Uber Slayer: How China’s Didi Beat the Ride-Hailing Superpower
bloomberg.comThis is just a Bloomberg PR piece. China has a plan for the internet: Don't let outsiders control critical IT infrastructure / personal data.
> It was widely suggested in press accounts that the Chinese government helped Didi in its battle against Uber. Cheng rejects that, noting that as the largest ride-hailing company, Didi had to shoulder most of the regulatory burden and paid tens of millions of yuan to cover driver traffic citations and other fines.
Here in 2015 is the president of Didi talking about the strong government support they received:
http://video.cnbc.com/gallery/?video=3000395252
I'm not criticizing them, if I were China I wouldn't want the USA in charge of that either, but let's not pretend this was a level playing field.
A thought that occurred to me recently, is that the main purpose of China's "Great Firewall" isn't so much silencing dissent[0], as it is protectionism.
By making it effectively impossible for foreign (i.e. American) internet companies to operate in China, they've ensured that the spaces occupied by Google, Facebook, Twitter, Amazon, Uber, and even to some extent Microsoft have been filled by Chinese-owned and China-based companies.
[0] Although that too, of course.
Oh definitely.
The interesting trade-off though (as with all protectionism) is where the boundary between promoting and hindering innovation is.
VPN blocking Facebook probably has good ROI for generating internal social networks (which in turn can be monitored).
Blocking Google, even more so GitHub, seems far more dangerous economically as you impair innovators' abilities to obtain research done abroad. (Google is especially bad since a large percent of English language sites have Google dependencies (css/js hosting) -- effectively most of the English internet has become unusable in China without a VPN since the complete *google.com block went in).
I agree, to that extent, I really wish Bloomberg or Business Insider articles weren't posted on HN. They often are fluff and don't understand the details, and thus just create unneeded noise on HN.
I disagree on a wholesale "ban". There are some excellent Bloomberg pieces that dive into details correctly (particularly items by Matt Levine and Megan McArdle over at Bloomberg view).
But these shiny, infographic-like Bloomberg Businessweek features tend to be light and fluffy. Ugh.
To be fair, Uber is heavily subsidizing rides in China (below cost). Even in the US, a foreign company doing this could have anti-dumping tariffs slapped on them.
The article is rather light on details on what Didi actually did better, but from using Uber as a foreigner in China, I could see significant problems in Uber's product execution:
1. The Android app continues to use Google Maps which is Great Firewall blocked, meaning I had to use a VPN to load uncached tiles.
2. Uber in turn IP blocks EC2, preventing me from using my main VPN while in app.
3. Entities on the map are shifted by the GCJ-02 offset (https://en.wikipedia.org/wiki/Restrictions_on_geographic_dat...), meaning entities on the map are shifted by 0.5 miles (look at those cars driving on parkland: https://www.dropbox.com/s/hhkrjyka5uq0pfb/PeopleUber.png?dl=...)
4. User support was absolutely non-existent. After I responded to an automatic email from the local community manager inviting me to ask any questions, I only received an automated emails telling me I won't be helped.
5. Help pages in the app on Uber China tell me to see the Chinese translation, with no actual link to said translation.
While the foreigner use-case in China is niche, the sheer amount of obvious (and easily resolvable) problems I ran into suggests that Uber has organizational problems ensuring both drivers and customers have good experiences in foreign countries.
One blog from Jan 2014 highlights how bad the execution was at launch: https://www.larrysalibra.com/5-problems-with-uber-in-china-a... - the most absurd being that Uber (at that time) only accepted US-branded credit cards which few Chinese nationals hold.
For 1 & 2: when was that as for the last 1.5 years at least they use Baidu and it works fine. 3 yes indeed, very annoying. 4. In the west support is good: from China they just do not respond unless a driver screwed me on purpose (doing the ride without me in the car which happened a few time): I would receive no mail but the amount paid would be set to 0 rmb. 5. Even most car brand names are in Chinese now it seems.
What we find most annoying that I always set pick up as an intersection of two streets (fuxing/wulumuqi but in Chinese) and then stand on the corner and yet they always call to ask where we are. We are on the f*cking corner I entered and we do not speak Chinese very well so a mumbling guy does not work well. They stopped allowing you to send sms to the driver, like you can in other countries which would allow to send google offline translated details... so often they just cancel the ride because we cannot explain. Uber Black still do their best so usually that is the easiest.
Last year things were definitely easier. Drivers were doing their best more for foreigners, you could sms and they would not call before, they would just pick you up.
