6 ms·

I think Uber's business model relies on self-driving cars. They're paying subsidies now to build a brand -- the name "Uber" has become synonymous with "pull out

by exelius 10y ago

I think Uber's business model relies on self-driving cars. They're paying subsidies now to build a brand -- the name "Uber" has become synonymous with "pull out your cell phone and call for a ride" -- that will be very valuable once self-driving taxis become a commodity (which will happen shortly after their introduction). Uber is not a bet on who can build the most profitable taxi company now -- it's a bet on a brand in an industry that will rapidly commoditize. Given that literally everything else Uber does has been replicated by at least one team at every hackathon I've been to in the last decade, there is very little sustained advantage from technology. The auto industry is at a crossroads: you have new upstarts like Tesla that are very obviously planning to convert to a transit-as-a-service model. The "old" auto industry (basically everyone that makes cars and is not Tesla) is still struggling to adapt to a more "continuous development" model like Tesla. Tesla's engineering process is far simpler -- an electric car replaces the complex internal combustion drivetrain (an engine block that requires separate air, water, oil and gasoline systems, plus the transmission) with a far simpler electric engine. The electric system in a Tesla is actually far simpler than your average car: the sensor package in a modern internal combustion engine is an incredibly complex piece of engineering. This gives them a huge cost advantage over existing automakers -- if Tesla is providing transit as a service directly to customers AND making/maintaining the vehicles themselves, that displaces a lot of revenue (auto sales/maintenance to companies like Uber). I think Uber will eventually merge with an auto manufacturer (and likely keep the Uber branding since it's likely to be the most valuable part of the company). They already have realized that Tesla is their biggest competition; and I think that the autonomous driving deal with Ford is simply testing the waters for a future acquisition. 15 years from now, most people will likely have 3 or 4 choices of how to get somewhere by car: Uber, Tesla, Lyft and likely a mix of local / regional companies. Brand value is powerful; and I guarantee you that at the end of this, the Uber brand will be worth more than what they've put into it.

