Uber backtracks after jacking up prices during Sydney hostage crisis
washingtonpost.com
washingtonpost.com
I guess the problem is that anytime an "extreme" situation happens (which could be any kind of accident or dangerous weather), Uber would then be expected to absorb the costs.
That said, maybe Uber shouldn't say anything during emergencies? I mean, what's the usecase here? People in the area are panicking for their lives...their first thought is to check Twitter, then check @Uber_Sydney to see if Uber has anything interesting to say about the terrorist hostage situation? Uber should've made no statement on this, and everyone would've more or less accepted the auto-surge situation. Instead, they not only inserted themselves into a citywide emergency for PR purposes, but then they did something that will never, ever be popular. Would really love to hear the PR strategy that inspired that original tweet by @Uber_Sydney
This is the best point I've seen here. Why say anything?
They've made the point of surge pricing known in the past, so there's zero reason to call out its existence as people are trying to get away from a police standoff. Maybe after the fact they could've made this exact statement and not come off as giant douchebags, but saying this as the situation is ongoing is astoundingly tone deaf.
Surge pricing strikes a lot of people as unfair, and the more extreme the circumstances, the less fair people are apt to think it is. Somehow, nobody at Uber understands this or that they need to shut the fuck up about it at especially sensitive times.
Edit: missing word
Another reason to like surge pricing is that, in addition to increasing the supply of drivers, it reduces demand by incentivizing riders to conserve: people are more likely to split a ride with strangers when surge pricing is in effect. That's a desirable property when there's an emergency and your goal is to get people out of an area as efficiently as possible.
Regression to the mean applies to panic as well. With a large enough sample size, all humans, wether they be passengers or drivers are susceptible to fear in the same ways.
The danger seems to have been way overblown, but that applies to both sides of this particular interaction.
Uber has a reputation problem and failed to generate public goodwill. They're the poster child of arrogant valley companies that everyone loves to hate. As logical as surge pricing is, it will be a hard sell until they win the people over.
As a result there was no optimal way for them to respond in this situation. - Disable automated surge pricing, less cars on the street, less people home quickly - Remain silent about automatic surge pricing, guaranteed backlash, more cars on the street
Instead they tried to get ahead of the backlash. Maintains their strategy, gets more people home, paints a huge target on their backs for everyone who hates them.
You're right that surge pricing is a counterintuitive concept for many people to accept, and it's a major gamble, but let's not confuse the HN filter bubble with the millions of people who actually use, and keep using, Uber.
Or, you know, most journalists.
http://reason.com/blog/2014/11/23/forget-ubers-asshole-probl...
If they want to show it worked, they should release numbers - proportionately how many uber drivers were 'enticed' back onto the roads? Show people that it actually works, rather than wave hands around saying 'mutter mutter logic mutter mutter save lives'.
Agreed that from a PR perspective for those who can't grasp simple economics, they should release their data on surge factor vs drivers on the road.
In this case on the one hand you have people desperately trying to leave a place, and on the other hand you have drivers sitting at home not wanting to risk their lives for $x/mile.
When Uber jacks up prices, you are creating incentives for those drivers to get out of their homes and start allowing people to desperately leave the places where they don't want to be in.
Highers prices>More drivers>More people leaving the dangerous places
There is nothing ethically wrong with it. I could even argue that the unethical thing would be not to jack up prices, since this would put less cars on the streets and consequently less people would be "rescued".
"I could argue that jacking up the prices for food in a starving area is ethical, since it would encourage diversion of food from other areas, bringing in merchants who would otherwise not visit"
Or maybe ethics is a little broader than just economic invisible-hand logic?
The optimal way from a user's perspective would be to let surge pricing do what it does. The optimal way from a PR perspective would've been to cap surge pricing (the hysterical lunatics currently screaming about this on Twitter et al would have a lot more trouble getting people to blame Uber for "there's not enough Ubers around!"). It seems in this case Uber chose what was optimal for users, and they were vilified for it.
Now obviously I'm talking about a sustainable optimal solution: what they ended up doing (free rides and refunds) is the most optimal for users but obviously not a steady-state solution (at the very least it would require them to keep tabs on every situation that might be considered an emergency and devote time to classifying those: everything from strong weather events to more anomalous things like the current hostage situation).
That said,
"at the very least it would require them to keep tabs on every situation that might be considered an emergency and devote time to classifying those: everything from strong weather events to more anomalous things like the current hostage situation"
You mean a 50B company with the ability to massively shift allocation of resources needs to pay attention to what's going on in the world, to deploy such resources effectively and ethically?
This seems to drive Google's approach (or non-approach) to customer service, as an example, and also leads to a lot of "why didn't I think of that?" moments when some company figures out they can make a lot of money without automation. (For a good example of the latter, look at all the attempts at electronic bill pay that tried and failed because of network effects before somebody finally realized that they could bridge the gap by mailing checks to anyone who wasn't yet in the system.)
