. This represents a very different cost structure.
They also have a more inelastic supply base, meaning they have cars whether demand is high or not. Whereas in the case of Uber and Lyft, supply can be variable. That's the idea behind "surge pricing" — to induce more drivers onto the road when it's needed the most.
Fewer Waymos also means that wait times are going to be longer and that their cars are probably spending more time driving around without paying passengers. That's a cost.
Whatever the reasons, lots of people seem to be willing to pay the premium. Part of this almost certainly has to do with the novelty of riding in an autonomous vehicle. I'd pay more if they were in Toronto today. But another reason seems to be that people really appreciate being in the car alone. I guess it's akin to driving your own car.
It, of course, remains to be seen how Waymo's cost structure and pricing model will evolve over time, but I have no doubt that privacy will remain a feature people are willing to pay something for. In the modern world, we are all going to have at least two places of solitude: bathrooms and Waymos.
Cover photo by gibblesmash asdf on Unsplash
Broadly speaking, cities tend to have better data on vehicular traffic than on pedestrian and bicycle traffic. This is because road design has traditionally prioritized the movement of cars, above all else. So it has felt right to bias traffic counts. But there are lots of places where pedestrians and cyclists greatly outnumber vehicles.
For example, I was on Queens Quay yesterday visiting my mom and, if you've ever been to Toronto's waterfront in the summer, you'll know that it has one of the busiest bike lanes/trails in the city — if not the busiest. But if you ask ChatGPT just how busy it is, it will more or less say, "I don't know. Really busy?" And that's because we don't have real-time usage data. We have estimates. And the same is true of pedestrian counts.
(If you're aware of a great dataset, please share it in the comment section below.)
But this is starting to change with the advent of AI traffic monitoring solutions that can handle multi-modal environments. Meaning they're capable of counting everything from pedestrians and scooters to cyclists and trucks. This is what cities need to make better decisions. And as this new tech becomes more widespread, I think it's going to show us just how much we've been missing.
Cover photo by Joshua Chua on Unsplash
Okay, so we know that Paris has transformed itself from a car city into a biking city. Between 2015-2020 the City doubled its number of bike lanes. Then in 2021, it announced that it wanted to become a "100% cycling city" and further add to its bike network. Today, it has one of the busiest bike routes in the world and more people cycle than drive. The Globe and Mail reported, here, that last year 11.2% of trips in Paris proper were made by bike, compared to only 4.3% by car. "You would not be wrong to call it a war on the car," Marcus Gee writes. However, the result was a "victory for the city."
At the same time, we know how people will respond to this data. Lots of people will read this article and immediately say, "yeah, but that's Paris, where the average highs and lows in January are 8 and 3 degrees, respectively. It simply won't work in Toronto where our January highs and lows average 0 and -7 degrees. We have snow to contend with; they don't." And of course, they wouldn't be entirely wrong in this argument. Cycling does tend to decline in the winter months in most cities. (Note: The months of April to November in Toronto are just as warm or warmer than Paris in the winter.)
But just for fun, let's look at Oulu, Finland which has been called the "winter cycling capital of the world." Oulu has a population of around 210,000 people, an extremely low population density of approximately 150 people per km2, and January temperatures that average between -7 and -15 degrees.
The autonomous vehicle narrative has historically gone something like this: remove the labor component of rides (i.e. drivers) and rides will become significantly cheaper. Then, people won't need or want to own a car anymore. They'll just Uber or Waymo or whatever around.
But as Waymo provides in and around 250,000 paid trips per week in the 4 cities in which it operates, the opposite has proven to be true — at least so far. A recent report by Obi (an app that aggregates real-time ride pricing) has just revealed the following for San Francisco during the period of March 25 to April 25, 2025:
In other words, Waymo is more expensive than Uber and Lyft, especially for shorter distances. Is this right? Well, Waymo may not have to pay drivers, but they do own and operate their own cars. Uber and Lyft do not. This represents a very different cost structure.
