"We've Been Here, and So Has Brother Xing": Three Entrepreneurs of Huaqing Jiayuan | This Is Wudaokou, Full of Young People ④

"This story feels very close to me — real, inspiring, get-rich-quick... Here, dreams feel within reach."

**After three pieces about Wudaokou, let's look again at some of the labels attached to this place — "code monkeys," "internet finance," "center of the universe" — all of them tied to entrepreneurs.**Most of the crowds moving through Wudaokou subway station are tied, one way or another, to the clusters of internet companies nearby.

The most famous of these is Huaqing Jiayuan. Finished in 1994, the compound was one of the first projects Ren Zhiqiang's Huayuan Group sold to individual buyers rather than to work units. It linked the ambition and hormones of the universities around Wudaokou to the internet 1.0-era companies lining Zhongguancun East Road across the street, and to Xierqi, two subway stops away.

Of all the tech gods who emerged here, the one everyone mentions is Wang Xing.

Here, hormones turned into money, into capital, or into changing the world.

Over the past year, Xu Can has changed addresses three times: Huaqing Jiayuan and then Zhichun Jiayuan in Beijing's Wudaokou, and finally back to the Jingbin Industrial Park in Tianjin's Wuqing district. When he moved out of Huaqing Jiayuan last June, he posted a photo to his WeChat Moments — a rainbow he had shot on his phone from the company rooftop in midsummer 2015. The caption read: "We've been here, and so has Brother Xing."

Xu Can's "Brother Xing" is Wang Xing. Around 2006, a batch of early internet startups was born inside the residential blocks of Huaqing Jiayuan, many of them started by former executives from BAT or the four big web portals. Wang Xing founded Xiaonei, then Fanfou, then Meituan. Wu Shichun and Chen Hua founded Kuxun. NetEase vice president Zhou Feng and several Tsinghua graduates built the first architecture of the Youdao search platform here. There was also Yijian Interactive, founded by Lin Yingming and Duan Hui; Meilishuo and Zhuaxia, founded by Xu Yirong and Chen Zhenyu; the instant-messaging service Hada.com, founded by Cao Weibin and Li Jinbo; and Baofeng Yingyin, the video player better known as Baofeng, founded by Feng Xin.

The nickname "grassroots Silicon Valley" dates from that period too. The advantages were obvious: it sat right next to the Wudaokou subway station, newly opened in 2003; it was close to the universities, which made hiring easy; and it had cafes and bars nearby, which made it easy to meet people and talk business. Most important, rent was cheap. In 2006, rents at Hualian Building and Tsinghua Science Park, across the street from Huaqing Jiayuan, had already reached about 7 yuan per square meter; an office with 200 square meters of gross floor area and just over 100 square meters of usable space cost 42,000 yuan a month, while a small two-bedroom apartment of 70 square meters in Huaqing Jiayuan rented for only 4,000 yuan a month — a tenth as much.

Different people have described the scene here at different times and in different settings. When Kuaishou CEO Su Hua and Cheng Yixiao were tinkering with short video, "upstairs someone was building drones, downstairs someone was doing group buying". Goumin.com founder Tang Yang could send one text message and pull several fellow alumni, all starting companies in Huaqing Jiayuan, down to the compound garden for a chat. Shanxi coal boss Huang Zhihua lived in Building 5 of Huaqing Jiayuan, negotiated an equity investment with a team in Building 6, and, hot on Meituan's heels, founded Ayatuan, which burned through 25 million yuan of investment in 14 months.

By results, the most successful of them all is still Wang Xing. Two months ago, Meituan, eight years after its founding, filed to list on the Hong Kong Stock Exchange; it has just passed its listing hearing, at a valuation of 35 to 40 billion US dollars.

No one has ever counted the startups inside Huaqing Jiayuan. People like to joke with an old line once applied to Zhongguancun's entrepreneurship street: throw a brick down from the second floor, and of the ten people it hits, five are starting a company, two have already failed at one, and three are about to start one — and the cleaning lady who picks the brick up is off to share her own entrepreneurial experience.

