Part IThe Story
When the market opened in New York on July 26, 2018, it was already night in Shanghai, and the man whose company was about to be priced at $24 billion was not in New York. He was in Shanghai, standing with a group of investors and customers at a "simultaneous remote bell-ringing ceremony." Most founders treat the ceremony on the balcony above the trading floor as the one photograph they will frame. Colin Huang skipped it. Pinduoduo, three years old, listed at $19 a share and raised $1.6 billion. Bloomberg's index put Huang's own worth at $9.89 billion, which made him, on that day, China's youngest self-made billionaire.
He gave two explanations to the mainland press, and they did not quite agree. The first was egalitarian: "We wanted to let more of our customers take part in ringing the bell, but it might have been a bit of a hassle to arrange visas for all of them to travel to the US. Isn't it better that consumers, investors, all of us, myself included, be here?" The second was medical. He preferred to avoid long-haul travel, he said, because he did not want an ear infection to come back.
One explanation was about the customers and the other was about his ear. The truth was probably some of each, along with a third reason he did not give, which was that Huang has spent his career declining to stand where people expect to find him. A Reuters photograph from that day was captioned, in full seriousness, as showing Huang speaking "during the company's stock trading debut at the Nasdaq Stock Market in New York, during an event in Shanghai." That caption puts him in two places at once and pins him down in neither, and it describes the subject of this profile better than most of what has been written about him.
That same month, July 2018, roughly a thousand merchants who sold on Pinduoduo gathered outside its headquarters to protest the platform's new penalties for counterfeits, and some of them fought with the company's security guards. So the bell rang in Shanghai and the sellers shouted in Shanghai. The founder was in the same city as both, and it is hard to say he was really present at either.
Eight years later the outline is easy to recite and hard to explain. Pinduoduo began by selling fruit to people in China's smaller cities. It overtook Alibaba by number of buyers and became the core of PDD Holdings, which is legally domiciled in the Cayman Islands, headquartered in Dublin, and staffed mostly in China. Its export sibling, Temu, bought Super Bowl ads telling Americans to "shop like a billionaire." Huang himself left. He resigned as CEO in 2020 and as chairman in 2021, gave the voting rights on his shares to the board, and said he would study food science and life sciences. In August 2024 he became the richest man in China anyway, and within weeks he no longer was.
The NPR correspondent Emily Feng, who covers China, summed up how little is known about him in one sentence: "We don't know if he's married, if he has children, what he does with his spare time." She also noted that he does not make public appearances. Even in 2018 he did not travel to the most public appearance a founder gets.
By the Numbers
The Pinduoduo Machine
$54.0BPDD revenue in 2024, up from $0.3B in 2017
$15.4BNet income in 2024, after losses every year through 2020
788.4MAnnual active buyers at end of 2020, edging Alibaba's 779M
RMB 4.17TGross merchandise value in 2021 (about $590B)
$1.6BRaised in the July 2018 Nasdaq IPO, three years after founding
9Posts on Huang's WeChat blog, Feb 2016 to Sept 2017, which laid out his theory of the business
$0Import taxes Temu appears to have paid in 2022, per NPR, versus about $700M for Gap
Tian Ji's Horses
There is an old Chinese story about a general named Tian Ji who was always losing chariot races to his king, because the king's horses were better at every level: the king's fast horses beat his fast horses, and so on down. A strategist advised him to change the order. Tian Ji ran his slowest horse against the king's fastest and lost badly. Then he ran his fastest against the king's middle horse and his middle against the king's slowest, and won both. He took the match two races to one without owning a single better horse.
Huang has used this story to explain his own life. In an autobiographical essay called My Middle School and University Years, he described the lesson as "local advantage": you can be weaker overall and still win by being stronger at the specific point where the contest is decided.
He had reasons to think in those terms. He was born on January 1, 1980, on the outskirts of Hangzhou, to parents who both worked in factories. His father never finished middle school. Accounts in Chinese describe a household that was not poor but was always short of money, and a boy who wore clothes handed down from his mother's coworkers and from relatives' children. Hangzhou would later become known as Alibaba's hometown, which adds some irony to the story. In the early 1990s it was a provincial capital, and the Huangs lived on its edge.
His elementary school, by his own description, was "ordinary." He was good at math, and good enough to win a prize in a math olympiad. That prize was the first of his lucky breaks. A teacher suggested he take the entrance exam for Hangzhou Foreign Language School. He didn't want to go, because he assumed a foreign-language school would be about languages, and he wanted "a school strong in Math, Physics and Chemistry." The principal talked him into it. He was twelve. Years later, on a blog that has since been taken down, he wrote: "In retrospect, [thank] goodness that I chose to go to HFLS." He described the school as known for its liberal approach.
His classmates there were the children of Hangzhou's political and social elite. Several English-language accounts repeat that one of them was a mayor's daughter. The specific classmate matters less than the change in his surroundings. The factory workers' son now had a view into the class above him, which, as China's economy started to accelerate, was about to grow much larger. He described himself to an interviewer, with a self-awareness that also sounds a little like boasting, as a "phoenix soaring out of a chicken coop."
He also expressed regret, in what may be the most personal thing he has said in public: "I was too goal-oriented and wasted too much time striving to be number one in class and a good student." He said he had not taken enough time to "purely enjoy youth." The man who would build a shopping app shaped like a game felt that his own childhood had been all competition and no play.