A big pet peeve is with Google and maybe someone here knows how to fix it; I use Google Translate with offline translations but when you are online, even though Google cannot be reached, it will try to do it online. It is beyond annoying when you quickly want to translate something, you have offline installed and yet it will just hang trying to reach the mothership...
Edit: now that I have access to Google to search for my Translate issue, I will give Netguard a go.
1/2: Are you on iPhone or Android? I was there a month ago and my Google Playstore downloaded variant only used Google Maps.
Great point on the driver calling being an issue for non-Mandarin speakers. I also experienced this. I actually asked drivers why they do this and apparently riders seem to put incorrect location information in -- this in itself indicates there is some problem.
Regardless, I don't begrudge Uber much here; the driver call culture is hard to fix (even in America, I'm called sometimes) and it's not unreasonable to require your customer base speak the local language. (It is unreasonable though for the phone app to assume your locale matches the local language [help screen problem] or to assume that you've downloaded the app from a completely local source [Google maps source issue])
> 1/2: Are you on iPhone or Android? I was there a month ago and my Google Playstore downloaded variant only used Google Maps.
Android. But my wife has iOS and that works as well?
> egardless, I don't begrudge Uber much here; the driver call culture is hard to fix (even in America, I'm called sometimes) and it's not unreasonable to require your customer base speak the local language. (It is unreasonable though for the phone app to assume your locale matches the local language [help screen problem] or to assume that you've downloaded the app from a completely local source [Google maps source issue])
Agree, but not being able to SMS while you can in every other country IS something to begrudge Uber for. If I could immediately SMS with a perfect Chinese sentence explaining where I am, which is how I always do it in countries I don't speak the language (well), then this saves all the hassle. They should put that back immediately. In Hongkong it has it (as it has it everywhere else) and the driver never calls...
Tried Netguard; doesn't work. Apparently Translate does not check the network, it checks if the network is on. While it it is on, it will try to reach Google for the translation. Meaning I need a separate offline device just for translate I guess.
#1 - was this because you were still using the US store downloaded app? I highly doubt they would release Uber to Chinese people with a non-working maps.
When I was in China early this year, they use baidu map on my iphone version. It looks like there was a partnership between baidu and uber back then. There's even a "get Uber" feature build inside Baidu Map app back then.
It uses Baidu for sure and has been for at least 1.5 years.
Edit: looked up the first time I used Uber in Shanghai.
Yes, I was using the playstore app.
But this is what I mean as an example of an easily fixable problem. If user is using the Play Store app, show a dialog or otherwise redirect them to another source. Even a simple help menu telling you about the problem and how to fix it is low-cost.
Why does Uber block EC2?
This is pretty standard, we block all "cloud" ranges for end-user-facing things (antifraud, etc)
Bots on EC2 scraping and bruteforcing and whatnot, free tier being abused for free proxies
Good to know for future projects.
Understandably this is something you should make the correct choice for yourself, you might want to be scraped on readonly articles or public pages for things like thumbnails on reddit from EC2, but deny all `HTTPS POST /register` or something from EC2.
Uber's problems extend far beyond China. While the app is nice, its the cheap fares that bring the vast majority of users to Uber, vs competitors or plain ol' taxis.
Those cheap fares are artificially low, funded by SV money and ignoring regulatory issues. But money does run out (or investors want a real return eventually), and regulators either change the law (nullifying their advantage) or crackdown completely.
Uber's network effects are almost non-existant. Every small geographic market has to be fought and won the hard, and expensive way. And when the money runs out and the fares go up, there is no market lock-in. Barriers to entry are low, potential competitors include anyone with a car, Uber just doesn't have a way to lock in market dominance.
Which is probably why they're rushing toward driverless vehicles as fast as they possibly can. Their own staff are an existential threat so they have to remove them as soon as possible. Except that puts them in direct competition with a whole range of extremely well funded, and well connected industry players.
So. Good luck Uber, this is just an example writ large of where Uber is heading in the not to distant future.
For those crying wolf about protectionism...
In general, I think most nations should have more closed door policies like China. My country (which won't be named) does not have a startup ecosystem anymore and we have become a consumption ground for one silicon valley startup after another. Instead of fostering local startups and paying some money (since many bootstrap and do not have big pockets), all the companies just take the freebies of silicon valley startups. We had a local chat start up but slack ate their lunch. Everyone here just uses the slack free tier. How do you compete against free?
Eventually, this has led to all the data of individuals and companies on US soil (or wherever those startups feel like stashing the data). We have no control or say over it. If we are to look for alternatives, there is none in our own country.
And for those who think the world is one giant land where everyone can roam freely, it's impossible for me to move to any of the western world because of visa restrictions. And US has the harshest work visas on the plante.
