Beleaguered Didi Is Said to Lose $585M in Just Six Months
bloomberg.com
bloomberg.com
You’re far more likely to die on the highway or being robbed.
Gasp! They'll know where you live!
Which is just nonsensical. They don't know who you are and don't need a delivery job to find a house to rob.
Didi allegedly did not respond to multiple negative comments about this guy's behavior, nor did they expose these comments to future passengers, which might have flagged him as a threat.
If not for Didi, or if they had been more safety-minded as a company, then this guy would not have had the opportunity to rape and murder this woman, allegedly. That's the argument.
If you are a predator, bring a driver affords you all kinds of opportunities to prey you people that you otherwise would not have. How often are you going to find a young woman walking in a dark park by herself late at night? Compare that to how easy it is to pick one up as an Didi/Uber driver.
At the end of the day, a ride sharing company's users literally entrust their lives to the company. It is the responsibility of the company to act on a way worthy of that trust. If they don't, it is definitely a reasonable response to #deleteDidi.
I mean if anything, it's more an argument that ride-share companies should vet their drivers, do basic background checks .. you know, treat them as real employees instead of this "gig economy" bullshit.
Even then, you're still going to get a psycho in the mix, but at least you can say you attempted to do due diligence.
There can be a fair argument as at what stage would it be necessary to remove the driver, but I think the main thrust of the issue is that to the clients, it is promoted as a taxi service and to the drivers, it is promoted as a dating service.
EDIT: This is literally what happened:
"As Huang Jieli—the woman who had served as Hitch’s general manager until she was demoted last week—said in an interview in 2015: “Like a coffee shop, or a bar, a private car can become a half-open, half-private social space. It’s a very sexy application scenario.” And in another interview (link in Chinese) in 2017, she said “the biggest motivation for our drivers is to share a trip with a passenger.”"
https://qz.com/1374431/after-female-passengers-murder-didis-...
Maybe you can argue that the first time that this happens. Well, it happened twice already.
> “Like a coffee shop, or a bar, a private car can become a half-open, half-private social space. It’s a very sexy application scenario.”
You're right. The vision and marketing are also to blame.
I agree with your characterisation of the problem:
> to the clients, it is promoted as a taxi service and to the drivers, it is promoted as a dating service.
The company is misrepresenting what the nature of the interactions between drivers and clients in its marketplace. This misrepresentation predictably increases the risk to women who use the service --- it should not be surprising that a service that allows drivers to see photos and share tags of clients before they agree to take rides structurally and predictably offers more opportunities for abuse. This is frankly pretty foul. I hope the company continues to get hammered for it.
For example, suppose didi was an online dating service, and clearly and unambiguously advertised itself as a dating service to all parties involved. How many women would choose to go on a first date with someone they'd never met before by getting a ride in their car? You're obviously going to be far, far more at risk and unable to get away if the person turns out to be abusive than if you meet in a more public location.
Quantity !== Veracity
Quantity is a metric that is easy to program and to understand for systems design, so we use it as an easy stand in, but we need to get better at finding Quality, not Quantity.
The number of negative reviews serve as a way to help strengthen the signal that a driver is truly "bad" in some way. Of course, the quality of those reviews should also be taken into account.
So yes, I don't have any easy answers. I think this a problem for our industry, and probably one we need to collectively solve sooner than later, giving how antagonists are using our systems against us.
Not something I want to see or use, but I think it's the only way to go about review-backed sites like Yelp or Uber. It's impossible to generally validate statements, but at least a persistent identity lets services decide whether to trust the person.
Anyway, why even bring that up if the solution is just to store the ride audio for a bit longer.
Ah, if only in the year of 2018 there existed a way to verify what kind of interaction happened during each ride.
Perhaps in a form of a phonograph record, or similar, that could be engaged when the ride starts.
If only.
If only.
In someone else’s car with a perfect stranger? I truly don’t care about privacy in that situation, and having an expectation of privacy there is delusional. In my car, I want the option for as much privacy as being surrounded by windows in a public space allows, but the whole point is that this would not be my car.
Come on.
I take it you never go shopping either; if you do, I have bad news for you.
I like it. If I wanted privacy I would call a friend or drive myself... cabs are for convenience. Security footage protects both drivers and riders from abuse, which translates to lower costs and greater availability.