The idea of local advantage runs through everything that followed. Alibaba and JD.com had already won the cities, the brands, the logistics networks, and the payment systems by the time Huang started. They had the faster horses. He looked for the race they were not running: fruit, smaller cities, WeChat, people who had never bought anything online. He put what he had against the parts of the market they had left alone.
An Answer in an Online Forum
Huang's network did not come from family, from the Party, or from his alma mater. It came from a reply he posted online.
As an undergraduate he answered a coding question in an internet forum, and the person who had asked it turned out to be William Ding, founder of NetEase. Ding was an engineer who had built a fortune in online games; NetEase later became one of China's most successful game companies. According to the investor Turner Novak's reconstruction of Huang's early years, Ding was impressed enough to introduce the student to other figures in Chinese tech: Pony Ma of Tencent, whose WeChat would later carry Pinduoduo; Wang Wei of SF Express, the logistics company; and, in the summer before Huang left for America, Duan Yongping.
Duan, like Huang, is a Zhejiang University graduate. He founded BBK Electronics, the company behind OPPO and Vivo, and stays out of public view as much as Huang does. In China he is often called its Warren Buffett, partly for his investing and partly for his habit of reducing hard problems to plain moral language. The word associated with him is benfen, roughly "doing what is right and proper to one's station," and it later became part of the vocabulary around Pinduoduo. Huang has called Duan the most important mentor in his life. Novak reports that Huang helped Duan with his investing, so the relationship ran in both directions.
These connections came at a time when Huang's world was widening quickly. He entered Zhejiang University's Chu Kochen Honors College in 1998, at eighteen, to study computer science. One account places him in the lab of Pan Yunhe, then the university's president and an academician of the Chinese Academy of Engineering. In his first year he was chosen as a fellow of the Melton Foundation, a program funded by Bill Melton, the American who founded VeriFone, the company behind the card-swipe terminals in shops around the world. Fellows received a computer, an internet connection, and a trip each year to another member country. Huang later credited the fellowship with giving him a more international outlook than most Chinese people of his generation had.
Then came Microsoft, first in Beijing and then in Seattle. The internship stipend is reported as 6,000 yuan by one account and $900 a month by another. Every version agrees on the point Huang himself made: it was more than his parents earned together. In Seattle it was reportedly more than $6,000 a month. A boy who had worn hand-me-downs now earned in a month, as an intern, more than his household had earned in a year.
He moved to the United States in 2002 for a master's in computer science at the University of Wisconsin–Madison and finished in 2004. A professor wrote him recommendation letters to the large technology companies of the day, and he received offers from Oracle, Microsoft, and IBM. Duan advised him to go to a young company in California that had not yet gone public. He turned down the safer offers and joined Google.
Lunch at $620,100
Google in 2004 was about to go public and about to become very rich. Huang worked as a software engineer on early search algorithms for e-commerce, which was an apt first assignment, and was soon promoted to product manager. In the now-deleted blog post he wrote about that period, he described it the way an engineer would describe a system: "The operating profits and the number of [employees] grew rapidly," and his own bank balance "suddenly grew a lot too."
He was more candid with Forbes in 2018: "It was a very good experience for me to go from a very poor background to all of a sudden [being among those who had] become extremely rich. It's exciting, but sometimes it also has some not so good side effects." He never said what the side effects were. Much of Huang's public record works this way: he states the result and leaves out the process.
In 2006 Duan Yongping won the annual charity auction for lunch with Warren Buffett, bidding $620,100. He brought along his wife, several friends, and Huang, who was twenty-six. There is a photograph of the young engineer standing next to the most famous investor in America. People who later tried to explain Pinduoduo kept coming back to that lunch, and Huang did too.
What he said he learned was neither a stock tip nor a strategy but a way of thinking: "A person's mind is easily corrupted. When you are making a judgement about something, you need to understand the background and the facts. After understanding them, what you need is not cleverness, but the courage to use reason and common sense when facing the facts. Common sense is self-evident and easy to understand, but we are blinded by the various biases formed through our growth and learning, and by the pursuit of our personal interests."
It is worth noticing who said this. Huang had spent his life being clever: olympiad prizes, elite schools, being "number one in class," and an engineering job at the company that had made cleverness its brand. Now he was saying that cleverness was a distraction and the real difficulty was having the nerve to state the obvious. Novak's account lists three things the lunch reportedly gave him: the power of simplicity, the use of float, and an interest in redistributing wealth, since Buffett had pledged nearly all of his fortune to charity. All three became central to Pinduoduo. Ten years after the lunch, Huang would write an essay arguing that Buffett's model should be turned upside down.
The same year, 2006, Huang went back to China with Kai-fu Lee to help build Google China, and found it difficult. One account says he grew tired of flying back and forth across the Pacific to settle small problems. Given what he later said about his ear and long flights, that detail stands out. He left in 2007. Only years later did he see what he had walked away from: "It was not until three or four years after I left Google, I started to realize how rare it was to come across and join a company like Google at the time. It was a once-in-a-lifetime opportunity (or at least once in 10 or 20 years) and I was extremely lucky."
He talks about his luck often. The olympiad, the principal, the forum reply, the Google offer, the lunch: he describes nearly every turn in his life as luck. That is either real humility or the habit of a man who prefers not to be credited with his own decisions. It may be both.
