True price of an Uber ride in question as investors assess firm's value
reuters.com
reuters.com
In other words, does Uber have a viable business model?
Some analysts think the answer is no.[1] They argue that Uber will never be able to charge prices in excess of the level required by a traditional taxi/limo operator to be profitable, for long enough to recoup the billions of dollars the company has burned by subsidizing the price of rides. Put another way, even if Uber drives every competing taxi/limo operator out of business, as soon as the company raises prices to cover costs -- vehicle, insurance, maintenance, fuel, credit card processing, license fees, and (at least for now) labor -- new entrants (and revived old entrants) would quickly come into the market, preventing Uber from keeping prices up.
[1] For example, see http://www.nakedcapitalism.com/2017/06/wheels-come-off-uber....
if they basically 'rent' autonomous vehicles from people then so can any other competitor, if they have their own autonomous tech then they have to pay for upkeep of vehicles which may negate the difference and even then only they have an advantage for the theoretical slice of time between when they have the tech and others don't
Autonomous cars will be a commodity, I think it's unreasonable for Uber to point to that as their competitive advantage.
Who actually has continually refreshed map and traffic data from users? Google, Apple, Uber, Lyft, Tesla, possibly some other auto manufacturers.
Better data allows you to provide better autonomous route planning, which causes more people to use your app, which gives you better data, etc.
Is it really that easy to create an Uber competitor once it's established a monopoly? If you're a driver, would you spend your time on an upstart with very few customers and similar prices? And if you're a passenger, would you use an upstart with few drivers? I'd have to have a significant financial incentive to do so (and that amount represents the excess profit Uber can charge).
This is what Warren Buffett calls an "economic moat", caused by what Silicon Valley calls "network effects". Those effect looks pretty powerful in Uber's case. I think a competitor would have to come in with heavy subsidies (heavier than what Uber offers now) once Uber is established.
Why do you think it will be cheaper to take monopoly power from Uber once they are established than it is for Uber now to establish monopoly power? I think it goes the other way.
Of course, Uber could be thrown off by regulation or lawsuits or a technological shift. Without one of those, I would pretty confidently bet on Uber.
It's a huge moat.
They would need to subsidize rides and pay drivers more but if they do it region by region it could definitely work without being exorbitant.
Many times its cheaper for me to Uber somewhere than to drive + pay for parking, and I will stop if its no longer the case.
I'm pretty price-insensitive in situations where I'm grabbing a few drinks with friends etc. and I leave my car at home, and I price that in as part of that evening's "fun" budget.
Uber and Turo have been great.
Uber is unusual in that, for once, it's the big investors who are the suckers.
It really seems like an underpants gnome situation.
$unit_profit = $unit_price - $unit_cost
In Uber's case, the interesting question is market flexibility on price. Or, another way, can Uber increase prices while maintaining market share, and if so, by how much?
In personal experience, in my market, I believe there's headroom. I never step out of an Uber thinking, "Wow, that was too expensive." (Surges notwithstanding.) Nothing will make me go back to taxis. So the alternative is Lyft or nothing at all. And I think there's room to grow.
I think the strategy of trying to maintain a monopoly to starve others out is daring. There's no guarantee fully autonomous vehicles will be the norm anytime soon enough for the business to capitalize on it. There was an article on here -- days, weeks? ago -- about how easy is it to fool the on-board cameras of autonomous vehicles. Slap some black rectangular bars on stop sign, and now all of a sudden it's recognized as a speed limit sign. I think we're farther away from Level 5 than what is led on. Which is a bit of a shame, I'd love to be in a world where I don't have to drive myself, but I'm also being realistic.
Ultimately, in my mind, uber is failing at, well, failing fast. Don't dance around the worry that your business might not be profitable at the price point you need customers to accept. Otherwise, I think they're just wasting time. You need to validate product-market fit as quickly as possible...
Those kinds of situations can be solved by training on adversarial examples. It's easy to fool them now, but perhaps not after they've prepared for that possibility. Some years ago, one could fool a web search engine by repeating a word dozens of times on a page. It's not so easy now.
Out of the infinite number of possibilities one can encounter while driving, can the systems be trained to apply common sense? You can't program an infinite number of if-then conditionals for every possibility. The system has to be able to "think" and apply reasonable interpretation, like any human would when taking in any visual or audible input and converting that into a vehicle maneuver.