Here, there's an inherent resistance to anything that would require Uber to deploy humans in a way that scales linearly with their business. Surge pricing can be automated, so it's OK. Modifying surge pricing based on the nature of current events requires a number of people that increases linearly with the number of locales that Uber serves, so it gets rejected.
Hint: it's not just about labor costs. Imagine Google ranked their search results by hand, and had to defend themselves against every website owner who thinks there's foul play going on? Being able to point to a ranking algorithm which they swear they don't bias has saved them countless PR headaches (and it's not a coincidence that when other Google properties starting showing up prominently is when they started having to deal with these kind of complaints from govts and others).
Uber's case is even simpler, since the surge factor algorithm is much simpler and much less secret that Google' ranking logic. If you hire a huge base of employees and task them with making tons of potentially PR nightmare decisions day in and day out, you'll be dealing with explosion after explosion and in every case, you can't point to an algorithm, but instead "Uber CHOSE not to count this weather event as an emergency and now some guy got charged a crazy amount of money to get his sick kid to the hospital!!!!".
I don't get why it's so hard to understand that it's a bad idea to put a ton of decisions every day into the hands of frontline employees where any single "bad" decision could turn into a PR nightmare.
I also think it's odd that you think "it's not our fault, it's the algorithm we designed and chose to deploy" gets rid of responsibility for the decisions.
If the end result of rich people funding new companies makes greed, then I vote to stay poor. There is so much potential with "airborne extensive genetics" to do good. But Uber is greed; greed is bad and toxic.
On the surge itself, I think it's a bit instructive to imagine two worlds, call them World 1 and World 2. In World 1, every person has the same ability to pay and attachment to money - their willingness to pay changes only by virtue of how much they need the ride. In World 2, every person needs the ride precisely as much, their willingness to pay is solely due to differences in their financial circumstance.
In World 1, the ideal response in every sense would seem to be surge pricing. Those willing to pay a higher price would have their greater need met, at a price they agreed to. Refunding them is fine but arguably unnecessary.
In World 2, we're clearly discriminating by class (what you think about that depends on your moral framework; I don't love it) by initially presenting the higher price, whether or not it's refunded. It's interesting to note that refunding in this case is particularly weird - you're giving money back to those who need it least - except when you consider that Uber probably has greater ability to pay than anyone in Sydney...
The real world, of course, doesn't resemble either of these.
In all cases offering more to drivers is desirable, other things being equal.
One thing that I haven't seen discussed is that there could be a distinction between moving people to some minimum safe distance, and moving people beyond that.
Media outlets are scrutinizing Uber's every move because scandal is popular and everyone wants to break the next Uber story.
They are doing this because public opinion for Uber is at a low, mostly due to Uber's own doing and shocking lapses in basic business ethics.
Uber got themselves into this mess by behaving deplorably, but one must acknowledge that there's a dogpiling/witchhunt going on as a result.
This story makes it sound like Uber deliberately turned on surge pricing to capitalise on what was going on, which is not the case. Those who have used Uber before know this is an automatic mechanism, simple supply and demand, which in most cases makes sense. It is just unfortunate that this happened when this hostage crisis was taking place.
Surge pricing is and always has been an automatic thing. It isn't something that someone in an Uber office switches on when they notice an increase in demand or a crisis going on they can take advantage of. Uber are a pretty questionable company who have done some rather questionable things, in this situation, Uber did everything right (for the most part).
While I understand the fear is real from some of the people fleeing, the people actually being held captive in cafe were the ones we should be focusing on, not the people fleeing who were safe surrounded by. Calling an Uber in a crisis is the last thing I would honestly be doing, I would be calling an actual taxi because I know there are more taxis than Uber's in Sydney.
As unfortunate as it might be, the system works for the most part. Obviously there were a lack of vehicles in the vicinity of the siege, the system realised this and compensated by increasing the cost. The alternative in this situation if there was no surge pricing, there would be no additional vehicles to meet the demand which means people have no transport. It is also worth noting the drivers get paid more when surge pricing is in effect, the incentive of more money brings drivers from further away into the area.
And you know what, once Uber found out what was happening, they made all rides in the Sydney area free which they didn't have to do. Does this sound the like an action of a company trying to capitalise on a crisis? I don't think so. If Uber didn't reverse surge pricing in the area, then yes, I would be calling them out like everyone else, but as soon as they noticed, they made all rides free and offered refunds to anyone charged surge prices. Anything for a story, right?
I think Uber could have handled the PR around this so much better, especially keeping the public in the loop around complexities (if any) of turning off surge pricing or making rides free. I think this also highlights Uber need to consider pricing caps on surge pricing, a 4x increase is just ridiculous crisis or no crisis. I have heard of some parts of the US which have hit a 6x increase before and then Uber intervenes and caps it.