The delivery man Xiao Ji has his own way of measuring it: by building, by number of parcels. Xiao Ji has worked the Huaqing Jiayuan route for seven years. Before 2013, in the compound's 23 twenty-story residential towers, each building held at least three or four startups, seven or eight people crammed into a single room, and each building's parcel volume ran to 60 or 70 pieces a day. That is nowhere near the several hundred a day an ordinary office building sees, but it was a lot for an ordinary residential block.

Xiao Ji has no idea what these companies actually do. From the outside they look no different from ordinary homes: red high-rises, shade trees, old people walking dogs or sunning their grandchildren in the compound's central garden. But these young people add him on WeChat to make shipping easier. On his Moments feed, someone moves in and someone else moves out every day. One CEO has since moved to the 768 Creative Industry Park in Zhongguancun and now has more than 100 employees; a few recent graduates just won a college entrepreneurship award, and picked up the trophy in the Great Hall of the People.

Next to the CEOs coming and going, Xiao Ji's life around Huaqing Jiayuan looks steady. Since arriving in Beijing from Fengning, Hebei in 2003, he has worked for three delivery companies — Xiaohongma, then ZJS Express, now SF Express. His home has gone from a 10-square-meter room in Zhongguancun at two or three hundred yuan a month to a 12-square-meter room in Qinghe at 900 yuan a month; breakfast has gone from a 40-cent sesame cake to a takeout order costing ten-odd yuan; his pay has climbed from a few thousand yuan to more than ten thousand. Not long after "Double 11" was born in 2009, his delivery bicycle was swapped for an electric tricycle. The one thing that has not changed is the work itself: out the door at eight or nine in the morning, done at six or seven in the evening, dropping off parcels, picking up parcels.

"If a company can't move out within two or three years, it's going to fold," Xiao Ji sums up. He counts off two names, Tingjiandan and Zanyi Interactive, both of which have moved out — leaving Huaqing Jiayuan for somewhere bigger is the sign that a company is on track.

1

Xu Can did spend less than two years at Huaqing Jiayuan, but he moved out not because the project had worked. It was because the money had run out.

Ten years after Xiaonei was born, Xu Can rented a 248-square-meter loft on the 20th floor of Building 20 in Huaqing Jiayuan for 25,000 yuan a month, one floor below where Wang Xing once had his office. He was 25 that year, about the age Wang Xing had been when he founded Xiaonei. He privately took these small details as a good omen.

When he moved into Huaqing Jiayuan, Xu Can had not yet been in Beijing a full year, and had just wound down his first venture: Aoxiu.com, an online social product for the photographer community.

Xu Can ended up at Aoxiu.com by accident. At 18 he bought his first DSLR, a Canon 50D; the next year he got a national photographer's qualification certificate and started taking freelance jobs to "get into the circle." Two years after graduating, at a photographers' event, he met his future partner, Sanshui, who talked him into starting a company in Beijing. At the time he was working at a construction plastic formwork company in Tianjin, earning 100,000 yuan a year, and had just bought a wedding apartment in the Jingbin Industrial Park at 7,000 yuan per square meter, planning to marry, have children, and live an ordinary life. Sanshui's argument was simple: "While you're young and have nothing tying you down, why not build something from zero to one?"

Sanshui had run product strategy at 360, and after leaving in 2009 went into wedding photography. The two of them judged that professional photographers were trapped by the traditional photo-studio model, limited to infrequent, high-priced single jobs, while plenty of people wanted to become professional photographers and could not find a convenient platform. Aoxiu.com could be an online platform built to serve photographers, offering training and equipment services, with sections for teaching, portfolio sharing, equipment talk, and wedding-and-portrait certification (AWP).

Aoxiu.com launched at the end of 2014 and drew more than 30,000 photographers within six months. The company moved from a residential flat in the Shiliuzhuang neighborhood near the South Fourth Ring Road to a 90-square-meter office in Jianwai SOHO at 15,000 yuan a month, then to Huaqing Jiayuan at 25,000 yuan a month, with staff growing from 8 to 35. In March 2015, Aoxiu.com raised 3 million yuan in angel funding.

Looking back, Xu Can thinks they were too confident. They did not know what to do with the money they had raised, so they threw it all at rent and salaries, instead of planning ahead or putting it into new products. And as a PC product, Aoxiu.com did not need that many engineers to maintain.

