Uber vs. Car Ownership(blog.samaltman.com) |
Uber vs. Car Ownership(blog.samaltman.com) |
Most people a) don't live in SF b) have your money c) have access to Uber and d) have a Tesla Roadster.
You're essentially talking to yourself and a few choice friends who live in the same bubble as you.
I'm not sure if $5,000/year is right, though. The car cost at least $109,000 new and probably more. At that rate it would take 21 years to fully depreciate, but the average life of a car is more like a decade. It's nonlinear, but the car is at most six years old. I think this may be an underestimate.
While we're at it, I'm struck by the maintenance figure of $3,500/year. That's a big pile of cash! I had a (much more boring, plain old gasoline) car of similar age until last year and the yearly maintenance on it averaged well under $500 and I drove it twice as many miles per year. I don't know enough about the Roadster to say whether that's correct or why it would be that way, but it reinforces that this is a highly unusual calculation whose numbers will apply only to a tiny number of people.
Taking a quick whack at the calculation for myself, I currently do around 8,000 miles/year, insurance is about $1,000, maintenance is $28 (preventive maintenance is bundled in the price of the car for another year, and nothing should break, so I'm just on the hook for the state inspection), depreciation is, somewhat wild guess vaguely informed by my perusal of the used car market last year, maybe $2,000, parking is zero, and my cost per mile is about 8 cents. (Amusing side note: I did that last calculation with Google and it defaulted to giving me the result in dollars per meter.)
Putting it all together, and I have about $3,000 in fixed costs and $640 in per-mile costs, so $3,640/year total for ownership. At $2/mile, UberX would cost me $16,000/year for the same distance driven, about five times the cost.
"just did the math. it's now cheaper for me to get rid of my car and uber everywhere (gas, parking, maintianence, insurance, depreciation) ."
And then he posted the calculation on Twitter.
Uber 9000 miles x $2 = 18000
Personal Car: 7000
Maintenance 2500
- new bumper (old one taken off while street parked) 1000
- replaced foggy headlamps 400. purely cosmetic reasons
- side window smashed 800. never park in oakland =(
- oil changes etc? (200 at most)
Insurance 600 (checkout AAA)
Depreciation 2000(?)
Gas 1800 ($4/gallon @ 20 miles/gallon)
For people who can get to and from work easily using public transportation, perhaps owning a car does not make sense as much as Zipcar/Uber. Owning a car can cost a lot of money, but for people who want one as a form of transportation it can be very affordable (i.e dont buy new). Also, the average car is expected to travel 10-12k miles a year.
Let me repeat that: from purely economical perspective, if driving more than 7000 miles/year, Tesla Model S P85+ is cheaper than Uber.
Here are the numbers. Basic assumptions:
gas : $4/gal (don't know if it's even a valid number anymore)
power: $.1/kWh
depreciation: $1/mile + $300/month (the $300 per months is a sandbagging factor, in reality it doesn't exist)
power: $.033/mile
maint: $600/year covers everything
tires: $1600 (P85+ goes through the full set once a year, depending on how you drive)
registration: $60/year
insurance: $600/year
parking: $300/year (suburbia, I hear you say. yeah, that's right)
---------
7000 miles => $13993
7000 miles => $14000
depreciation: $1500/year
gas: $1470 for the 7K miles
maint: ~$2000/year (it's old)
tires: $1000/year (chews through one set of good, soft tires. I like fast corners)
tabs: $100/year
insurance: $600/year
parking: $300/year
---------
7000 miles => ~$7000The point could be taken further to work out how much people spend on average on a car and at what point they would get rid of their car.
Possibly for general use it is more applicable to self-driving cars than Uber but it's still interesting.
This is a lazy analysis.
Are there folks who will trade car ownership or public transportation for Uber and services like it? Absolutely. But what is the size of that market? How much of that market has Uber already tapped?
Sam's post assumes that Uber's current valuation doesn't take into account this segment of the market and that this segment of the market still offers Uber significant growth potential. Maybe he's right, but he provides absolutely no evidence and doesn't even attempt to come up with hypothetical numbers for consideration.
Literally everyone. Everyone needs to move around their city. The size of the market is the size of the total city scale transportation market which includes car ownership, taxis, public transportation, good walking shoes, pedi cabs......etc.
How much of that market has Uber already tapped?
An insanely small percentage outside of SF and even in SF its small.
One question you didn't ask: Can Uber get even cheaper than it is now? My guess over a 10-15 year time horizon is absolutely yes when you remove the cost of the driver. I think the state of the world that Sam is referring to is when ubers (or services like it) are the predominant way people move about cities. In that world, who is best positioned to win? Uber.
Your approach to market sizing is eerily reminiscent of Ali G's[1]. Amusing, but not credible by any standard.
For example, a 2014 Honda Civic with 1 mile on it (example) is ~$17k, while a 2013 with 6k miles is ~$15k (KBB values). There should be very few things not covered by warranty since they're <3 years and 36k miles.