A fine piece of shoddy journalism. Shame on you Washington Post and every other media outlet making this into a far bigger issue than it actually is.
http://cl.ly/image/2U3l242l350p/Image%202014-12-15%20at%2012...
"We are all concerned with events in CBD. Fares have increased to encourage more drivers to come online & pick up passengers in the area."
Somebody wrote that sentence, and it's a gross one. The same doesn't belong on the Post, it belongs on the person who wrote that and upon the culture that allows for sentences like that to be written.
The person who wrote that sentence explained what happened, of course a real person wrote it, it's a blog post. Automatic surge pricing kicked in, like it does anywhere in the world because there was high demand and not enough Uber drivers. While the app is smart, it can't automatically detect when there is a hostage crisis happening in the area and turn off surge pricing...
Such an uninformed response from the likes of a Hacker News user is actually worrying. I thought everyone here mostly understood how Uber works, obviously not.
I don't understand how a human being can write a tweet like that explaining the price increase, for that price increase to stay in place for an hour, and only after significant blowback, for them to say "oh, we'll not charge riders."
That to me is the part that's actually worrying.
All of this Uber hate is getting ridiculous.
Here are the facts:
0. There is a demand to leave the CBD due to a hostage taking.
1. 4x surge pricing kicks in
2. The person who tweeted about about being concerned with the CBD events was obviously aware of the CBD situation since they tweeted about it
It may have been an algorithm that increased pricing to 4x. But it's clear that Uber can (and will) reduce pricing or cover fares in certain emergency situations.
We know this because that's exactly what they did an hour after tweeting about 4x surge pricing. There's no reason they shouldn't have been able to do this when they tweeted about surge pricing in the first place.
It's a damn sight more disappointing to see this on HN.
Isn't that exactly what surge price is for on Ubers business side of view?
It's very important to keep in mind that the alternative to surge pricing isn't cheaper cars, it no cars (or, more accurately, much too few cars). If there were plenty of cars available, the surge pricing wouldn't have kicked in.
That's what Uber claims all the time. Why does every one believe their very own narrative? Why isn't it enough to tell their drivers that there is business at location X and they'll show up? I mean, like it normally works too.
For $25, you can go several kilometers in a regular taxi[1] - ie, Uber, even at normal rates, represents a significant premium over regular taxis.
If you're correct, that merely telling people that there's a spike in business somewhere means supply will flock to the demand, then it follows that there would be plenty of regular and much cheaper taxis for people to take. So, if you're correct, the only people who would take Ubers in the area are the people who love Uber so much that they will pay a premium of maybe $80 for the privilege of fleeing a scene of terrorism in an Uber rather than a taxi.
Or, maybe you're wrong: Perhaps merely telling taxi drivers that there's business available is not in fact enough to get them to go take it.
Of course I might be wrong - I haven't tried to run a cab service ever.
Edit: Or asked differently - why is surge pricing 4x or sometimes even 5x? Why isn't it like 1.10x? My guess it's profits 101.
It's obvious that Uber is an experiment on many levels. Their approach to public policy and regulation is one, another is whether users will accept a priority of better availability over uncertainty of prices (and so the risk of feeling fleeced), rather that the "old style" total certainty of prices over uncertainty of availability.
1) If demand becomes more inelastic with respect to price, you can raise prices more than you will reduce rides, allowing uber to pocket more money.
2) If supply is elastic with respect to price, more drivers will be ready to provide rides as prices increase.
Uber is saying it's purely case 2 because that sounds like better PR, but any profit-maximizing business that knows any marketing 101 is going to seriously interested in case 1.
case 2 is questionable - how elastic is the supply? The supply of drivers is limited - there are only so many uber drivers. The question is, if you raise prices 4x, will you get 4x as many drivers or something less than that?
> The question is, if you raise prices 4x, will you get 4x as many drivers
Nothing in the surge pricing argument is predicated on that relationship being in place. As long as the relationship is positive, surge pricing works. Presumably, Uber has a data-driven model that allows them to calculate which level to surge to to satisfy a given demand - which is probably the same model they use to set base prices - if the model consistently suggests a 0.9x surge, it means they should lower the prices, which, incidentally they did in London four months ago (also in direct contrast to your case 1).
Yes (ignoring the possibility of wanting to discourage long-term competition by keeping margins low), from a profit-maximization perspective, but not necessarily a quantity-of-rides maximization as uber claims.
> only incur the wrath of users by raising the price when it rains? That's really poor marketing.
Some would argue that surge pricing is already poor marketing, given the negative reaction people intuitively have to it
> As long as the relationship is positive, surge pricing works.
No, depends on the elasticity of demand. If demand is unit elastic then an increase in price will result in a equivalent decrease in rides. In order for surge pricing to increase the quantity of rides supplied supply has to be more elastic than demand.