Cab companies like Uber already track routes and riders, that's the valuable and privacy-sensitive information... security footage? Doesn't add much.
What's preventing anyone who doesn't like you from getting you fired?
Yes, establishing due process is hard, but "X% of respondents complain" has never been a due process (it hurts people when it takes too many complaints to fix the thing [and sometimes with things like harassment and murder One is already too many complaints], and it hurts people when it is easy to get X% from a mob). We need to get out of the mindset that just because computers are great at tabulating quantity and numbers are "objective" that doesn't make that the right process, and certainly doesn't make that the due process.
Absolutely, too, we should include workflow steps like "extra monitoring/background checks" in our processes. Often those get ignored because they likely require people/labor and when designing systems we don't always want to include those elements.
Isn't this a characteristic of the asset-sharing industry in general? From Uber's God View [1] to AirBnB's out-of-home campaigns [2] ?
[1] https://www.theverge.com/2014/11/19/7245447/uber-allegedly-t...
[2] https://archives.sfweekly.com/thesnitch/2015/10/21/passive-a...
On top of this there's additional competition from the huge numbers of people using bike sharing like Mobike or Ofo - which not only taking a bike much cheaper, but it's often faster because you avoid traffic jams.
Amazing how much the market has changed in such a short time.
And self-driving cars, flying cars, delivering food, freight, a credit card. Probably more I can't think of.
After Uber pulled out, Grab stopped handing out promotion codes that would save you anywhere from $2-5 bucks per trip (or occasionally more), and they slashed the rate you earn reward points for every dollar spent by ~13-16 times.
Grab became the market, quite literally. It's such a drastic change in just a matter of weeks, and what makes it worse is that there is practically no competition in the ride-hailing scene over here (for now, at least, and hopefully).
Anecdotally, I don't have good experiences with taxis here, and sometimes they're more expensive than Grab.
They're fighting Grab in Thailand and Philippines for now I think.
Recently didi has seen a few new, well funded competitors and coincidentally a rash of negative PR.
Taxis apps are a very solid business model. 2 digits commission for every ride. There is hardly any costs or work involved to match a guy with a nearby taxi.
Subject to certain assumptions of how much customers are willing to pay, price/volume sensitivity, the geographic areas where the company can operate economically, customer acquisition costs, how much drivers need to make to work for you, fixed vs. variable costs, etc.
I suspect that to survive the "ride sharing" app companies will end up significantly increasing prices when it becomes clear there's no magic bullet on the horizon.
ADDED: Which means of course volume will go down and, almost certainly, valuations. But there is almost certainly a business model for some scale at some highish price in sufficiently dense areas.
That leaves 8 figures (over $10B) to be distributed between taxi apps (at least 10% fees). It's a simple business that is very profitable.
There will undeniably be multiple competitors and they will vary per region, but that doesn't affect the viability of the market.
When I interviewed for my current job, (corporate file sharing,) the founder and CEO was very aware of this. He used the terms "race to the bottom" for companies that compete on price. Then, he explained that we had to focus on solving real problems for our customers that our competitors didn't solve.
7 years later, our cheapo competitors are out of business and we're a leader. This, in part is because we offer a feature set that no one else offers.
What does this mean for ride sharing? It means that Didi needs to stop discounting its services and offer features that their competitors don't have. They might have to offer features that the rest of the market thinks are crazy; but crazy enough that Didi's chunk of the marketshare just won't choose the competitor.
Uber, Didi, Lyft are free to prove me wrong by turning a profit. So far, that doesn't seem to be happening.
In New York, for example, hailing a hack outside of Manhattan or the trendy boros was virtually impossible. "Car service" became a thing because cabs left more than half the city unserved. But even those were limited in availability in some areas.
In Las Vegas, it was rare to see a taxi outside of the entertainment corridor. If you called for a taxi, they'd ignore you. The new state taxi commissioner even tried it, himself. He phoned three times for a taxi from a mall just a mile from The Strip, and every time he was ignored.
That's why Uber works in many places. And the app showing the position of the other cars encourages drivers to fan out into underserved areas.
Some people have even done that math and found that Ubering around town is cheaper for them than owning+maintaining+insuring+fueling a car.