The data today suggests that autonomous vehicles get into less accidents per vehicle mile than human drivers, which is promising. But who knows if autonomous vehicles are being used in complex situations representative of human driving conditions: poor weather, non-highway driving, etc. If we're putting them on the 101 in sunny weather, then yes, I'd expect them to perform reasonably well and definitely better than the average human.
When's the last time you saw someone put black tape over a stop sign? These adversarial examples are not simple use cases. They're situations that humans rarely if ever encounter.
> apply common sense ... "think"
Ever seen the "invisible rope prank"? Humans are easily fooled as well.
> if-then conditionals
That's not the way statistical machine learning works. It's layers upon layers of probabilistic functions.
> who knows if ... representative of human driving conditions
I've read reports that a few of the autonomous vehicle companies are indeed testing/training on regular roads, which would allow a representative sample of human driving conditions. Frankly, that's just basic science -- don't experiment with a biased sample.
The irony is that they are not. They are subsidizing far less than any other competitor. Show me one competitor with lower losses as a percent of gross bookings.
Based on my rough calculation:
1) Say, 100 rides at $2 each = $200 2) Double the price (surging), 40% ride drop: 60 rides at $4 each = $240
Uber still made 20% gain in revenue due to price doubling, right? That's a good thing for Uber, isn't it?
[1] https://en.wikipedia.org/wiki/Price_elasticity_of_demand
>Uber still made 20% gain in revenue due to price doubling, right? That's a good thing for Uber, isn't it?
the next big good thing would be market segmentation : those 40% at $2 while the rich 60% at $4 - the total would be $320 :)
It's already been a game changer in the restaurant world, where proprietors no longer have to price the cost of service into the menu. Sure, customers don't like it, but it does the job of price discrimination where it matters.
I submit that the US would not be able to maintain as dynamic a restaurant ecosystem as it has without tipping. Uber's business model is not sustainable with human drivers without tipping. Eventually drivers will leave once the VC money dries up and ridership declines due to price sensitivity.
Tipping is the safety valve that will allow ride markets to clear properly. Uber wanted a clean, easy user experience, right up until it becomes unsustainable. Now it is, and the only people who can keep the Uber service afloat is riders.
Just like in the restaurant field, where if they price service into the menu, the perceived pricing uptick is enough to make people think restaurant dining is unaffordable. Sure it's worked for one restaurant, but the owner does not intend to bring the practice to his other restaurants.
Eventually tipping culture will crop up around ride share services and everyone will know how to be a good citizen in that space, and assholes will be free to be assholes.
So the alternative is not cab service or pricing. The service can always be superior, just by not having terrible queuing and apps. The pricing should naturally be lower, as taxi companies charge way higher prices.
1: The local cab companies so hilariously mismanaged things they just invited to be destroyed.
I read on the Uber Driver's forums that many (the majority, going by the polls) will now be rating riders lower by default and then increasing the rating after a tip has been received.
I've since basically stopped using Uber. Between the increased prices since legislation was introduced here in Toronto, the drop in driver quality as more of the taxi drivers move to Uber, the forced surge pricing by colluding groups of drivers and now the expectation of tips it has essentially lost all of its appeal to me.
The off-hours public transit, which was once made unattractive by Uber, has returned to being my default go-to for now.
Never in anything I have read about Uber has there been anything more clear than the fact the Uber drivers make a pittance -- nothing near a fair wage.
I know what you are referring to, but is this really the practice? Do you have evidence that supports this claim?
The corollary to tip-based salaries is that the employer is required to provide the difference up to the local minimum wage should the server's total tips be under the minimum wage, as per the Fair Labor Standards Act [1]. Given that this is a known cost, I would assume that to some degree the prices are set to accommodate such costs to at least a percentage of the required funds. I know based on anecdotal evidence that most restaurants are run pretty tightly, but your average chain that pays minimum wage surely must be aware of the potential for slow periods. For tourist areas with on/off seasons, I know the solution is to simply bring on and let go seasonal workers, but I would still assume that it's easier to factor in the potential cost for minimum wage than it is to get caught without available funds.
Another point of comparison is Juno in Manhattan which took a smaller cut from drivers in order to poach the best ones from Uber and Lyft. I've always felt that competition for drivers will drive the profit margin down to what's needed to support a small team of developers pretty quickly.
It's still an open question whether there's a 100 Billion Dollar Business. That's what they sold to investors, and since they cashed the check, that's what they must deliver.
Or some hedge fund will buy the pieces for pennies on the dollar and sell them for scrap.
As an entity contracting human driven vehicles? My bet would be yes.
As for its desired autonomous future? There's no telling.