Is the dockless bike-sharing model a failure in China?
Depends on how one defines "failure"
Photos and flyover video montages of these “bicycle graveyards” — which, to be honest, are beautiful in a terrifying and haunting way — are merely another share-worthy meme in a long line of Chinese memes, following the trail of “ghost cities”, the rise of an ominous-sounding “shadow banking” sector, “building roads and bridges (and high-speed rail) to nowhere” and my personal favorite, the “treadmill to hell”:
How did China use more cement in the past three years than the US in the entire 20th century?
Could China's shadow banking create a similar situation as the 2008 US financial crisis?
The common thread with these memes is an overly simplified narrative that does not scratch much below the surface, consider the wider context, and often conforms to the pre-existing notion that economic development in China is generally unsustainable, to a large extent fake, and carried to its natural conclusion, will end in tears — accompanied with a healthy serving of schadenfreude piled on thick and heavy.
It is very tempting to read about these meme-worthy mangled mountains of multi-chromatic mass-transit mobility machines — say that three times fast! — and conclude that this is yet another example of China’s propensity to waste, and another sign of the impending apocalypse. But if you take the time to scratch below the surface and dig one or two levels deeper, you’ll realize that truth and reality are a lot more nuanced and very different from the simple memes that dominate your various news and social media feeds.
First of all, it’s an extremely young industry — a veritable toddler that a mere four years ago existed only in the minds of a couple of recent college graduates. It is a shame (or perhaps laziness) that some can scan a few photos, skim a few articles or even come across a mass of unused bicycles on a random Chinese city street, and think they have enough ammunition to draw a firm conclusion about how this toddler of a business model is a “failure”. When a drooling three-year old stumbles on the sidewalk or runs into a lamp post, do adults roll their eyes and come to the conclusion that “this kid will just never make it”? Of course not. So why are we doing that here?
Even more significantly, by deciding at this early stage that you already know how the story ends, you close yourself off from diving deeper and considering alternate endings. And by doing that, you might just be missing out on a chance to gain a deeper and better understanding of the complex, rapidly evolving, and not black/white place that is China.
You see, when I started to see these multi-colored bicycles rapidly clog the streets of Shenzhen and Shanghai and other cities in China, I made a conscious decision to dig deeper. And what I found was a story that was absolutely fascinating in the way that it touched on so many important themes about China’s rapidly changing economy — from the role of mobile payments as an enabling platform technology, to the “Game of Thrones”-like battle being played out by Alibaba, Tencent and up-and-comers like Meituan — which as I will delve into below, now owns Mobike, one of the two companies that dominate the space today.
You can read a lot more about those specific ancillary themes here (What does the bike-sharing mania say about the Chinese economy?) but for now let’s turn our attention back to our drooling toddler, starting from the very beginning with a key protagonist in the story, the Bicycle.
The Amazing Bicycle
Anyone who has ever owned a bike knows that they are not the easiest things to maintain. Like the brawny Leafcutter Ant, bicycles carry loads that are multiples of their own weight — perhaps one turn too many, in my case! — while having to travel at high speeds on paths of varying quality under all sorts of weather conditions. To keep these faithful machines in good working condition, its gears need to stay lubricated, its tires inflated, its brakes aligned, etc.
This is my trusty made-in-Taiwan Giant FCR1 city bike. It’s a good-looking bike with something like twenty-one different gear-speed combinations, fast tires, really strong brakes and an aluminum-alloy frame that makes it super light but still sturdy.
I store it in a bicycle rack in the basement of my building, protected from Mother Nature. I only ride occasionally, taking it out maybe once every few weeks when the weather is nice. Because I do not ride that often, I usually need to check that the tires are fully inflated before riding. Once, I forgot to pump up the tires and I ended up blowing out the front tire when I hit a pothole and had to replace the inner tube at typical NYC extortionary pricing.
There is a lot of maintenance involved and I do not even ride it that often and despite the care and attention of its owner — i.e. someone who actually cares about the well-being of the bicycle — it still breaks down every so often.
Now imagine a bike that is used multiple times per day, by riders that don’t care about its well-being, and typically sits outside exposed to the elements. As you are probably beginning to imagine, shared bikes get old rather quickly, especially when they are used heavily.