I have requested Ubers in almost a dozen different countries and been able to easily get a car without worrying about language barriers or being pressured into increased tourist prices. I actually deleted Uber due to the concerns about their company culture and ended up re-downloading their app because it's so helpful for international travel.
The last time I was in Berlin, Uber drivers were required to be taxi operators. The Uber app just hailed a taxi driver, so there was no real difference in getting a Uber vs hailing a cab.
Instead of each Taxi company trying to stay on top with their own individual apps (which are probably made by like 3 US companies who sell/rebrand their services to everyone), why not an app system that connects you to local Taxi services?
Support your local Taxi company. Put in your destination and get a list of rates and estimated wait times from companies .. you could put fancy Limo companies right next to the Taxis in the rare case someone wants to pay some crazy amount for a limo (or so they can use the same app to book one for a big engagement/later time).
Is there anyone in this space?
I've heard many stories about people saying that cabs and private hire companies simply wouldn't go to <neighbourhood> or pick up people who sounded non-white-affluent, meaning a large swath of people were discriminated against and couldn't get rides.
With Uber/Lyft, the driver doesn't know what I look like, what I sound like, or where I'm going until I'm in the car. And the driver has no reason to discriminate because he's getting paid regardless.
It varies by city of course but there were at least two largely orthogonal issues with cabs in a lot of places. General crapiness (unreliable, hard to get dispatched, won't go some places, credit card readers broken, etc.) and high cost.
e.g. I get a car to and from the airport were I live. It's comfortable and reliable but not cheap.
Yes. Outside of a few markets in the US it was a gamble to get a cab at any price. You'd call and it would be something like "30-45 minutes" and then an hour later you'd call again and hear the same thing. And this just on a regular night, forget about it if there's an event going on.
Where did you live? Downtown Manhattan?
Pre-Uber try getting cab service like that in Harlem or the other boroughs. You mostly rely on gypsy cabs who were always sketchy and always tried to cheat you. Coming back from the airport one time, one driver even tried to put my bags in another cab and tell me he'll finish the ride.
If the pre Uber cab service were so great, Uber wouldn't be such a popular success there right now.
And God help you if you were black and trying to get a ride.
Sounds like you weren't dealing with a TLC-licensed driver, which all medallion cabs and all legal black cabs (including Lyft and Uber) are.
The introduction of low-cost options for Uber/Lyft were arguably as important a development as their original ride-hailing app ideas. Once the fleet was no longer constrained to professional drivers, and anybody with a spare hour or two (and a car) could go online, the size of the fleet exploded. That meant the likelihood of having a car near you - no matter where you are in a city - dramatically increased. Sure, call your local car service from home when you're heading to JFK and he'll be there in 10. How about when you're in the Village at 1am? How about Red Hook? Not to mention this allowed the service to be provided in cities all over the world which didn't previously have a car service at all.
> 10,000 highly-paid developers
Uber probably services more than 10,000x the number of rides than your car service did. I suppose you could try employing 10,000 "ladies with a phone" to handle that volume - but of course human dispatch quickly becomes unable to handle those kinds of volumes. Sorry, those developers are necessary for scale.
> and providing a marginally better level of service
Car-sharing with strangers was impossible before Uber/Lyft operated at scale. Share a car to work with some of your colleagues, who happen to live near you, sure. But to car pool anywhere at any time requires a huge fleet, and the technology to match riders who are taking the same route. That is an very difficult problem to solve, and yes it takes capable engineers. That car pooling allows fares which are probably around a third of what your car service charges. More improvements are on the way which should see short-medium trips of a comparable cost to public transport.
Not to mention providing this service in under-served and otherwise unprofitable areas, to minorities, at all times of day, all across the world. Plus of course a consistent name and brand, by far the biggest value-add ;)
Now integrate with other modes of transport - public transport, bikes, and yep even scooters. These companies are positioned to shape entire cities' transit systems (they already have, for better or for worse). They're not the same as your small car service, and they're not over-engineered. They're certainly not setting out every day worried about justifying their Silicon Valley existence.
New York is a bad example. The system that was here before was pretty exceptionally good.
Standard protocol if you were in the boroughs was to simply ask for the local car service number. Restaurants, bars and bodegas all had it ready. If you were in the city, just flag down a yellow.