Four Companies and an Ear
The usual summary is that "after several moderately successful ventures in gaming and e-commerce, Colin Huang got sick and retired." That sentence, from Bloomberg, covers about eight years.
The first company was Ouku, founded in 2007. It sold consumer electronics online, mostly phones and electronic learning devices, and largely acted as a distributor for products from BBK and OPPO. In other words, it sold Duan Yongping's goods. Huang sold it in 2010 for $2.2 million. That is a respectable amount, and a modest one next to what came later.
The second company was Leqi, sometimes spelled Lequee. It was an e-commerce agency that helped brands, including foreign ones, sell on Alibaba's Taobao and on JD.com. This is the experience that taught him the most. Leqi put Huang inside the incumbents' platforms as a hired operator, where he could see how traffic was bought, how merchants were squeezed, and where the systems were weak. Running a shop on someone else's marketplace is the best way to learn what that marketplace gets wrong.
The third company was a game studio, Xunmeng. This is where the record gets unreliable. Pinduoduo's own materials have said Huang founded an online games company called Xinyoudi in 2011. The Financial Times reported in 2020 that no company by that name had ever existed. The confusion is minor and probably clerical, but it fits a pattern. Even the founding stories of Huang's own companies tend to come out slightly blurred.
Then he got sick. An ear infection was serious enough that around 2013 he stopped working, at about thirty-three, with money in the bank. By Bloomberg's account he stayed home for a year thinking about what to do next. Nobody has described that year in detail, and Huang has not filled it in. All we have is the outline: a man who had already built and sold companies, who had studied Buffett, who had seen the incumbents' platforms from inside, sitting still and thinking.
When he started working again, he combined all three previous companies into one. Ouku had taught him e-commerce. Leqi had taught him how Taobao and JD.com operated. Xunmeng had taught him what keeps people engaged with a screen. Cathay Innovation, the first global fund to invest in Pinduoduo, put it plainly: Huang wanted to "leverage his unique blend of experiences in e-commerce, modern search, and gaming." Pinduoduo was literally incubated inside the game studio. It launched in September 2015 as a project within Xunmeng Games.
The ear also stayed part of the story. Five years later he cited it to explain why he would not fly to New York to ring the bell.
Nine Posts
In February 2016 Huang started a personal WeChat blog. He stopped posting in September 2017, after nine entries. The originals are gone, but people across the Chinese internet had copied and reposted them, so the record survived without his help. That is a fitting fate for a man whose company spread by people sharing links.
The posts are personal and wide-ranging, and they are probably the closest thing to a candid self-portrait Huang has produced. One is about his school years and the Tian Ji story. One, inspired by how the First World War forced garment production to scale up quickly, sets out a theory of what he calls supply-side reform. He identifies three ways production gets transformed: by entirely new demand, as when the airplane was invented; by new uses for old goods, as with wartime needs; and by aggregating scattered demand, as ride-hailing did. The third is the one he built a company on.
Another post is about Gresham's Law, the idea that bad money drives out good. He clearly disliked that dynamic, and he wrote about how grassroots participants might keep a market healthy. This is odd in hindsight, because within two years critics were blaming Pinduoduo for exactly that: for "consumption downgrade," for counterfeits pushing out real goods, for cheap and bad products crowding out the rest. Some commentators who read the blog later, including the writer at Jianshan who went looking for it, concluded that the critics had mistaken "consumption substitution," meaning finding cheaper equivalents, for degradation. Others did not.
The final post is titled "把资本主义倒过来," or "Reversing Capitalism." It opens with praise that turns into a challenge:
"Warren Buffett is an admirable capitalist, a pure capitalist through and through. His entire career can be described as a relentless, focused, and rational effort to move money around to enjoy the fruits of compounding interest. I enjoy reading his letters to shareholders. For decades, he has repeated the same simple, yet difficult purity. One end of his empire is built on insurance, and the other is on investment."
His argument, in brief, is that insurance makes money from poor people's need to manage risk. The poor pay premiums and the rich collect the float. Huang proposed what he called "reverse insurance." Ordinary people would pool their future demand and commit in advance to buying something they will need later. His example was buying winter coats in summer. Manufacturers who know with certainty what will be bought can plan production precisely and pass the savings on. In this scheme, the certainty of what people intend to buy becomes a product, and the money flows from capital down to consumers instead of the other way.
The boy who wore other families' old clothes grew up to write about buying winter coats in summer. He never pointed out the connection himself, but it is there.
Years later, the tech executive and newsletter writer Rui Ma explained the same idea to NPR's Planet Money in simpler terms: "So instead of me necessarily having the goods on hand, what if you just wait, you know, a week, two weeks, whatever it is to receive your item? Because then I can take this aggregate and accurate demand, go to the suppliers and say, now you know exactly what to make. Make it." Nick Fountain, the show's host, called it "the most just in time, the most lean operation ever." Ma agreed: "That is the dream."
The blog ended in September 2017. Ten months later the company went public. By then Huang had already published his theory of the business, in nine posts on a platform owned by one of his investors, and then stopped writing.
Fruit, Friends, and the Free Salesforce
The company started with fruit. In April 2015 Huang founded Pinhaohuo, a site that sold fresh produce directly. Fruit is perishable, seasonal, has no brand, and is bought often, which makes it a poor product for a traditional retailer and a good one for testing his theory. If enough people commit to buying the same box of oranges, the orchard can sell straight to them, the middlemen drop out, and the price falls. Pinduoduo, launched that September out of the game studio, applied the same model to general merchandise. In September 2016 the two companies merged.