The financial term for “getting old” is “depreciation” and this is an important point that we will come back to later. So let it stew for a bit while we jump into our time machine and make a quick stop in late 2014 in Beijing, when a handful of recent graduates from Peking University who loved cycling decided that they wanted to start a business together.
The “Little Yellow Bike”
At the 2014 World Economic Forum in Davos, Li Keqiang (China’s second-in-command) formally introduced the notion of “mass entrepreneurship” (大众创业万众创新) — a concept that had already been broadcast informally for some time. Answering the call, five members of Peking University’s cycling club decided they wanted to focus on bicycle tourism and called their business “ofoofo” due its resemblance to a cyclist hunched over on his/her bike, not to mention its easy pronunciation.
The bicycle tourism idea didn’t pan out and neither did many subsequent follow-on ideas. By April 2015, according to the account given by then 24-year old principal co-founder Dai Wei (戴威), ofo was down to its last 400 yuan (around $60) and struggling to pay suppliers. Things were getting a bit desperate.
That is when they stumbled upon the idea of shared bikes. Over the May 1st “Worker’s Holiday” (Chinese equivalent of Labor Day), Dai Wei thought long and hard about the pain points of owning a bicycle for college students. Having lost five bicycles in four years of college, he had firsthand experience with the pitfalls of bicycle ownership. They began to develop the idea of sharing and renting out bicycles on Peking University’s campus and wanted to be ready for the students when they returned in the fall.
At 8 am on the morning of September 7th, 2015, the six ofo employees stood in front of a row of yellow bicycles marking the first official day of operations of its shared bicycle service at Peking University. The team had somehow convinced 500 students to register for the initial launch.
By the second day after launch, there were 300 rides ordered. After ten days, 1,500 rides were being ordered daily. By the end of October, ofo was providing over 4,000 rides per day to the university students. The team knew that they had finally hit upon a viable business idea. Like Facebook a decade earlier — which had begun at Harvard — ofo began expanding its business model to other college campuses.
A Simple Narrative of the Spectacular Rise and Fall of Shared Bikes in China
In China, good ideas do not stay hidden for long. Or even for a short time.
By late 2016 — less than a year after initial conception — dozens of companies had been funded in the space and the industry had already entered the “land grab” phase. With 60,000 bicycles, ofo was serving over half a million rides per day and was on the verge of expanding outside of college campuses into cities with plans to increase the number of bicycles one hundred fold. To do this, it had recently raised $130 million in funding and was in the process of raising another $450 million (in a round that closed in early 2017).
The speed of adoption and the very visible impact on people’s lives in such a short time period was breathtaking. With so much going on, Chinese people proudly annointed bike-sharing as one of the “Four New Inventions” alongside mobile payments, high-speed rail and e-commerce.
The problem was that everyone else was pursuing the same strategy. Another startup with a seasoned entrepreneurial team, Mobike, had launched in Beijing and was about to announce a $100 million funding round from well-known investors like Hillhouse, Qiming, Sequoia and Tencent.
The race — pun intended — was on. Over the course of the next two years, on the back of billions of dollars in largely VC-backed equity, over 20 million bicycles would enter the streets and alleyways of large and small cities across the nation.
2017 is when the problems of rapid scale began show up in the news reports, especially in Tier I cities where the front lines of the “land grab” war were taking place. At some point in Shanghai there were around 1.7 million shared bikes, which comes out to one shared bike for every 15 residents. Experts thought the optimal number was around 500,000 shared bicycles, or one for every 50 residents.
A brutally competitive environment led to rumors and allegations of companies sabotaging competitors’ bicycles, and users abusing the system by parking bicycles inside their own private property, throwing bikes into the river and other anti-social behaviors. City residents complained of crowded, messy sidewalks where many of the bicycles did not even work properly. Local officials began taking measures to curb the amount of bikes on the street and soon thereafter, pictures of these massive “bicycle graveyards” began to show up.
Bike-sharing companies began to drop like flies, many merging their operations in an effort to stay afloat while others were shut down or acquired out of effective bankruptcy, sometimes refusing to refund deposits that had been paid by users when they signed up to their apps. Even ofo, the original dockless (or station-less) shared-bike company with all of its funding, ran into expansion issues and announced that it had to massively scale back its once-promising international operations and cut back on bicycle orders.