A car was rarely more than a few minutes away.
It was a short term gain for Uber, but now that the business has gone on long enough drivers are pricing in the maintenance costs to what they need to drive. That's going to drive up costs for Uber and other ride hailing companies because having every individual maintain their car personally is inherently less efficient than having a centralized fleet with professional mechanics who also make sure to perform preventative maintenance that most car owners avoid
I was referring to Uber/Lyft pools in that paragraph: "Share a car to work with some of your colleagues, who happen to live near you, sure." Agreed, ride sharing is a much misused term.
I can't give an informed opinion on just how profitable driving for Uber/Lyft is right now. Everything I've read on the subject has been (potentially) biased, devoid of real data, or since shown to be incorrect. I'd love to read studies on this if there are some available. I do think the best solution is to make as much data on profitability available to drivers, so the market can correct itself if necessary, like you suggest.
However I don't think the ride hailing companies are blind to the economics of their markets. If there is a problem with the economics, that's an existential risk and they're not going to be sitting on their hands. If we're comparing ride-hailing with traditional car services, it's the ride-hailing side which have a whole lot of knobs they can twiddle to make the market work, and they're adding more all the time (e.g.: Uber/Lyft Pool can go a long way toward mitigating driver "opex" - minimizing those empty seats, like any efficient transport operator.)
I just don't think the efficiency of centralized maintenance is really that big a deal. Even if we ignore all the other factors at play, if that's the make-or-break feature Uber could easily provide centralized mechanical support for their fleet. They already have driver centers in each city they operate in.
The fact that The Oatmeal[1] has a comic on this means that it's more than just me.
- I don't need to make calls, I have list of places I go to most often and calling a taxi is couple of clicks away
- I don't need to worry about payment, it's just "Thanks, bye" and you go out of the car.
- I know what car will come to me
- I can see where the car is while it's coming to me
- When abroad it's so much easier to to use an app instead of talking to a driver who barely speaks English (like Thailand)
So it's much more than just a brand.
So while I wouldn't necessarily say the business model isn't valid, I would say the business model should be more akin to an airlines model than a winner-take-all Facebook or Google model.
Really? Have you checked financial reports of box.com and dropbox.com? They are both publicly traded multi-billion dollar companies
Dropbox made a huge loss in 2017 https://markets.businessinsider.com/news/stocks/dropbox-ipo-...
2Q18:
"Net Cash Provided by Operating Activities of $111.9 million and Free Cash Flow of $102.2 million"
"Total revenue was $339.2 million, an increase of 27%"
"Paying users totaled 11.9 million, as compared to 9.9 million for the same period last year."
"GAAP gross margin was 73.6%, as compared to 65.4% in the same period last year."
"Non-GAAP net income was $48.0 million, as compared to $20.0 million in the same period last year."
Accepts foreign CCs, very good English translation in app, and text based translation to communicate with drivers.
And quite inexpensive; Shanghai at least charges similar distance costs as Lyft and one third the time cost.
They now have a txt message system that should be relied on:
1. It has translation built in.
2. It has pre-canned responses
In practice, with #2, the most common pre-canned response you give is "follow navigation" and my drivers tended to just respond with the pre-canned "ok". Drivers don't call me -- their app even tells them I'm not using a Chinese language app.
Never had any trouble.
It is not perfect, but this method allows me to move around easily with didi without speaking chinese. I had no urban mapping problems, but I do test whatever address I want to go in baidu maps first, or any other chinese map app. That way, in case of problems, I can just show him the location or use whatever voice guidance, but I rarely needed this.
I wish this tired "asking hard questions" trope would stop. I think this maybe started with Zuckerberg. It's nothing more than a rhetorical device. It allows CEOs to acknowledge a PR problem without having to provide any meaningful or prescriptive guidance to the culture that created that PR problem in the first place. You will never read a blog post by a CEO titled "The Answers To Those Hard Questions."
>"Cheng said in an email, according to people who’ve seen the memo and verified its authenticity. “The problem is within ourselves. Our desire for success caused us to forget our mission."
Ugh. When success and your mission can have such orthogonality, it could be argued that your "mission statement" was disingenuous to begin with.
(I spent a few seconds considering if Beleaguered is a first name...)