VW, BMW, and some others have also started including free maintenance for things like oil changes and wipers and brakes as part of the new car warranty.
On the flip side, I'm impressed your insurance is so cheap.
[1] http://www.therichest.com/expensive-lifestyle/entertainment/...
For them, driving is just a means to an end. If there was something better that was cheaper and fulfilled their needs, I think they'd sell their cars in a heartbeat.
That being said, I do love driving, and you're gonna have to pry my car from my cold dead hands :).
It's a calculation about the cost of car ownership vs. taking alternative means of transportation.
I'm simply saying that it's often quite nice to drive, as Sam knows quite well as the owner of a Tesla Roadster.
Tbis service was perfect! Need to go shopping for the weekend? Take the vw golf. Cat needs to go to cat doctor? Take the smart. Want to help a friend with moving? Take the mercedes sprinter. Want to go to the beach with some friends? Take the vw passat.
All in all i paid 20 to 50 euro for _all_ my car needs and i never had to think about refilling gas, switching to winter tires, do checkups and so on.
Unfortunately i had to move to a different town where such a service does not exist and i really really feel limited. I need a car only once or twice a month so buying one is out of question. And renting from a normal car renting company is too expensive.
Being in my early 20s I frequent bars often. I like my car but I can't go to a bar in my car.
Previously, to accommodate my lifestyle I lived in downtown Seattle. This year I moved away, not very far, a 30-40min walk, and am saving about $1000 a month. According to Mint I've only spent $150 (about $6 a ride) on UberX this month.
All in, gas, insurance, car payment and parking spot we were spending over $900 a month on unavoidable monthly costs, at times this number would also go up based on using the car to get around the city and having to pay $10-40 to park the car at wherever we were going.
Add it all up and that same money can buy a lot of Uber/Lyft/Zipcar at 90% of the convenience. I've yet to see how much we'll spend monthly on various transportation services but so far it looks like we'll be saving close to $400/m.
1. People will move back to cities as cities become more attractive (thanks in part to things like the shared economy). There's already some evidence of this trend among millennials. Reverse white-flight if you will.
2. Uber will eventually work just as well in suburban areas. The fundamental economics don't change (higher resource utilization => lower costs). It will take longer to get to penetration levels that result in sufficiently high utilization. But we'll eventually get there.
No, it won't.
> The fundamental economics don't change...
They don't change for Uber, but they do for drivers.
> It will take longer to get to penetration levels that result in sufficiently high utilization.
You seem to be under the belief that supply drives demand here. That's not the case.
The economics of utilizing a taxi service can be markedly different for suburban consumers than urban consumers. Point B is frequently more distant from Point A in suburbia, so the average cost of a taxi ride tends to be higher. Thus the relative attractiveness of paying for a taxi ride decreases.
"as of 2010, more than half of all people live in an urban area. " [1]
[1]: http://www.who.int/gho/urban_health/situation_trends/urban_p...
It's not really that helpful to wonder about the worst scenario, especially without any imagination whatsoever.
For everyone else, Uber is far more expensive than owning their own car.
How is Uber doing in LA, I wonder? Just watching it here on HN, you could be forgiven for thinking it was SF-only (due to the demographics of the site)
That said, there are already taxis and car share programs. Uber may be able to take customers from them but I'm not sure that Uber really provides something compelling enough that it would push a large number of people who are at the margin over the edge. Disclaimer: I've never actually used Uber
With a taxi, unless you're going from a dense main drag (of either people or taxi-needing people, ie. bars) you call for a cab and it may or may not come some time between now and 2 hours from now, depending on if the cab that's supposed to pick you up got flagged down for a better fare.
And then, in most places, the cabbie will tell you he can't take anything but cash. After some grumbling they might let you pay with a credit card, or they might make you take them to a cash machine.
With Uber you say you want a car and it tells you where the car that's coming is and how long it'll take to get to you. And then when you get there you pay by clicking a button.
1) UberX is covered by personal insurance, drives a lot of other people around (hence the cost is shared over many rides).
Your personal car also is covered by personal insurance but drives only "you" around.
Of course, UberX has an insurance advantage!
2) UberX driver is doing it part time, so he/she certainly has a lower paying job (~minimum minimum wage) or no job at all, at which point, the UberX driver's hourly rate is wayyy less than that of a Silicon Valley techie.
3) Since an UberX waiting for a passenger is probably idling somewhere or is driving around the high density of pickups area, the vehicle, technically is not required to pay for parking at any point, but if you driver your own car, you have to leave it somewhere, and pay for the parking (at work, and in SF, at home too).
Bottomline: UberX can be (strike that), will be cheaper than owning and operating own car once Uber + autonomous vehicle becomes a reality.
why?
1) Insurance: Less human error --> Lower premium
2) No need to pay a driver: 50% of UberX price goes to pay the driver (assuming 30% costs, 20% Uber commission). If Uber just keeps that 20% of commission, and costs remain the same, the fares can go down an another 50%!!!!