The mechanism was the team purchase, and the name explains it. Emily Feng gave NPR the translation: "Duo duo means more, more. And pin means to merge. So the whole idea is that you're merging orders. You're bulk buying together with friends." The official English tagline, "Together, More Savings, More Fun," is clumsier and says the same thing. You find a product, you send the link to friends or relatives, and if enough of them join, everyone pays less.
The links traveled over WeChat. Tencent's messaging app reached nearly every smartphone in China, and Tencent became a major Pinduoduo shareholder, owning 17.7% by one count. The two companies had a common interest: Tencent blocked links to Alibaba's Taobao inside WeChat, and Pinduoduo grew in that protected space. Each customer who shared a deal was doing the company's marketing for free, and the people receiving those links were largely in what Chinese analysts call Tier 3 and Tier 4 cities, and in rural areas Alibaba had largely ignored. Many of them had never bought anything online. Cathay Innovation said, in the way of an investor celebrating an IPO, that Pinduoduo had brought "hundreds of millions of people" into digital commerce "in a way that is at the same time both more fun and economically attractive."
Feng's own introduction to the app is a small version of the whole model. She had started boxing in Beijing and noticed a woman at the gym wearing "really, really hot tights." She asked where they came from. "She sent me the link, and I was shocked. These tights were, like, five bucks, and they got cheaper if you bought more than three pairs. So I ended up getting five pairs." A stranger's leggings, a shared link, a group discount, and a purchase larger than she meant to make: the conversion happened in the time between sets.
Pinduoduo's IPO prospectus described all this as "an exemplification of a multidimensional space seamlessly integrating cyberspace and the physical space. A combination of Costco and Disneyland, driven by distributed network of intelligence agents." The "intelligence agents" sound like science fiction until you realize they were people sending each other links over WeChat. The shorthand that stuck was Costco plus Disney: Costco for the value, Disney for the entertainment. The Disney part came from Xunmeng. Pinduoduo filled its app with games, including virtual orchards where users who tended digital trees received real fruit. Huang, who regretted not playing enough as a child, made shopping into play.
He was clear that the point was not luxury. He told Caijing in 2018:
— Colin Huang, interview with Caijing, 2018[Pinduoduo] is not about letting people in Shanghai feel like they are living a Parisian life, but making sure that people in Anhui have kitchen paper and fresh fruits. The goal is not to be cheap, but to make users feel like they got a good deal.
The last sentence of that quote is the precise one. The key word is feel. Low prices were the means, and the satisfaction of having got a deal was what he was actually selling.
Less was said publicly about the financial side. Novak's 2020 analysis described Pinduoduo collecting deposits from suppliers and paying them on stretched terms, so that customer payments sat with the company long enough to create float. Buffett ran on insurance float. In Novak's account, Pinduoduo built its own float out of shopping transactions and spent it on acquiring customers, which was exactly the kind of capitalism Huang had written about reversing.
The growth was extraordinary. The company passed 100 million users in July 2016. In April 2017, less than two years after launch, it raised more than $100 million in a round led by Banyan Capital, at a valuation above 10 billion yuan, about $1.5 billion. Bloomberg reporters who visited the "airy Shanghai offices" that spring found a founder with a buzz cut and "an unrestrained laugh" who expected to double headcount to 1,000. "A few companies have tried this before, but no one has really been able to do it," Huang told them. "We felt we had a competitive advantage." Huang rarely talks to the press, so when he does he uses few words. "We felt we had a competitive advantage" understates it.
In the first quarter of 2018 Pinduoduo sold $10 billion worth of goods to 56 million users. By the end of 2020 it had 788.4 million annual active buyers, slightly more than the 779 million on Alibaba's Taobao and Tmall combined. A company that started by selling oranges through group chats had more customers than the company founded by a former English teacher from Huang's own hometown.
Ten Times the Value
Low prices and no brands came with a reputation for fakes. Chinese media had long criticized Pinduoduo for shanzhai goods, the knockoffs and near-copies that fill China's markets, and after the Nasdaq listing the criticism became official. China's State Administration for Market Regulation opened an investigation. The share price fell. The company published an open letter saying that in a single week in August 2018 it had closed 1,128 stores, removed more than 4 million listings, and blocked 450,000 suspected counterfeit listings before they went live. Later it reported removing more than 10.7 million suspicious items and blocking 40 million suspicious links.
The penalty system that went with these numbers showed how Huang approached problems. Pinduoduo promised to compensate buyers at ten times the value of any counterfeit sold on the platform, three times what China's Consumer Protection Law required. Against sellers, it froze ten times the trading volume of any item found to be fake. The legal scholar Angela Huyue Zhang, in her study of Chinese tech regulation, High Wire, describes how the policy played out. A thousand sellers protested at headquarters in July 2018, the same month as the IPO, and scuffled with guards. Some sued, and by 2021, at least, Pinduoduo had won a significant majority of those cases.
So the platform had sided with its customers against its merchants, and it was willing to fight about it in court and outside its own building. The penalties were set above what the law required so that the company would not have to wait for the law to act. It is a recognizably Huang approach: if you cannot yet earn trust over time, you pay for it with punishments that exceed the official ones, and you accept the anger that causes.