Being so vertical, the business grew slowly. By June 2015 the money had run dry and employees were leaving one after another, so Xu Can and Sanshui decided to shut Aoxiu down and pivot to a more consumer-facing project, the Meipian app — "helping photographers find clients, and bringing good service to people who can't spend a lot but still want nice photos."

In other words, the product mostly used the photographers they had already gathered to sell shoots booked online, with a "99-yuan shoot-as-you-like" package in 176 cities: a one-hour shoot with a professional photographer in your own city, eight retouched photos of the client's choosing, extra retouched photos at 10 yuan each.

When Meipian launched in October 2015, Beijing was in the middle of a smog spell. They lined up 25 cafes across Haidian, Chaoyang, Xicheng, and Dongcheng districts and ran five rounds of a marketing campaign called "Most Beautiful Cafe Goddess," which added more than 40,000 followers to their WeChat official account. In Xu Can's words, these were their first "seed users." The campaign later spread to 54 cities.

The two of them argued fiercely over whether to launch the 99-yuan product first or the 399-yuan one. It turned out the cheaper one opened the market faster. At its best, Meipian was handling 400 orders a day and 400,000 yuan in transaction volume, at an average order value of around 300 yuan. To their surprise, the strongest response came not from Beijing, Shanghai, or Guangzhou, and not from parents buying family photos, but from single women aged 25 to 40 in second- and third-tier cities earning 8,000 to 10,000 yuan a month.

But trying a sharing-economy model in photography meant solving at least four problems: the public has no habit of booking shoots often; how to balance standardized photographer service against customers' demand for personalized shoots; how to keep photographers from defecting to other platforms or taking jobs on the side; and, once the first three were solved, how to lift margins enough to keep the platform running.

After 48 updates in six months, Xu Can and Sanshui believed they had found a workable system. It included screening incoming photographers and at least three months of training, a fairly mature set of service standards, a commission (50%) lower than the cut photo studios take (80%), and a multi-tier product line (99, 399, and 1,699 yuan). Traffic came mainly from word of mouth generated by group-buying bundles, plus city and national rankings of photographers derived from customer scores.

But under the Zhihu question "What do you make of the app 'Meipian'?", several answers pointed out that a 99-yuan shoot, counting round-trip travel and editing time, actually costs a photographer three hours, so a 50% commission works out to 23.33 yuan an hour; and that Meipian photographers' finished work was mediocre, not good enough for first-tier-city customers.

And then there was the most fatal problem of all: how often people buy. Sharing-economy companies like Didi, Mobike, and Flower Plus all had to subsidize their way to market share early on, using money from investors. If Xu Can and Sanshui wanted scale, they would need, at a minimum, to find investors willing to burn money on Meipian.

In July 2016, an investment firm sent a term sheet offering 10 million yuan. The two of them turned it down, thinking some of the terms too harsh, including giving up part of the company's control. By March of the following year, order volume was still growing 5% a month, but the money was essentially gone. Xu Can regretted his earlier arrogance and began meeting investors back-to-back at Bridge Cafe near Huaqing Jiayuan.

He saw everyone he could see, and no one would invest. One investor told him: "Your product is fine, it just needs a huge injection of cash up front, and nobody's going to invest in that right now, so wrap it up quickly. Do you have a background like Wang Xing's?" He was probably referring to the story of Wang Xing, who, unable to raise money after founding Xiaonei, borrowed 500,000 yuan from his father to pay salaries and held on long enough to sell the company to Oak Pacific Interactive for 2 million US dollars.

Neither Xu Can nor Sanshui had a father who could lend them 500,000 yuan. And with a pregnant wife, Xu Can was not going to sell the apartment in Tianjin. With the rent on the Huaqing Jiayuan apartment up from 25,000 to 28,000 yuan, Meipian moved to Zhichun Jiayuan in June 2017: 120 square meters, 12,000 yuan a month. They did not need the space anyway; staff was down to ten.