With all this going on, were the naysayers right all along? Was dockless bike-sharing on its way to a quick demise and yet another example of “credit-fueled inefficient capital spending” in China? Were the photos and flyover videos of spectacular “bicycle graveyards” accurate metaphors for the prospects of the industry?
Not so fast. Let’s step back for a moment and think about what we are looking at in these pictures in the context of the much bigger picture.
Thinking about the Big Picture, Slowly
Let’s face it, China is a big place. You’ve heard it all before. 1.4 billion people! Miles and miles of seemingly endless high rises! The world’s largest consumer of steel, cement, coal and luxury goods! The list goes on.
It is the level of scale that is hard to truly comprehend for the human brain.
Let me explain what I mean by “truly comprehend”. We can easily tell the difference between two cookies and three cookies. We might even be able to tell the difference between fifteen and sixteen cookies. But unless you are Rain Man, there is almost nobody on the planet that can instantly process and count the 246 toothpicks that I just spilled on the ground.
The brain’s physical limitation shows up in other areas of study, like anthropology with the concept of Dunbar’s number i.e. the cognitive limit on the number of real relationships humans can have (around 150 to 200).
And this all kind of makes sense. When we were figuring out how to survive on the African savanna, there was really no need for our brains to figure out how to count more than a few dozen of anything. “I see five apples, everyone in the family gets one apple tonight. I see two lions, I better run.” To comprehend numbers past a couple hundred, the human brain had to start abstracting (and fortunately for human civilization, it had the capability of doing exactly that).
By now you’re probably thinking … how is any of this relevant?
Quick! Take a look at this picture and within ten seconds, tell me how many bikes you think there are in that pile? Definitely more than a hundred. Probably more than a thousand. Five thousand? Twenty thousand? Half a million?
The answer is around 10,000 bicycles — this was a picture taken on January 13, 2018 in the city of Xiamen as you may have noticed in The Atlantic article referenced at the top in the question link.
If you had more than ten seconds, many people might have been able to get close … perhaps by estimating the cubic volume of that pile and doing some basic arithmetic based the space taken up by a haphazardly placed bicycle … or perhaps by running a “how many jellybeans are in this jar?” style poll and averaging out the responses from a few dozen people to get a pretty good approximation.
But in a snap judgment type situation (like skimming through a news article), the best the human brain can probably come up with is … “a lot”.
And yes, 10,000 bikes is definitely “a lot” … in certain contexts. Is it “a lot” in comparison to the number of total shared bikes in Xiamen? Is the beautifully documented existence of this pile and even many others significant in the context of all of the shared bikes being utilized across the country? How long did it take to accumulate this pile of bike — a day, a month, the past year?
If my (often drooling) toddler saw a pile of say 1 million bikes next to that pile of 10,000 bikes and was asked “how many?”, he’d probably answer, “a lot a lot?”. And frankly, for the numbers we are talking about, that response would about as accurate as any adult’s, including Rain Man himself.
The point is, as much of an impression these photos make on your “fast” brain, until you start to process everything with your “slow” brain, you really can’t begin to answer this question of “failure” on the basis of looking at some photos and watching some videos.
So let’s move away from “thinking fast” and making snap judgments. As former Nobel Prize in Economics winner Daniel Kahneman might say, let’s “slow down” and really think about this question with the rational part of our brain. This means diving into the actual numbers that are coming in from the field in real-time.
From Rapid, Iterative Design Cycles to Bicycle Graveyards
Looking at those massive piles of mangled bicycles, many naturally assumed that most of those bicycles should never have been manufactured in the first place. These assumptions were certainly egged on in many cases (implicitly or explicitly) by the authors of these articles themselves. In other words, “bicycle graveyards” fit perfectly in the “China is wasteful and over-invests in capacity” theme.
And while there was certainly some truth to this (i.e. the “land grab frenzy” I described above) it’s really a lot more complicated than this simple narrative. This is the point at which the idea of bicycle depreciation re-enters the story.