The fixes did not settle anything. In April 2019 the Office of the United States Trade Representative added Pinduoduo to its list of Notorious Markets for counterfeiting and piracy, and as of 2023 it was still there. In January 2019 hackers exploited a loophole in the platform and took coupons worth tens of millions of yuan, and the company reported the theft to police. In February 2020, as the pandemic began, it disclosed removing 500,715 items and closing more than 40 stores to stop sales of counterfeit and substandard masks.
That April, in his first letter to shareholders after the IPO, Huang made a broader argument. Chinese e-commerce, he wrote, suffered from "forced exclusivity." Merchants were made to choose sides, and Alibaba's payment system Alipay and its logistics arm Cainiao were set up to exclude rivals. Pinduoduo, he said, gave merchants and customers choices in waybill systems, cloud services, and payments. Then he wrote the line that served as his mission statement:
— Colin Huang, first letter to shareholders, April 2019Our strategy has never been to disrupt a monopoly in order to create a new one of our own, but to disrupt in order to provide a choice.
He was saying this while the company was on the U.S. government's counterfeiting list, in court with its own sellers, and building a dependence on WeChat, which itself excluded Taobao. Huang either did not notice the contradiction or did not think it needed addressing. The letter was sincere and self-serving at once, which may be the most accurate thing to say about Pinduoduo as a whole. It did offer people a real choice. It also wanted that choice to end up being Pinduoduo.
Einstein in the Annual Report
Huang's letters became stranger as the company grew. Pinduoduo's 2019 annual report, published in April 2020 during the pandemic, contained a founder's letter that read like an essay on physics:
"When this tiny virus was dropped into our world, it acted just like a catalyst in a test tube, accelerating the formation of a whole new world. Inevitably, some dimensions of the previous world are being restructured, some rules are being rewritten... It is the time of reestablishment."
It discussed Newton. It discussed the history of time. It took on Einstein directly: "When Einstein wrote down his famous E = MC², he elegantly (in some sense also arrogantly) depicted a physical world in his mind. However, what he did not explain in his theory of relativity is the relationship between the human mind and the physical world, nor the relationship between energy and information."
Some investors appreciated the ambition. Others said, reasonably, that the letter skipped what they needed to know, given that the company had just reported weaker fourth-quarter revenue than expected and was warning about the first quarter. The numbers were in the filing anyway: RMB 1.01 trillion in gross merchandise value in 2019 from 585.2 million active buyers, average annual spending up 53% to RMB 1,720 per buyer. Huang left the arithmetic to the filing and used the letter for relativity. Around the same time, another of his shareholder letters compared Pinduoduo's path to Yao Ming's, the very tall basketball player whose career showed an unusual body can still find its game.
The pandemic did what the letter predicted. China's lockdowns pushed shoppers online, and Pinduoduo, which had built its business on groceries, became an essential service. It ran a program to help rural farmers sell produce online while physical markets were shut. In August 2020 it launched Duo Duo Maicai, a service for preordering groceries and collecting them at pickup points. It launched a wholesale platform, Duoduo Pifa, with no commissions or promotion fees to attract suppliers. Its share price rose 120% in 2020. Huang owned 43.3% of the company, and in six months his wealth increased by about $25 billion. By the end of June 2020 he had passed Jack Ma to become the second-richest person in China.
A week later he resigned as CEO.
The announcement on July 1, 2020, named Lei Chen as his successor. Chen was the company's CTO and a co-founder, the engineer who had built the systems that Huang's essays described. Huang said he would step back from daily management to focus on long-term strategy and corporate structure. Michael Norris, writing for TechNode, pointed out that strategy and structure are what a CEO is supposed to do. He said institutional investors had not been told in advance, and he described analysts' use of the word "unexpected" as "a polite way of saying, 'We don't know what the bloody hell is going on.'"
More departures followed. In August 2020 Huang resigned his directorship at Hangzhou Weimi Network Technology; on August 6, at Shanghai Xunmeng Information Technology, Pinduoduo's operating entity; and on August 19, at Pinduoduo (Shanghai) Network Technology. Meanwhile he gave away equity. He donated 2.37% of the company to a charitable foundation and 7.74% to the Pinduoduo Partnership, a group of senior managers modeled on the partnership structures other Chinese tech companies had used to keep control in place after a founder left. By June 2020 his stake was down to 29.4%. The foundation was the Starry Night Charitable Trust, set up with the founding team to "support fundamental research in biomedical science, agriculture, and food." The Hurun Report ranked him China's leading philanthropist of 2021 after he pledged $1.85 billion, and the foundation gave $100 million to Zhejiang University.
On March 17, 2021, the same day Pinduoduo reported having more annual buyers than Alibaba, Huang resigned as chairman and handed the voting rights on his shares to the board. His letter explaining it was brief: "After resigning as chairman, I will focus on my lifelong interests and devote myself to research in the fields of food science and life sciences."
The timing is notable in hindsight. Late 2020 and 2021 were when Beijing turned against its tech billionaires. Ant Group's IPO was halted, Alibaba was investigated and fined, and Jack Ma largely disappeared from public view. Huang's fortune peaked at $71.5 billion in early 2021 and then fell 87% in about a year as the pandemic boom ended and the crackdown spread. The record does not show what he expected or knew. It shows a sequence: he got rich very fast, gave up the CEO title, gave away equity, created a partnership, endowed a science foundation, gave up the chairmanship, and handed over his votes. For someone who prefers not to be seen, it amounted to taking himself off the list of founders a government might decide to target.