> We had a local chat start up but slack ate their lunch. Everyone here just uses the slack free tier. How do you compete against free?
The point of division of labor (aka how civilization got built) is that it's not _necessary_ to spend time re-doing the same thing others are doing. Slack is giving you something for free that you don't _have_ to spend time or money building now.
Now you have X engineers free to work on something actually innovative, not just a clone of X, and getting a cheap/free service from slack.
The point of economic policy should not be to produce pointless jobs.
Nations, communities, individuals: they value autonomy. Building their own networks rather than submitting to whatever is being fed by silicon valley is not a pointless job.
Apparently not, otherwise they would have used the locally produced. But they didn't, they chose the freebie SV tools.
The problem is the world is divided into factions, and China is a massive country that just happens to be a separate faction, which means they can't divide labor with the US, they have to compete head on, or else they won't get a fare share of the pie.
US had a head start on development, which means these two factions didn't have a level playing ground to compete for a piece of the pie. Therefore China must protect to get back on level playing ground.
The pie is not a fixed size. The Chinese could produce something new and innovative instead of clones and sell those to the US, making the pie bigger for everyone.
Protectionism is bad in general - it works until everyone starts doing it, then everyone is worse off. But the benefits of protectionism in the short term is not equally distributed and China definitely has greater incentive.
It worked for their manufacturing ecosystem, maybe it'll work for tech too.
They tried, and US government bans Huawei
The US government banned Huawei for many reasons.
One of the reasons that should hit home to the HN crowd is that Huawei simply up and 100% cloned some of the highest-end Cisco routers in their early days.
While I'm no fan of patents, when you clone something down to the bugs (which is how Cisco demonstrated the cloning) you deserve to get some penalty.
Huawei was banned from government contracts, not from the U.S. [edit: afaik]
Also, it appropriated so much of Cisco's IP (quite poorly), at least as it was establishing itself as a network equipment vendor, that I wouldn't call Huawei innovative, not back then.
It was a pretty clear case of cloning very expensive tech for a dime with MASSIVE government support.
Not only from government contracts. I remember watched a news clip years ago which interviewed the owner of a local ISP in midwest, who tried to use huawei's equipments and got FBI knocking his door.
I see your point of huawei cloning cisco's IP back when it first started. But today it's the biggest applicant for IP globally.
So where do you draw the line? Should there be one computer company, one car company, one airplane company, one movie studio, one agriculture company, etc? That's the natural conclusion of your logic. Either you add a bunch of special cases (that doesn't work in X, Y, and Z industries, or in country A's economy or culture), or your logic is flawed. And if you add special cases, you have to justify them.
So either you think that there really should be one company per thing (for whatever definition of thing), in which case that's a whole 'nother ballgame, or your theory is logically inconsistent.
The problem is, say you have developed an app that is on par with Slack.
You launch and are trying to grow your business.
Then Slack shows up and is giving away a similar product because they have VC money to burn. How do you compete?
I completely understand Chinese (or any other country) entrepreneurs, why let some foreign company come in and take all the money (ultimately a lot of it outside the country) when you can do the same internally. The idea is not a rocket science, all about execution. And locals know the specifics of China more than a bunch of silicon valley grads.
Cheers to these guys for making it work in China. And Good luck to Uber fighting strengthening competition all around the world.
> why let some foreign company come in and take all the money (ultimately a lot of it outside the country) when you can do the same internally.
Because those engineers and entrepreneurs could have spent their time producing something globally new and innovative instead of creating an Uber clone. Then they could have sold that new thing to the world and everyone (global consumers and the aforementioned Chinese) would be better off than the status quo.
Or - considerably more likely - they could have created a significantly less successful Uber clone, like many engineers and entrepreneurs have managed in other countries where Uber operates. I mean, if it were that easy to come up with revolutionary new business models that the entire world needs, we wouldn't see quite so many cookie cutter startups in Silicon Valley either...
Actually, the more plausible arguments against protectionism and state aid for favoured corporations are the precise opposite to the one you're making: it's good for consumers to have a wider choice of ride-booking apps, and not a choice guided by the government. If the ridesharing apps are likely to tend towards being a monopoly anyway then there certainly are advantages to having it a locally owned and run one.
The question becomes what the consequences of that should be. Since you're arguing in favor of elaborate protectionism, that becomes a central issue.
Should Xiaomi be artificially blocked from selling smart phones in the US? Should all Huawei products be blocked? Should the policy of the US Government be to directly assist Apple versus Samsung (or any other competitor) in the US market? Should the US begin directly subsidizing immense steel production, using the US Dollar reserve currency to intentionally bankrupt China's steel producers via Fed printing to pay for it all? To what extent should the US be using the US Dollar to destroy foreign corporations to gain a competitive advantage? Or better yet, using the NSA to do so. How far should protectionism go?