As with any early-stage venture, as ofo began to scale in 2016, the company began to realize that you could not just use any old off-the-shelf bike as part of its shared bike fleet. First, because it was a dockless system, you had to install a GPS-enabled box that could lock and unlock the bike remotely. But more importantly, you had to design a bike that could withstand the rigors of being used (and frequently abused) by the general public. For instance, regular bike tires with an inner tube pop all of the time, so perhaps using solid rubber tires made more sense. Steel bikes rust easily when exposed to the elements, so maybe it made more sense to go with more expensive aluminum-alloy bikes.
As detailed excellently in this podcast, the shared bike companies had to essentially re-design the bicycle from the ground up for this new use case. One implication of this is that the first couple generations of shared bicycles had design problems which led to a high percentage of them having to be discarded. In other words the depreciation rate was very high for the first couple iterations of bicycles as the entrepreneurs figured things out and learned from their mistakes.
On top of this, the bikes were actually being used in a big way. Low fees (1 RMB, 15 cents for a typical ride), widespread availability, and smartphone-enabled ease-of-use led to tens of millions of urban residents signing on and using the service in the first 18 months. Shared bikes were incorporated into the rapidly developing local delivery networks that were driven by China’s e-commerce boom.
So what happens when you have millions of bicycles on city streets that are being used fairly heavily, a large portion of which were not designed for this type of usage? You end up with a lot of broken bicycles.
The original goal was to design bicycles that could last four years. This means that if there were, say, 10 million active bicycles out there, the natural rate of depreciation would mean upwards of 208,000 bicycles per month having to be replaced. That is twenty piles like the one you see above created every single month based on normal usage. And since we know the real-world depreciation rate was much faster than 4 years, this further increases the rate at which discarded bicycles are generated.
The point here is that the mere existence of massive “bicycle graveyards”, taken out of context, could mean multiple things, not necessarily just the “waste” thesis that many naturally assumed or were led to assume. Indeed, one of the major reasons why there were so many discarded bikes is because they were actually being used out there in the real world! And isn’t that the whole point?
Even the over-building issue itself is likely a one-time, non-recurring effect of China’s “scale fast and fail fast” entrepreneurial mentality. It took a mere three years for the industry to go through Gartner’s “hype cycle”, and after about two years of “disillusionment”, the industry has effectively shrunk down to only a handful of players that are no longer focused on “growth at all costs” but figuring out how to compete with each other for market share in other ways (like improving the service).
In other words, China’s shared-bike industry is settling into a less chaotic and more rational market equilibrium, which means the wastefulness of the frenzied “land grab” phase is unlikely to repeat itself again in the same way … at least in the bike-sharing industry (although in other new industries, this cycle is very likely to repeat itself).
Capital Efficiency: Docked vs. Dockless
On April 4, 2018, Meituan, an O2O (online-to-offline) powerhouse that is part of the next generation of Chinese mega-unicorns, announced the acquisition of Mobike, ofo’s key competitor, for a total consideration of $2.5 billion. As the leading player in the local on-demand delivery sector (primarily for restaurants), the acquisition of Mobike was seen as a strategic acquisition ahead of its impending initial public offering.
From Mobike’s perspective, this was a chance to bolster its balance sheet resources in the cutthroat battle for dominance in the shared bicycle industry, and partnering with a fellow Tencent equity investee made a lot of sense. A few months later, Meituan completed its initial public offering in Hong Kong, raising over $4 billion to replenish its war chest to continue its own brutally competitive fight with Alibaba, Didi and others.
One of the side effects of Meituan’s acquisition of Mobike and the subsequent listing is that we now have some more insight into the numbers behind the bike-sharing industry. Up to this point, the public had to rely primarily on patchy, selective, sometimes-inflated disclosure from private companies or leaked, often-stale information from previous fundraisings.
For example, on page 23 of Meituan’s listing prospectus, we catch an audited glimpse into the operational performance of Mobike through the first four months of 2018:
“48.1 million active bike users”
“7.1 million active bikes”
“1.0 billion rides completed in the four months ended April 30, 2018” which annualizes out to 3 billion rides per year, or 8.2 million rides per day. I will note that these include two of the coldest, least bike-friendly months of the year, especially in the frigid Northeast.
We already knew how much capital had been invested in Mobike to get them to that point, approximately $825 million during that frenzied “land grab” phase from August 2016 to June 2017.
Doing some simple arithmetic, we can start to see some interesting operating metrics:
Mobike “spent” about $117 of equity capital per active bike. I put “spent” in parenthesis because they also collected refundable deposits, revenue from paying customers, while also having to run operations, account for broken bikes etc.