To stop the bleeding, Sanshui even went to Sanya, in Hainan, planning to open a separate studio doing travel photography. He talked to Xu Can about "the commercial value of consumer-grade imaging" — retouching alone could earn 50 cents to a yuan per photo. Xu Can pushed back hard: "If we open a studio, how are we different from all those photographers in Sanya making money off tourists? Can you even call that a career?"

After Sanshui took the team to Sanya, Xu Can spent a month alone in the office at Zhichun Jiayuan, until the company folded at the end of March this year. In more than three years in Beijing he had kept one habit: eating, sleeping, and working all in the office. He preferred the feel of a residential building to a shared workspace; it felt less restless. In that last month he went over the past two years again and again in his head, trying to find where things had gone wrong. Sometimes he thought the problem was money; sometimes he thought he himself was not "elite" enough — no prestigious school, no time abroad, always turning up to meet investors in a T-shirt and jeans, no talent for talking up grand business visions.

After that, Xu Can changed his phone number to escape the constant "why did you stop?" He worked for two months in Meituan's smart-payments division, but kept saying "Meipian" when he meant "Meituan."

At the end of June, thinking of the daughter about to be born, Xu Can decided to go back to Tianjin and join Ziroom's marketing department. The eager stream of posts filling his WeChat Moments switched from Meipian to Ziroom, along with the discomfort of going from boss to employee, broken only by the occasional reflection. One, from an afternoon in mid-August, read: "I'm starting to understand why Mao Kankan chose that path." He explained that it was the helplessness and loneliness of giving everything you have and still failing to meet the expectations of the people around you.

The wave of people starting companies in Huaqing Jiayuan alongside Xu Can was far smaller than it had been ten years earlier. Even Xiao Ji could see it: "After 2014, you were lucky if a building had even one startup." That was largely down to the crackdown on shared rentals, and to broader changes in the startup climate.

These companies were also far less known than Wang Xing's generation; the only examples Xu Can could name off the top of his head were Wanmen University and Kuaikan Comics. The first does online education and said last year that it had broken even and was preparing a Series A round, aiming to raise 30 million yuan. The second runs a comics platform, closed a $177 million Series D round at the end of last year, and published the comic behind the recently released film Take My Brother Away.

Xu Can still keeps a photo taken one night while working late at Huaqing Jiayuan, in which he wears a black hoodie and black-framed glasses and looks radiant. "Some experiences are a lifelong asset, impossible to forget as long as you live."

2

Almost at the moment Xu Can returned to Tianjin, Bridge Cafe, the ground-floor shop on the east side of Huaqing Jiayuan, also closed after 15 years in business.

The year Xiao Ji came to Beijing to deliver parcels, an American named Kang Wei was 24, freshly graduated from the Hopkins-Nanjing Center at Nanjing University. In Wudaokou he got to know Zhuang Songlie, the owner of Sculpting in Time, and wanted to open a cafe of his own. Because the rules for registering a foreign-invested company were demanding, Zhuang Songlie licensed his own brand to Kang Wei. After biking around the neighborhood to scout locations, Kang Wei picked a spot on the Caijing East Road side of Huaqing Jiayuan and named it "Sculpting in Time Tongwenguan." It had 45 employees and nearly 500 square meters of floor space across three stories. He handed day-to-day operations to a Chinese woman named Lili and went back to Harvard's Kennedy School of Government to keep studying.

At that point the Wudaokou subway station had not been built, there were no startups yet inside Huaqing Jiayuan, and there was no Hualian Building or Tsinghua Science Park across the street; the only ground-floor shop on the Caijing East Road side was a bar called Wujiaoxing. The first customers were drawn simply by the unusual events at Tongwenguan: English training, lectures, live music, debates, bilingual contests, poetry evenings.

In 2005, Tongwenguan was registered as its own corporate entity, and in 2008 it was renamed "Bridge Cafe." Kang Wei brought a chef over from the US to develop the menu, switching the signature dishes to American breakfasts and paninis, and redid the decor in an American country style. It no longer offered English classes, but the cafe filled with Chinese and international students from nearby universities meeting for one-on-one tutoring. Kang Wei later explained that Bridge wanted to build a way of living around international exchange: "Here you can meet people you'd never meet even in the US — Cubans, for instance."