He kept the money. That is the part the news coverage of his "quitting" tended to omit. He had left the stage without leaving the company.
Shop Like a Billionaire
In September 2022 PDD launched Temu in the United States. The name has been read as "team up, price down." NPR's Nick Fountain asked the company how to pronounce it, and a spokesperson told him it "has many pronunciations, but most people go with Temu." Fountain was baffled: "You're the company. You should tell me how to pronounce it." Feng summed up the company: "Yeah, they don't like sharing a lot."
The app was the same model translated into English. "The factories are the same. The producers are the same," Feng said. "They're just repackaging the same app, turning it into English and calling it Temu." Rui Ma remembered buying headphones on launch day for $2 and guessed she was among Temu's first hundred American customers. The games came too. Users fed a digital pet fish in a game called Fishland to earn discounts, and browsed listings like the "vomit egg yolk egg liquid pinch-happy decompressing toy," priced at $1.43. Fountain, testing the app for his show, ended up in what his co-host called "a shopaholic fugue state," ordering about $400 of listed value for $77.88 with free shipping, including three bike bells.
What made the prices possible was partly Huang's demand-aggregation theory and partly a U.S. customs provision from the 1930s. The de minimis exemption let packages worth less than a threshold value into the country without duties or formal declarations. It was originally meant for travelers mailing souvenirs home. In 2015 Congress raised the threshold from $200 to $800. One of the members of the House Ways and Means Committee who voted for the increase was Representative Earl Blumenauer, an Oregon Democrat who had been Portland's commissioner of public works and who had given out roughly 50,000 bicycle lapel pins over his career. His pins, he admitted to NPR, were made in China. "We have supply chain issues," he said.
"Nobody had a clue that there would be a flood of commerce from Asian countries, primarily, that would overwhelm the system," Blumenauer said. "This completely blindsided Congress, blindsided me." By 2024 roughly a billion de minimis packages a year were entering the United States, about 60% of them from China, and estimates had Temu shipping as many as a million packages a day. In 2022 Gap paid about $700 million in import taxes and H&M about $200 million. As far as NPR could determine, Temu paid nothing. When Fountain showed Blumenauer a bedazzled bicycle pin on Temu for 98 cents, the congressman answered: "There's a lot you can do with slave labor."
He was not saying Temu used forced labor. His point was that, given how the system was set up, nobody could check. A congressional report found that Temu "conducts no audits and reports no compliance system" to verify compliance with the law barring goods from Xinjiang. Senators Tom Cotton and Marco Rubio wrote to the Biden administration about forced labor and intellectual property. At least twenty states filed consumer-fraud suits. Temu said its third-party code of conduct "forbids any form of forced child or penal labor."
Problems surfaced in China as well. In March 2023 Google suspended the Pinduoduo app from its Play Store after a Chinese security firm found malware in versions distributed through Chinese app stores. A Kaspersky researcher told Bloomberg that some versions "exploited known Android vulnerabilities to escalate privileges, download and execute additional malicious modules, some of which also gained access to users' notifications and files." Six security teams consulted by CNN, from Finland, Russia, the United States, and Israel, along with a Chinese firm, classified the app as malware or potential malware. Two days after Pinduoduo issued an update addressing the issue, it disbanded the team of engineers and product managers responsible and moved most of them to Temu. In 2024 the Financial Times and the Wall Street Journal reported that the company had sued former employees for breaking non-compete agreements, using video of them going to work at rival companies as evidence. The company said it did not "engage in any illegal or unethical surveillance practices of current or former employees." Bloomberg mentioned, almost in passing, that Huang had "set a punishing work schedule for his own employees."
Meanwhile the company moved its legal base. In February 2023 Pinduoduo became PDD Holdings, and that year it moved its domicile from Shanghai to Dublin. The man who would not fly to New York had a company with its legal home in Ireland, its shares in New York, its staff in China, and its packages landing in American mailboxes.
On March 6, 2024, the Financial Times's Dan McCrum, an investigative reporter best known for his work on Wirecard, published a long article titled "The mysterious rise of the Chinese ecommerce giant behind Temu." It raised questions about PDD's thin disclosures, its pseudonymous staff, and why its balance sheet had not grown in line with revenue. The stock fell. Robert Wu, a China analyst who had met Huang and his team years earlier over a business deal, wrote a response granting the secrecy and rejecting the suggestion of fraud: "They are always the David in a market where a Goliath already dominated. Before, their Goliath was Alibaba. Now that Goliath is Amazon." In his view, secrecy was simply how an underdog protected itself. Both views are plausible, and the company did not try to settle the question.
Through all of this the money kept growing. PDD's revenue was $34.9 billion in 2023 and $54.0 billion in 2024, with net income of $15.4 billion, from a company that had lost money every year through 2020. On August 8, 2024, Huang, then forty-four, became China's richest person according to Bloomberg's index, with a fortune of $48.6 billion. He replaced Zhong Shanshan, the bottled-water magnate who had held the top spot since April 2021, and he was the first tech founder to hold it since Beijing's crackdown began. "Ma and Jeff Bezos have been corporate leaders in their moments," said Brock Silvers of Kaiyuan Capital, "but the times have changed and Huang is seeing great success with a different, less visible approach."