3) More people moving to autonomous UberX will lead to increase in premiums for manually driven cars (Insurance company has to make profits!).
Until this Autonomous UberX happens, all other explanations of UberX is cheaper than driving, needs to be taken with a pinch of salt
UberX will eventually charge prices more like $0.60 per minute plus $0.20 per mile.
A rough estimate based on the distribution of miles driven is that 30% of Americans should give up their cars under these numbers. Of course, many won't because owning a car isn't purely a financial decision and humans change slowly.
Discussions and imagination move much faster than reality but the world will catch up.
Technical note: The price per hour should be (time-related depreciation)/miles + hourly wage divided by the utilization rate. Mileage should be the sum of fuel costs per mile, maintenance cost per mile, and mileage based depreciation.
Running these numbers for a driver expecting $20 per hour (Average for a Taxi driver is ~$12) with 2/3 utilization and driving a Prius yields a pricing structure of ~$0.60 per minute and ~$0.20 per mile.
I have a Chevrolet cobalt that I bought 5 years ago for $6000. I will have this car until it dies so depreciation isn't really a factor. Instead I just divided $6000 by 5 years. I live in Canada and gas here is the equivalent of about $5.29 per gallon. It's been a good car and has only cost me about $2000 per year in maintenance (I do some minor stuff myself). Factor in registration, license, insurance and at 12000 miles per year at my measured fuel efficiency, I come in at a little less than $0.50 per mile.
If it was going to cost me $2 per mile I think I would have to just stay home.
In conclusion, paying someone to take you in their Toyota Prius is less expensive than paying for your own Tesla.
Still, I see sama's point. I own/maintain my own motorcycle and I uber/lyft when I can't do that. Best of both worlds.
i could definitely get a car that depreciates less. it's old battery tech at this point.
An uberX at $2.00/mile would have cost me $80k.
I could have bought 4 cars and driven them the same mileage for the price of uberX. I've been able to drive this car deep into the cascades, olympic peninsula, etc. for awesome camping trips--something uberX would never let me do. I also never worry about parking on the street because, hey it's a $3000 car who cares?
edit: Car is a '99 Jeep Cherokee if anyone cares, with over 180k miles on it now (and should easily go to 250k+ miles with regular maintenance).
* Or UberX can be lucrative for people who doesn't need 10k miles per year.
This makes a lot of sense. New Yorkers might charge more. I would guess the biggest objection would be safety -- how do you know you're not getting in the car of a kidnapper, thief, rapist, etc.?
On one hand, I think people greatly over-estimate that risk. On the other, many criminals likely would take advantage of the service to find victims.
What do the Kazaks do about that?
That's kind of an aside; I just like old cars and working on them when I don't have to work on them.
Anyhow, the real part I was going to say is that I moved to Atlanta with my wife. She has to drive all over creation because she teaches ballet. We have a new hyundai accent which runs great, but my number one priority was to be close to work so I wouldn't need a car. I live about 1.5 miles from work. I walk to work a lot of days (not so much in the summer). This city sucks for having no car, but if you really plan around, you can. The biggest thing was being willing to live closer to work.
Obviously it's not that easy for people in tons of circumstances that aren't mine, but I wanted to throw my anecdote into the mix fwiw.
I used to commute exclusively by bike in oklahoma. I rode about five or six miles to work and five or six miles back. Had to structure where I looked to live around that as well.
Again, not saying this is viable for everyone, but if you approach it as a lifestyle, it can make a bit more tractable, depending on your context (I have no kids, which makes it a lot easier to)
I didn't do it for the outright savings so much as I hate commuting. I can't stand staring at the road, and eventually I run out of podcasts.
Say you averaged going 45mph. Then your 9000 miles took 200 hours. And you'd spend $4000 extra to use Uber, so not having to drive would cost $20/hour.
Depends how you'd use that time, but for many it would be well worth $20/hour to be able to do other things, like bill hours.
i.e do I do my own laundry, clean my own place, drive my own car, cook my own food?
Te type of customers who would use Uber in LA live in the high-density areas (generally K-town, Hollywood, West LA, and DTLA), and Uber does well in those areas. For anyone living in the low-medium density areas, i.e., the suburbs, Uber simply isn't economical.
I lived in Santa Monica, which wasn't too high density, but I recall that taxi prices were absolutely outrageous. Hopefully Uber has caused those to drop some.
Perhaps it doesn't work "as well", but I live in a very-much suburban area that borders on rural. I know several farms from which where one can consistently get an UberX with a 5-10 minute wait.
For example, the further you travel, the more important cost per mile becomes in deciding car vs. Uber. As with every other own vs. rent discussion, the more you use something, the more compelling ownership is.
I'm sure it can scale up more than now, but how much farther?
When you reduce the relative importance of the per-unit cost and drive up the relative importance of centralized costs, you make a market that has stronger natural tendency to form a monopoly, because there is more of a positive feedback loop reinforcing the position of the leading player.