Lili arrived in Beijing two years before Xiao Ji and Kang Wei did. Her family is from Luanping County in Chengde, Hebei; a shared car ride home from Wangjing West costs only 50 or 60 yuan, or she can take a two-hour bus. Her relatives and friends are all in Beijing, some running businesses, some married in. After finishing high school, Lili followed them into UBC Coffee at Zhongguancun's Haidian Book City, working her way from clerk to manager; her husband ran the bar at the same shop.

Although she picked up most of what she knows about running a shop at UBC Coffee, Lili thinks it was Bridge Cafe that shaped her taste. In 2003 she saw a cafe like this for the first time: dense rows of single wooden tables and chairs, green-shaded reading lamps, red curtains against white walls hung with art. Not everyone who came in was clutching a laptop and dressed for a quick business meeting — there were also students bent over books, conversation partners practicing a language, shy young couples. The lesson was immediate and concrete: "Why would anyone want to come in? Beyond function, a cafe needs some culture of its own — it isn't just about drinking a cup of coffee."

After the 2008 renaming, more startups appeared nearby. Lili turned three small classrooms on the third floor, originally used for spoken-English training, into one 56-square-meter room for parties and small events. Besides student organizations from Peking University and Tsinghua, the most active customers were startups from the Yousheng Building and Huaqing Jiayuan. CreditEase held many events here, and whoever dealt with Lili would always mention how often founder Tang Ning had come by back when he was starting the company at Huaqing Jiayuan.

It was also in 2008 that Bridge Cafe went to 24-hour operation. There was little profit in it, and it added costs like night-shift subsidies and utilities. Lili's main thought was that it was hard to clear people out at one in the morning anyway, so she might as well serve the all-nighters. Besides entrepreneurs, the all-nighters were mostly students. In the month to month and a half before exams, the place was packed every night, and people only drifted out around three or four in the morning. Others tired themselves out at the bar downstairs and came up for breakfast at four or five before heading home.

Across her 14 years at Bridge Cafe, Lili's life stayed regular: in at 8:30 every morning, reading reports and customer feedback on the way, then walking the shop floor by floor from bottom to top after arriving, checking food prep, safety, and hygiene. After that came outside partnerships, scheduling events, planning menus, prepping ingredients, setting up the space, and odds and ends. The dozen-plus servers' lives ran just as strictly by a rota set every Thursday: the early shift from 8:30 a.m. to 5 p.m., the mid shift from noon to 9 p.m. or 2 p.m. to 10 p.m., the evening shift from 4:30 p.m. to 1 a.m., and the night shift from 12:30 a.m. to 9 a.m.

The servers were either young people of eighteen or nineteen or older women whose children lived in Beijing. The young ones turned over fast — some just wanted to try it, saving a bit of money to go cycling; others wanted to earn more and later open a cafe or nail salon in Beijing or back home. Lili remembers a girl from Shandong who met her husband at Bridge Cafe, a man from the northeast who helped out in the kitchen. After they left, the two opened a cafe in Qingdao, and a year later took a stake in a restaurant business in Wuhan, opening several chain bars.

The older women were steadier. Most regulars at Bridge Cafe knew Sister Ma, who had worked there since 2005. Sister Ma is a Beijinger; she did not wash dishes or clean, but sat at the front taking orders and running the register. People remember the little notebook she carried, filled edge to edge with notes on how to work the computer and how to pronounce the names of new dishes. In her ten years at Bridge Cafe, her son went from high school to college, then graduated, got a job, married, and had a child. After her grandson was born, Sister Ma quit to raise him at home. But on weekends when she could not sit still, she still came in to help — riding the bus over from Huixin Beili with her grandson, then riding slowly back at sunset.

If it were not for outside controls and the constant rise in labor and rent costs, this way of life might have gone on.

From 2015, Lili felt distinct pressure from the crackdown on shared rentals. As restaurants usually do, Bridge Cafe provided room and board for staff, normally on the first or second floor of a nearby compound, to avoid disturbing neighbors, with men and women in separate flats. Demolition moved fast around Wudaokou, and the staff dorms changed several times. In 2015 the staff lived on the fifth floor of a residential building in the Dongwangzhuang compound, five people sharing a 60-square-meter apartment. After late-night shift changes disturbed the neighbors, they made one round of adjustments, tearing out the top bunks and keeping only the bottom ones. But before long the landlord no longer wanted to rent to staff, and new places got harder and harder to find. The local subdistrict office also tightened its oversight of shops. "Meetings every couple of days — the subdistrict office holds one, then the police station holds one, once a month, going over all kinds of new policies."