komali2 10y ago
Agreed, and the author even sets us up for this - again and again he mentions how the largest sink is the drivers, the largest cost is the drivers, the biggest blocker is the drivers. Remove the drivers from the equation and things suddenly become a lot simpler. Imagine if instead of becoming an uber driver, you retrofit your car with an Uber Automated Car Taxi Converter(c), "guaranteed to put your car to work for you!"
- amyjess 10y ago
  After having two nightmare drivers this week where I actually had to end the ride early and get out of the car, I can't wait for self-driving cars, and whatever service that rolls out self-driving cars first is going to get all of my business.
  - exelius 10y ago
    Yeah; that's ultimately why Uber treats driver quality as a short term problem and they don't do a lot more pre-employment checks... it's a problem they'll fix by removing the drivers entirely; so better to sweep it under the rug and by the time they get in trouble, it will be too late...
- exelius 10y ago
  > Imagine if instead of becoming an uber driver, you retrofit your car with an Uber Automated Car Taxi Converter(c), "guaranteed to put your car to work for you!" No way; see my comments above about the differences in car design. Combining electric vehicles and autonomous driving is a HUGE combination that trades mechanical complexity (difficult to diagnose, time intensive, permanent fixes are extremely costly) for software complexity (less deterministic but can easily be patched). This greatly reduces both the initial cost and the maintenance costs of electric vehicles (really, the only thing you have to swap out are the tires). Reduction in fleet maintenance costs alone is huge. My theory is that the Tesla Model S only costs as much as it does because they're using it as a premium option to fund R&D of the self-driving auto platform. I'd bet if we saw the total marginal cost for a Model S it would be SIGNIFICANTLY lower than a comparable internal combustion / hybrid automobile. Your current car is also designed at least partially for driver comfort. If you no longer need a driver, then you no longer need a steering wheel, or control of any kind really... and you can offer a completely different transportation experience. Do you even need a windshield? What happens to the interior of the auto as VR/AR gain popularity? But I'm pretty sure whatever the future of the car is, your current auto won't be along for the ride. I'm not even sure that individual auto ownership will make sense anymore -- companies with large fleets will almost certainly be able to offer various levels of guaranteed pickup times. If you're wealthy and want your own car, you would just pay more for a "dedicated" car instead of a "shared" one.
CardenB 10y ago
Uber does not have an autonomous driving deal with ford.
fencepost 10y ago
The problem with Uber paying subsidies now to build a brand is that for people paying attention to them their actions independent of the drivers are building even more of one. Living in an area covered by both Lyft and Uber (and probably with many/most drivers working both), I will pretty much exclusively use Lyft - and I pay more for it as well, as I've tipped the drivers on every one of my rare trips. In the long term I expect their prices to rise to the point where it's more comparable to taxi services, but they still have some advantages there at least to my mind.
aetherson 10y ago
Guys, if you're paying $2B/year to "build a brand," you should maybe think about exactly how valuable that brand will be.
- exelius 10y ago
  So I do this for a living... and I'd estimate the value of Uber's brand at $20-30 billion right now and growing. Brand perception among mainstream consumers is extremely positive and they're aggressively expanding globally. Think about how many times you've heard someone say "Hey, who wants to call the Uber?" That's brand value right there. They are essentially price inelastic given how many tiers of service they offer. Uber is only having a cash bonfire right now to keep the drivers happy so they don't make it a terrible experience for the customers. But rest assured as soon as they are allowed to, Uber will replace the drivers, lower the price and capture some percentage of what they used to have to pay drivers for themselves. That's when Uber goes from a brand investment to a cash cow.
MarkMc 10y ago
Self-driving cars will remove Uber's one significant competitive advantage: The network effects of having lots of drivers. I very much doubt Uber's brand will be a significant asset in 10 years. I cannot see any reason why the economics of self-driving cabs would be any different to airlines.
- exelius 10y ago
  The airline industry is much harder to differentiate because it is so capital intensive. Airplanes are super expensive because failure is so catastrophic -- thus they are built with multiple levels of redundancy (and production / maintenance costs increase exponentially with the complexity). Elon Musk has spoken about this exact situation in relation to SpaceX: basically NASA builds $5 billion rockets with a 1% failure rate, and SpaceX builds $30 million rockets with a 5% failure rate -- but the idea is that because SpaceX's solution is so much cheaper, they will have more launches and thus will have more failures. Cars are much less dangerous. In a failure scenario, a car can easily be designed with a simple mechanical failsafe to disable the vehicle and wait for help. The lower cost of failure in an automobile entirely changes the business model: it allows you to apply modern product design processes (agile design, rapid feedback, tight integration of R&D and operations, etc.) because any individual failure of the product isn't going to cause headlines. Anyway, that's the difference: a catastrophic failure in an airplane is $200-500 million in financial losses ($150-300 million for the airframe, and the bad press/settlement from killing ~200 people). A catastrophic failure in a car is the loss of a $60,000 vehicle and maybe death payouts of under $1 million. The only way we'll be able to ethically apply these design principles to flight is drones / UAVs -- which is exactly why they will quickly come to dominate air freight once someone figures out a viable model. Anyway, due to all of this, automated taxi companies won't be taking on nearly the level of risk that an airline does. IMO there's also more room for differentiation for "specialty" taxis: let's say you and a group of coworkers need to go to another office an hour and a half away. Why not rent a "meeting" car with Internet access, whiteboards and tables so you can knock out a meeting or two on the way? Or are you hungry on the way somewhere? Order a food car and have your meal waiting for you when it picks you up. Are you tight on cash? Order a 10-passenger bus that will pick you up on the way. Because the risk and cost are lower, there will be a lot more variability in cost and thus in business model. IMO Uber's brand value is in owning the customer billing relationship -- that is ENORMOUS in the consumer product world. They've reduced transactional friction to almost nothing (THAT is Uber's true brand value -- even if consumers don't consciously realize it), so it's going to be very hard to displace them.

by exelius 10y ago

komali2 10y ago

Agreed, and the author even sets us up for this - again and again he mentions how the largest sink is the drivers, the largest cost is the drivers, the biggest blocker is the drivers. Remove the drivers from the equation and things suddenly become a lot simpler. Imagine if instead of becoming an uber driver, you retrofit your car with an Uber Automated Car Taxi Converter(c), "guaranteed to put your car to work for you!"