The title lasted only weeks. Later that August PDD reported disappointing second-quarter revenue, and Lei Chen, by then co-CEO, warned of "many new challenges ahead from changing consumer demand, intensifying competition, and uncertainties in [the] global environment." The company lost about $50 billion in market value in one day. Huang's net worth dropped $14.1 billion, to $33.5 billion. By May 2025 Forbes estimated it at $39.1 billion. Small self-published Kindle biographies, such as The Inspiring Journey of Colin Huang: From Factory Floors to Silicon Valley and Beyond, tried to turn his life into a rags-to-riches story. He did not participate.
On Planet Money, after the questions about de minimis, forced labor, and the future of retail, Fountain opened his Temu package: nineteen items, including the three bike bells. He tested them one at a time on air. The electric one was "pretty annoying." The second was "perfect," and he said it would go on his bike that week. The third, he said, was "basically just a useless piece of plastic and metal that was flown from across the world just so I could throw it away."
Six years earlier, a man with a bad ear had chosen not to cross that ocean to ring a bell, and had brought his customers to a room in Shanghai to ring it with him. Now his customers rang his bells on their own, in places he will probably never visit. One bell went on a bicycle and one went in the trash.
Part IIThe Playbook
The principles below are drawn from what Huang did, not from what he said about himself, although the two overlap more than usual for a founder who rarely speaks. Several are uncomfortable. Pinduoduo's record contains real innovation alongside counterfeits, malware, punishing work hours, and aggressive legal action against former employees, and a playbook that left those out would be dishonest. Each principle below is meant to be useful without being an endorsement.
The principles
- 1.Win the race your opponent isn't running.
- 2.Collect demand before you summon supply.
- 3.Make the customer the channel.
- 4.Enter through the lowest-status product.
- 5.Sell the feeling of the deal, not the price.
- 6.Write the theory down before you need it.
- 7.Borrow judgment from people who've already won.
- 8.Compound your apprenticeships.
- 9.Buy trust with penalties the law doesn't require.
- 10.Treat secrecy as an underdog's weapon, and know its price.
- 11.Leave the stage before the stage becomes a target.
- 12.Arbitrage rules written for another era, and plan for the day they change.
Principle 1
Win the race your opponent isn't running
The story of Tian Ji's horses is Huang's own account of how he got ahead, and it is the strategic core of Pinduoduo. In 2015 the received view was that Alibaba and JD.com had locked up Chinese e-commerce. They had. They had locked up the parts they were competing for: urban customers, branded goods, search-driven shopping, and their own payment and logistics systems. Huang did not compete with them there. He matched his strengths against the markets they had left alone: fruit, smaller cities, social sharing inside WeChat, and people who had never shopped online.
This is a more disciplined idea than "find a niche." A niche is a small market. Local advantage means finding the place where this particular contest will be decided and concentrating everything there, even if you are weaker everywhere else. Pinduoduo reached 788.4 million buyers without beating Alibaba on logistics, brand relationships, or payments. It won on cost of acquisition, through referral links, and on price, through group purchasing.
The ordering is what matters. Tian Ji did not get faster horses. He chose which of his horses would lose. Huang chose to lose on brand prestige, on reputation for quality, and on the opinions of urban media in order to win on scale.
Tactic: List the three dimensions on which the incumbent is unbeatable, decide in writing to concede them, and put all your resources into the one dimension where you can win.
Principle 2
Collect demand before you summon supply
The "Reversing Capitalism" essay is, if you remove the philosophy, a description of how to restructure a supply chain. The ordinary retail model forces manufacturers to guess demand, build inventory, and carry the risk of guessing wrong. Huang's alternative is to gather demand first, whether through group purchases, preorders, or consumers willing to wait a week or two, and only then go to factories. In his example, people buy winter coats in summer and manufacturers schedule production around orders they already have.
This also explains why so many odd products appear on Temu. When a factory can test ten designs instead of committing to one, every listing works as a small market test, which is how Fountain ended up comparing three bike bells. In this model the platform's job is not to stock goods but to tell producers what to make. The C2M label, consumer-to-manufacturer, describes the direction of that information: it runs from consumers to factories.
⇄
Supply-driven vs. demand-driven retail
How Huang's 'reverse insurance' rearranges who carries risk and who captures value.
| Dimension | Conventional e-commerce | Huang's model |
|---|---|---|
| Starting point | Manufacturer forecasts demand Risk | Platform aggregates committed demand Scale |
| Discovery | Shopper searches for a known item | Friends share links; feed proposes items |
| Customer acquisition | Paid traffic to the highest bidder | Customers recruit customers Scale |
| Inventory | Built ahead, stored in warehouses | Ideally made after orders arrive Try |
| Who funds the float | Policyholders, in Buffett's insurance model | Supplier deposits and payment terms, per Novak |
| Customer patience | Two-day delivery expected | A week or two traded for price Scan |
Tactic: Before building inventory or capacity, find a way to get customers to commit to purchases in advance, and price that commitment low enough that they actually prefer to wait.
Principle 3
Make the customer the channel
Pinduoduo's growth came from people sending links. The team-purchase discount only worked if a customer recruited other customers, so the incentive to share was the product itself rather than a bonus added later. Feng's five pairs of tights, bought because a stranger at a boxing gym sent her a link, show the whole mechanism in one transaction.