Rising rent and labor costs were another problem. In 2007, staff wages averaged around 2,000 yuan a month; an 80-square-meter apartment near Huaqing Jiayuan cost only a bit over 3,000 yuan a month, and a 228-square-meter ground-floor shop rented for 180,000 yuan a year. By June this year, average wages had risen to around 4,000 yuan a month, a 60-square-meter apartment on Xueyuan Road went for 6,000 yuan, and Bridge Cafe's annual rent had reached 2 million yuan.

This year the three-year lease came up, and Bridge Cafe could no longer carry a rent that was still rising. On the surface business looked brisk — on busy days 600 to 700 customers came through, and on weekends the place was nearly full from 11 a.m. to 10 p.m. — but the average check was only 40 yuan, tables turned over slowly, and the four slow months of winter and summer vacation both ran at a loss.

Looking for another spot around Wudaokou turned up nothing. The storefronts available were either too big or too small, or came with eye-watering transfer fees. Lili also was not sure whether moving would change how people felt about Bridge Cafe. The hotpot restaurant T6, once above Xiangzhufang, moved a couple of years ago to the food court across the street; the space, the interior design, and the menu barely changed, but foot traffic dropped sharply. Whenever Lili went back, something felt missing.

Notice of the closing went up on the cafe's official account a week ahead. On June 29, the last day of business, the place was packed one final time. Some customers called ahead: "We want your carrot cake — save me a slice." The cookie jar, the mugs printed with the logo, and the staff uniforms were all carried off as keepsakes. A teacher from the Beijing Film Academy stopped to talk with Lili and mentioned that she used to pull all-nighters here as a student, and had met her husband while they stayed up writing scripts. A man who drove in from Huairou said he used to have no money and now had his own company, and wanted to do something — he would be willing to invest and go into partnership. A former employee who had quit while pregnant took a cab in from Changping with her child, and the two of them posed for a photo in front of the sign.

Lili also suddenly remembered a boy from Peking University. He always came alone, would call ahead to reserve a spot, and made a habit of paying an extra 50 yuan to keep a small table beside him. Sometimes he came to study in his lab's white coat, sometimes he was lost in a video game, but he had not shown up in the past six months or so. Lili guesses he graduated, and wonders what he would say if he heard Bridge Cafe had closed.

After placing her staff at friends' cafes, Lili rested for two months. She is not worried about making a living — a shared-office company has already approached her to run its food and drink business, and over the years she and her husband have opened several cafes around Beijing. There was one near the Beijing University of Posts and Telecommunications, and another inside the Jingwei Pharmaceutical office building in Fengtai, which they ran for eight or nine years.

Shaped by her years at Bridge Cafe, Lili still prefers opening shops near schools; the latest one opened inside Anhui University of Technology.

3

"It's more or less certain that for the foreseeable future I'll be dancing with the internet, for better or worse." Ten years ago, 21-year-old Huang Haijun wrote this on his personal blog.

The next ten years bore him out: he became the first product manager on NetEase's Weibo project team; he was employee No. 9 on Zhihu's founding team, and its 83rd registered user; and from the end of 2014 until now he has started two companies, one a failure, one still going — moving into and out of Huaqing Jiayuan twice.

On the day his daughter was born four years ago, Huang Haijun registered a domain name using the full pinyin of her name. Four months ago he made the same suggestion to two friends about to become parents: "Making the domain's Whois registration date match the child's birthday forever, witnessed by the internet — that's not a bad little inheritance for an old-school internet person to leave a kid."

In 2005, the year Wang Xing founded Xiaonei, Huang Haijun had just been admitted to Jinan University to study e-commerce. Looking back on college, he described it as "a 20-year-old's body striving for a 30-year-old's way of thinking and acting; nearly 10 hours a day at the computer traded for a faith in the internet."