They also have a more inelastic supply base, meaning they have cars whether demand is high or not. Whereas in the case of Uber and Lyft, supply can be variable. That's the idea behind "surge pricing" — to induce more drivers onto the road when it's needed the most.
Fewer Waymos also means that wait times are going to be longer and that their cars are probably spending more time driving around without paying passengers. That's a cost.
Whatever the reasons, lots of people seem to be willing to pay the premium. Part of this almost certainly has to do with the novelty of riding in an autonomous vehicle. I'd pay more if they were in Toronto today. But another reason seems to be that people really appreciate being in the car alone. I guess it's akin to driving your own car.
It, of course, remains to be seen how Waymo's cost structure and pricing model will evolve over time, but I have no doubt that privacy will remain a feature people are willing to pay something for. In the modern world, we are all going to have at least two places of solitude: bathrooms and Waymos.
Cover photo by gibblesmash asdf on Unsplash
Broadly speaking, cities tend to have better data on vehicular traffic than on pedestrian and bicycle traffic. This is because road design has traditionally prioritized the movement of cars, above all else. So it has felt right to bias traffic counts. But there are lots of places where pedestrians and cyclists greatly outnumber vehicles.
For example, I was on Queens Quay yesterday visiting my mom and, if you've ever been to Toronto's waterfront in the summer, you'll know that it has one of the busiest bike lanes/trails in the city — if not the busiest. But if you ask ChatGPT just how busy it is, it will more or less say, "I don't know. Really busy?" And that's because we don't have real-time usage data. We have estimates. And the same is true of pedestrian counts.
(If you're aware of a great dataset, please share it in the comment section below.)
But this is starting to change with the advent of AI traffic monitoring solutions that can handle multi-modal environments. Meaning they're capable of counting everything from pedestrians and scooters to cyclists and trucks. This is what cities need to make better decisions. And as this new tech becomes more widespread, I think it's going to show us just how much we've been missing.
Cover photo by Joshua Chua on Unsplash
Okay, so we know that Paris has transformed itself from a car city into a biking city. Between 2015-2020 the City doubled its number of bike lanes. Then in 2021, it announced that it wanted to become a "100% cycling city" and further add to its bike network. Today, it has one of the busiest bike routes in the world and more people cycle than drive. The Globe and Mail reported, here, that last year 11.2% of trips in Paris proper were made by bike, compared to only 4.3% by car. "You would not be wrong to call it a war on the car," Marcus Gee writes. However, the result was a "victory for the city."
At the same time, we know how people will respond to this data. Lots of people will read this article and immediately say, "yeah, but that's Paris, where the average highs and lows in January are 8 and 3 degrees, respectively. It simply won't work in Toronto where our January highs and lows average 0 and -7 degrees. We have snow to contend with; they don't." And of course, they wouldn't be entirely wrong in this argument. Cycling does tend to decline in the winter months in most cities. (Note: The months of April to November in Toronto are just as warm or warmer than Paris in the winter.)
But just for fun, let's look at Oulu, Finland which has been called the "winter cycling capital of the world." Oulu has a population of around 210,000 people, an extremely low population density of approximately 150 people per km2, and January temperatures that average between -7 and -15 degrees.
Here's a video comparing winter vs. summer cycling in Oulu.
I'll be the first to admit that I don't generally cycle in January and February. Partially because it's cold and partially because I don't want road salts all over my business. But the correct framing isn't that it's not possible in a city like Toronto; it's that I'm too soft. That, and we need to invest in the right infrastructure if we want more people to do it.
Here's a video comparing winter vs. summer cycling in Oulu.
I'll be the first to admit that I don't generally cycle in January and February. Partially because it's cold and partially because I don't want road salts all over my business. But the correct framing isn't that it's not possible in a city like Toronto; it's that I'm too soft. That, and we need to invest in the right infrastructure if we want more people to do it.
Share Dialog
Share Dialog
Share Dialog
Share Dialog
Share Dialog
Share Dialog