The average Mobike active user uses the service about 5 times per month.
Each active bike is used on average 423 times per year (about 1.2 times per day).
Now how do we judge whether these operational metrics signal any sort of capital efficiency? Looking at its financials for signs of operating profit is one way but Mobike’s financials are not broken out and in any case, a good case can be made for why at this stage in the market when the industry is still growing rapidly and there are still fierce competitors out there, it makes sense to prioritize market share growth over accounting profitability (in addition to other strategic considerations).
But one way of benchmarking how Mobike’s capital efficiency is by comparing to alternative systems.
So let’s take a look at Citibike, a well-known docked bike-sharing system in New York City that is now owned by Lyft:
Total cumulative funding for Citibike has been around $200 million.
Citibike’s total system capacity is around 12,000 bicycles and ridership (during the good weather months) is around 50,000 per day.
There are around 150,000 active Citibike users. So the average Citibike user rides around 10 times per month and the average Citibike is used 1,500 times per year (roughly 4 times per day).
Citibike invested approximately $16,700 of capital per bike. The main reason why this is so high is because of the need to build the docking station.
Citibike is still unprofitable as the system operator.
The averaged-out cost of a Citibike ride is around $1.82 — compared to 15 cents for a typical Chinese ride.
Now some rough comparison numbers:
Mobike took in around 4 times the invested capital of Citibike and is providing 165 times the number of daily rides.
Over its lifetime (since May 2013), Citibike has cumulatively served around 70 million rides. Mobike serves this number every eight days.
Individual Citibikes are used more intensively (four times per day vs. Mobike’s once per day) but for the same cost of putting out one docked Citibike on the street, you can deploy more than one hundred dockless ones.
While this comparison is not necessarily indicative that Mobike is capital efficient (it is still losing money, after all), it is pretty convincing that at the very least, Mobike has been significantly more capital efficient than Citibike.
The Drooling Toddler
From 2014 to 2015, I was living in Taiwan. We were renovating our place and I was enjoying some time off after working hard in the investment industry straight through since graduating from college. I distinctly remember using a docked bike-share service called U-Bike in Taipei and thought it was comparable to the first year of using Citibike service in Manhattan. I had generally satisfactory experiences with both.
With the renovations completed, we moved back to States in the fall of 2015. This was around the time that Dai Wei and his ofo co-founders were first launching their bicycle sharing service on the campus of Peking University. I had boxed up all of my papers and personal effects into around forty banker boxes and remember sitting in my office trying to figure out how to deal with everything … should I keep it, scan it, file it, throw it away, etc.
I am proud to report that I have finally finished organizing these boxes! It only took a little over three years! Sigh …
Where am I going with this? Okay, I have to admit that this little vignette has zero relevance to the dockless, bike-sharing story in China, except to highlight one thing:
In the time it took me to organize a few dozen banker boxes here in Lower Manhattan, an entirely new industry was created from scratch in China that today collectively serves billions of rides per year and has had a massive impact on urban development, mass transit patterns and the lives of hundreds of millions of Chinese people. It is an industry that is almost undoubtedly going to be permanently ingrained in the fabric of urban China — possibly with electric upgrades (also here) by more or less the same players that have survived the frenzied competition environment of the last two years.
At the 5-year college reunion:
Glenn: So, what have you been up to?!
Wei: Oh … I just … you know … nothing crazy … started a company … revolutionized urban transport in China … you?
Glenn: Hmmm. Well … I did finally clear out some boxes.
This is an industry that is literally younger than my toddler — who I am also happy to report, is finally starting to drool less. I am not even going to get into the carbon-saving potential or all of the knock-on effects of collecting trillions of heretofore inaccessible data points and how they can be used to inform future urban development, allocate resources and help optimize consumer habits.
If after getting through all of this, you still think that the three-year old dockless, bike-sharing model in China is a failure … on the basis of some anecdotal, totally-removed-from-context-but-striking-in-an-admittedly-beautiful-yet-haunting-way photos and videos … the only thing I have left to say is that your standards for what qualifies as a “success” must be higher than two hippies riding a hot air balloon at a Grateful Dead concert.
This was originally published on Quora in December 2018.