Huang Haijun was born in Wushan County, Chongqing. When he was nine, the "village school" was merged into the "township school," and only then did his morning walk to school get a kilometer shorter. It was not until he got into a middle school in the Wushan county town that he first met a computer class — a textbook plus a paper keyboard. In 2002, when most of his middle-school classmates dropped out to work, Huang Haijun was the odd one out, getting into a key high school, where he first met a search engine at an internet cafe.

Filling out his college applications, "e-commerce at Jinan University's Shenzhen campus" turned, in his eyes, into a string of tantalizing words: Shenzhen, internet, commerce. They sounded far more interesting than machinery, petroleum, or manufacturing. The summer before he started, Huang Haijun taught himself the Wubi input method, imagining that he was about to shed 13 years of near-total obedience and start stirring things up.

In 2006, when personal blogs were taking off, Huang Haijun started one too, first on Sina, then moving to WordPress. The content tracked whatever was current online, with his views on internet tools, Google's product line, microblogging (Twitter), and trends in social-media marketing. There were also year-end summaries and travel notes, which showed a young man who was intensely curious, highly organized, and strict with himself: in 2012 he drove 3,000 kilometers, read 19 books, and followed more than 3,300 questions and 2,600 people on Zhihu. In 2013 he watched 65 films and TV shows, 40% fewer than the year before.

This blog, first called "E-commerce Blog" and later "Huang Haijun's Internet Diary," gradually built a following in internet circles, and it turned out to be decisive when he looked for a job after graduating. Huang Haijun's dream then was to become a product manager at Tencent, but campus recruiting, info sessions, and internal referrals all led nowhere. Then, unexpectedly, NetEase product manager Lei Xiaoliang, whom he had just met on Fanfou and Gtalk, saw the blog and messaged him out of the blue on Gtalk, asking whether he would be interested in a product designer job at NetEase Beijing. After three rounds of interviews he got the offer he wanted from NetEase's website product department, and headed north at once.

Huang Haijun's first home in Beijing was in a compound called Zan'anchu — "temporary shelter" — on Zhanchunyuan West Road in Haidian district, at 900 yuan a month. The apartment hunt had a geeky flavor: no agents, just social networks. His landlord was a fellow user of Maopao, a social networking site built at the time by Wang Huiwen, now a senior vice president at Meituan. The place was less than a 20-minute walk from the NetEase Building, and came with two cats, Xiaosi and Xiaowu, which gave him a bit of a feeling of home.

But as the name "temporary shelter" suggested, he did not stay long. Five months after joining, Huang Haijun became the second person on the Weibo project team. For a full 40 days the whole team did closed-door development in a five-star hotel near Peking University, seven days a week, more than 18 hours a day. A month later, NetEase Weibo launched. By then, five months had passed since Wang Xing's Fanfou was shut down and Sina Weibo went live.

Over the following 14 months, Huang Haijun poured everything into building and rebuilding Weibo's core features, and the results were poor. In his blog he wrote that the four big portals were locked in a brawl at the time, all pouring star power, traffic, and media resources into their own microblogs. "NetEase was on the back foot; we wanted to differentiate through the product. At the time Foursquare was still the darling of the capital markets, and the mobile internet was nowhere near as widespread as it is now, but many people had already caught the scent of a gold mine. We turned our attention to the dimension of 'location,' upgrading Weibo from 'who's doing what' to 'who's doing what, where.'"

At one Weibo meeting, Tang Yan asked Huang Haijun: "Can this feature let me meet the people in the Aiba cafe downstairs?" Making friends was not Weibo's core goal, and the question caught Huang Haijun off guard: "Uh, not yet, Mr. Tang."

In May 2011, Lei Xiaoliang and Li Zhiwei took that idea and left NetEase, joining Tang Yan to found Momo. Two months before that, Huang Haijun had also left NetEase, becoming Zhihu's ninth employee and, as its only product manager at the time, building private messaging, the timeline, and the bookmarks feature. He felt that while he still had little weighing him down he should take "a few more impulsive risks." Between 2012 and 2014 he took two more, at Douban and Wandoujia, focused on building mobile product experience, anxiety mixed with self-consolation: "I wasn't early to the shift from PC to mobile, but luckily not too late either."