amyjess 10y ago
After having two nightmare drivers this week where I actually had to end the ride early and get out of the car, I can't wait for self-driving cars, and whatever service that rolls out self-driving cars first is going to get all of my business.
- exelius 10y ago
  Yeah; that's ultimately why Uber treats driver quality as a short term problem and they don't do a lot more pre-employment checks... it's a problem they'll fix by removing the drivers entirely; so better to sweep it under the rug and by the time they get in trouble, it will be too late...
exelius 10y ago
> Imagine if instead of becoming an uber driver, you retrofit your car with an Uber Automated Car Taxi Converter(c), "guaranteed to put your car to work for you!" No way; see my comments above about the differences in car design. Combining electric vehicles and autonomous driving is a HUGE combination that trades mechanical complexity (difficult to diagnose, time intensive, permanent fixes are extremely costly) for software complexity (less deterministic but can easily be patched). This greatly reduces both the initial cost and the maintenance costs of electric vehicles (really, the only thing you have to swap out are the tires). Reduction in fleet maintenance costs alone is huge. My theory is that the Tesla Model S only costs as much as it does because they're using it as a premium option to fund R&D of the self-driving auto platform. I'd bet if we saw the total marginal cost for a Model S it would be SIGNIFICANTLY lower than a comparable internal combustion / hybrid automobile. Your current car is also designed at least partially for driver comfort. If you no longer need a driver, then you no longer need a steering wheel, or control of any kind really... and you can offer a completely different transportation experience. Do you even need a windshield? What happens to the interior of the auto as VR/AR gain popularity? But I'm pretty sure whatever the future of the car is, your current auto won't be along for the ride. I'm not even sure that individual auto ownership will make sense anymore -- companies with large fleets will almost certainly be able to offer various levels of guaranteed pickup times. If you're wealthy and want your own car, you would just pay more for a "dedicated" car instead of a "shared" one.

CardenB 10y ago

Uber does not have an autonomous driving deal with ford.

fencepost 10y ago

The problem with Uber paying subsidies now to build a brand is that for people paying attention to them their actions independent of the drivers are building even more of one. Living in an area covered by both Lyft and Uber (and probably with many/most drivers working both), I will pretty much exclusively use Lyft - and I pay more for it as well, as I've tipped the drivers on every one of my rare trips. In the long term I expect their prices to rise to the point where it's more comparable to taxi services, but they still have some advantages there at least to my mind.

aetherson 10y ago

Guys, if you're paying $2B/year to "build a brand," you should maybe think about exactly how valuable that brand will be.

exelius 10y ago
So I do this for a living... and I'd estimate the value of Uber's brand at $20-30 billion right now and growing. Brand perception among mainstream consumers is extremely positive and they're aggressively expanding globally. Think about how many times you've heard someone say "Hey, who wants to call the Uber?" That's brand value right there. They are essentially price inelastic given how many tiers of service they offer. Uber is only having a cash bonfire right now to keep the drivers happy so they don't make it a terrible experience for the customers. But rest assured as soon as they are allowed to, Uber will replace the drivers, lower the price and capture some percentage of what they used to have to pay drivers for themselves. That's when Uber goes from a brand investment to a cash cow.

MarkMc 10y ago

Self-driving cars will remove Uber's one significant competitive advantage: The network effects of having lots of drivers. I very much doubt Uber's brand will be a significant asset in 10 years. I cannot see any reason why the economics of self-driving cabs would be any different to airlines.

exelius 10y ago
The airline industry is much harder to differentiate because it is so capital intensive. Airplanes are super expensive because failure is so catastrophic -- thus they are built with multiple levels of redundancy (and production / maintenance costs increase exponentially with the complexity). Elon Musk has spoken about this exact situation in relation to SpaceX: basically NASA builds $5 billion rockets with a 1% failure rate, and SpaceX builds $30 million rockets with a 5% failure rate -- but the idea is that because SpaceX's solution is so much cheaper, they will have more launches and thus will have more failures. Cars are much less dangerous. In a failure scenario, a car can easily be designed with a simple mechanical failsafe to disable the vehicle and wait for help. The lower cost of failure in an automobile entirely changes the business model: it allows you to apply modern product design processes (agile design, rapid feedback, tight integration of R&D and operations, etc.) because any individual failure of the product isn't going to cause headlines. Anyway, that's the difference: a catastrophic failure in an airplane is $200-500 million in financial losses ($150-300 million for the airframe, and the bad press/settlement from killing ~200 people). A catastrophic failure in a car is the loss of a $60,000 vehicle and maybe death payouts of under $1 million. The only way we'll be able to ethically apply these design principles to flight is drones / UAVs -- which is exactly why they will quickly come to dominate air freight once someone figures out a viable model. Anyway, due to all of this, automated taxi companies won't be taking on nearly the level of risk that an airline does. IMO there's also more room for differentiation for "specialty" taxis: let's say you and a group of coworkers need to go to another office an hour and a half away. Why not rent a "meeting" car with Internet access, whiteboards and tables so you can knock out a meeting or two on the way? Or are you hungry on the way somewhere? Order a food car and have your meal waiting for you when it picks you up. Are you tight on cash? Order a 10-passenger bus that will pick you up on the way. Because the risk and cost are lower, there will be a lot more variability in cost and thus in business model. IMO Uber's brand value is in owning the customer billing relationship -- that is ENORMOUS in the consumer product world. They've reduced transactional friction to almost nothing (THAT is Uber's true brand value -- even if consumers don't consciously realize it), so it's going to be very hard to displace them.