Should the US use the dollar to start massive state enterprises, printing dollars to pay for it all, to buy up foreign corporations. For example, maybe the US should create a massive potash conglomerate from a state owned enterprise backed by the Fed directly. Then use some USD printing to merge up with other big potash players around the world, and use political leverage and threats to force other nations into allowing those mergers. Japan for example has been buying up its own stock market, maybe the US should go a step further than that. If China wants to compete using big state owned enterprises, why shouldn't the US respond in kind, and then some? China also directly funds non-state owned enterprises through various banking channels. Maybe the Fed should be providing 0% interest loans to US corporations to advantage them to acquire specific major foreign competitors for essentially no cost.
Should the Fed provide Apple, Microsoft, Amazon, Google, etc. with a nearly unlimited budget to begin manufacturing all of their hardware & components in the US, along with perpetual subsidies to keep that production local? Might as well take advantage of available leverage.
If not, then why is it ok for China's government to set up intentional barriers to entry or to encourage rather drastic dumping policies in several big industries? Should foreign companies not be allowed to buy majority ownership positions in domestic US corporations? Should the US Government deny all attempts by foreign corporations to buy domestic corporations? Why is it considered ok for China to directly advantage and subsidize so many massive state enterprises, if the US and other nations aren't supposed to behave that way?
Should the US Government look the other way, while a major US entrepreneur steals tens of billions of dollars from a Chinese company, as Jack Ma did with Yahoo and AliPay? Is that the kind of protectionist playing field that is desirable?
Facebook, Twitter, Google and numerous other US companies have been directly blocked from competing in China in all sorts of ways. Which companies of China's should the US be blocking in response to that?
Because Uber invests most of their money on US soil, on US engineers, that go back into the US economy.
They just want to release an app that harvests money from Chinese consumers.
So if China can protect a local company that will clone uber, but with the difference of investing in local human capital and infrastructure, why not take that opportunity?
Why shouldn't the US begin very aggressively promoting its own domestic corporations, directly, at the expense of foreign companies as well then? Using all means available to the US Government to do so.
Nobody here that is saying it's great for China to protect its local ecosystem, will think it's ok for the US Government to do likewise and doing everything in its means to leverage its particular advantages in favor of companies.
How about the Fed set up a US manufacturing fund. Instantly create a trillion dollar sovereign fund out of thin air, to decimate or buy-up foreign manufacturing competition. Of course the Fed wouldn't be so obvious about it, you'd want other nations to be the frog getting boiled in the pot, you want them to go to sleep with assurances that it's nothing sinister. Next thing you know, Tesla is getting $100 billion in free capital to go after BMW and Mercedes with subsidized vehicles. Or maybe the US should start a state owned lithium enterprise and buy up all the lithium supply, and use the leverage in the Western Hemisphere, to block out all foreign access to lithium in South America. And leverage all of that into a huge benefit to domestic electric car production. China did the same thing with rare earth metals. Or how about Uber get a 0% loan from the Fed for $50 billion, to go after the rest of the world (examples intentionally exaggerated). Those are the sort of policies that China has in place right now across numerous major industries.
This is Trump's campaign platform, which is to start protectionism for the US. And I agree, it would be a necessary step. You can bitch and whine about other countries being protectionist, but the appropriate response is to actually do it yourself (in the US).
So if China can protect a local company that will clone uber, but with the difference of investing in local human capital and infrastructure, why not take that opportunity?
China is the textbook beneficiary of global free trade -- as Donald Trump's candidacy shows, there's significant appetite amongst the American electorate for renewed protectionism. Certainly many voters think (wrongly, I would argue) that Chinese manufacturing "harvests money" from American consumers.
Having said that, I don't think there's any evidence that the Chinese government put an anticompetitive thumb on the scales against Uber. That doesn't mean there's a level playing field for Western companies in China, but Uber's failure can be chalked up to a number of strategic missteps.
Giving an American company where your people are, where they are going and with whom they are likely meeting and traveling was a non-starter before we confirmed every Tom, Dick and Harry at a three-letter agency has the ability to stalk their - and anybody else's - kids and ex girlfriends.
through anti-competitive policies?
What kind of anti-competitive policies?
Dumping their product below-cost?
china is a difficult market. if it wasn't for the perpetual carrot of 1 billion consumers i doubt many would bother.
Wait, so we're supposed to believe this was some kind of merit or business acumen thing? Is anyone really that gullible?