In November 2014, Huang Haijun left Wandoujia and moved into Huaqing Jiayuan, sharing a three-bedroom apartment with an interior design company, and finally began preparing his first venture. The previous tenant, apparently also a startup, had just moved out, leaving no bed or sofa — just rooms full of desks and chairs.

That same month, NetEase Weibo announced it was shutting down; a month later Momo listed on Nasdaq at a valuation of 3 billion US dollars, with headcount grown to 400 and monthly active users at 69.3 million. Caught up in the moment, Huang Haijun thought back to that meeting and his exchange with Tang Yan, and wrote on his blog: "This story feels very close to me — real, inspiring, get-rich-quick... This is why I love the internet so much. Here, dreams feel within reach."

But for most people the dream only ever gets close.

Huang Haijun's first product was called Mojing. It was a photo community built on location: users found people nearby, rated their looks, and got to know each other through the rankings. In the ten months after its June 2015 launch, the team grew from 9 people to more than a dozen, raised 1.5 million yuan in angel funding, and moved from Huaqing Jiayuan into a shared office in Zhongguancun.

But as with Xu Can, the money ran out fast, and the project never gathered enough users, let alone a sustainable way to make money. Looking back, Huang Haijun points to the rising cost of acquiring app users, fiercer competition in making money from social apps, and problems in how the product team was organized. At bottom, though, this was a product that died of unclear positioning — whether for hookups or for showing off good looks, Momo, Meitu, or WeChat Moments had already filled the gap. Mojing shut down in April 2016.

Just a month later, Huang Haijun began his second venture. A real-estate agent he had met at Huaqing Jiayuan a year earlier, practiced and shrewd, called him at exactly the right moment: "Still starting companies? Thinking about moving back? I've got a place." Like Xiao Ji, these agents keep their own timetable in their heads, tracking entrepreneurs' movements through their Moments posts.

Huang Haijun's second project was called Zhirenshe. It came entirely out of his own experience: helping startups find people, and helping internet job-seekers with years of experience find jobs. He kept a tight rein on costs — no app, just a WeChat official account plus a few dozen WeChat groups. The team is now under 20 people, more than half of them professional headhunters.

In official terms, Zhirenshe serves only internet companies at the Pre-A stage or later, for jobs paying a cash salary (excluding shares and options) above 250,000 yuan a year, and offers recruiting through social platforms like WeChat, Zhihu, Douban, and Weibo — free distribution, pay only for results. That also includes two offline meetups a month, and job-seeker interviews published online on a regular schedule.

"The backbone workers with 4 to 6 years of professional experience are in short supply. On the other side, professionals with more than 3 years of experience don't trust job boards — they won't submit resumes there, and they put much more faith in a referral from someone they know." In other words, Huang Haijun wants the acquaintance-referral market that headhunters and job boards have not covered.

The business is not especially innovative, but it lasts; it is hard to scale, but it can turn a profit without a big round of funding. At least Zhirenshe can now afford to move out of Huaqing Jiayuan again, into an office building in Zhongguancun. Once more it has moved into the traces left by another startup of uncertain fate: desks, chairs, a typewriter, a water cooler, potted plants. Over the green wall the previous tenant had carefully painted, Zhirenshe's brand-new blue logo now stakes its claim.

Huang Haijun now looks more like an observer, tallying the hiring needs of startups at different funding stages, listening to CEOs speak at meetups, sharing the latest internet jokes and in-depth product articles on his Moments. Zhirenshe's community still has room to grow — three or four offline events a month, and the community itself might become a product. Until that is done, Huang Haijun does not plan to think far into the future.

In the middle of last year he set aside, for now, the writing outpost he had kept for 12 years, moving his attention from his own blog to a WeChat official account. "Huang Haijun's Internet Diary" stopped on July 13, 2017. That last post was titled To Think a New Product Through, First Answer These 17 Questions. In it he wrote: "Unique timing and historical background — especially huge era-level factors — often give rise to unique product models. As the saying goes, the times make the hero."

(At the interviewees' request, Xu Can and Sanshui are pseudonyms.)

Cover image from The Social Network