Social commerce had mostly failed in the U.S. As Bloomberg noted in 2017, Facebook's and Twitter's buy buttons were dropped because people don't want to be sold to while socializing. Huang reversed the order. Pinduoduo was a shop with social features, not a social network with a shop attached, and the social behavior was something users did for their own benefit. They were not being advertised to. They were saving money by recruiting their friends.
There was also a structural reason it worked. WeChat had blocked Taobao links, so Pinduoduo was spreading through a network where its largest competitor couldn't easily follow. The choice of channel was part of the strategy.
Tactic: Build a product feature that is only valuable when users bring in other users, and launch it on a distribution channel your largest competitor is shut out of.
Principle 4
Enter through the lowest-status product
Pinduoduo started with fruit, through Pinhaohuo, a self-operated produce seller. Fruit was unattractive to incumbents. It had no brands, spoiled quickly, and produced small orders. Those same traits made it a good entry point. People buy fruit frequently, so they come back often. It has no brand, so there was no premium to protect. It is perishable, so cutting out middlemen produced a noticeable improvement in freshness and price.
Fruit also built a reputation that turned out to be valuable. Pinduoduo became, by Wikipedia's description, China's largest agriculture platform. It helped farmers sell during the 2020 lockdowns, launched Duo Duo Maicai for grocery pickup, and received the state honor of "National Advanced Collective in Poverty Alleviation." Selling kitchen paper and fruit to people in Anhui looked like a weak position to Shanghai analysts. It looked very different to policymakers concerned with rural development.
Tactic: Pick a product that incumbents look down on but that customers buy every week, and use it to build the habit and the reputation before you expand.
Principle 5
Sell the feeling of the deal, not the price
"The goal is not to be cheap, but to make users feel like they got a good deal." In Huang's sentence, the important word is feel. Pinduoduo, and Temu after it, put enormous effort into the experience of getting a bargain: games, virtual orchards that deliver real fruit, Fishland, inflated list prices against low actual prices, and free shipping. Fountain's purchase of $400 in list value for $77.88 was designed to feel like a win, and it worked. He said he hadn't slept in days.
This is the Disney half of "Costco plus Disney," and it is the half rivals had more trouble copying. Alibaba and JD.com both released group-buying apps, but copying a mechanic is easier than copying a culture. Huang had built a game company before he built a shopping company, and he understood that people will keep playing something that gives them small, frequent rewards.
There is a moral cost here, which NPR pointed out: some of what gets sold is wasteful, and the third bike bell went straight in the trash. The principle still holds. Sometimes the thrill of the bargain is worth more to customers than the bargain.
Tactic: Find the moment in your purchase flow where the customer realizes they got value, and design it on purpose, with feedback, progress, and a reward, rather than leaving it to chance.
Principle 6
Write the theory down before you need it
Huang wrote nine blog posts between February 2016 and September 2017, before Pinduoduo went public, and in them he set out nearly the whole business: local advantage, supply-side reform through demand aggregation, the risk of bad goods crowding out good ones, and reverse insurance. When the company later did things that looked unusual, the reasoning was already available, written down by the founder before anyone was paying attention.
That gives a company two kinds of discipline. Internally, a written theory gives a fast-growing team something to check decisions against without the founder in every meeting. That matters a great deal if the founder intends to leave, as Huang did within five years. Externally, it lets you appear consistent while saying very little, since the theory is published and the ongoing details stay private. Huang's later shareholder letters, with their Einstein references and the Yao Ming comparison, were less useful to investors because the useful theory had already been written.
The posts also included Huang's doubts, such as the Gresham's Law concern about bad driving out good, which critics later turned against him. Writing honestly has that risk. It gives critics something to quote, but it also tends to be the thing people remember.
Tactic: Before you scale, write a short public essay explaining why your model should work, including the conditions under which it would fail.
Principle 7
Borrow judgment from people who've already won
Huang's mentors came one after another: William Ding through an answer in a forum, Duan Yongping through Ding, and Warren Buffett through Duan's $620,100 lunch. Each one shaped a later decision. Duan steered him to pre-IPO Google over Oracle, Microsoft, and IBM. Duan's company's products filled Ouku. Buffett gave Huang ideas about float and simplicity, and a model he later decided to invert. Ding's company was in gaming, and Pinduoduo was eventually built inside a game studio.
What Huang took from Buffett was not a playbook but a way of thinking: "what you need is not cleverness, but the courage to use reason and common sense when facing the facts." For a former math prodigy, giving up cleverness was a real change in habit. Mentors are most useful when they make you question what you are already good at.
The relationships were also reciprocal. By Novak's account Huang helped Duan with his investing. That suggests mentorship works best when the student has something to offer back, even if it's only good questions.
Tactic: Find one person who has already won in a field next to yours, offer them concrete help before asking for anything, and pay attention to which of your strengths they tell you to rely on less.
Principle 8
Compound your apprenticeships
Each of Huang's earlier companies taught him something Pinduoduo needed. Ouku taught consumer electronics e-commerce and distribution. Leqi taught him how Taobao and JD.com worked from the merchant's side, including where they were weak. Xunmeng taught him engagement, and Pinduoduo was literally incubated inside it. Google had taught him search for e-commerce. His year off with an ear infection gave him time to connect it all.
Cathay Innovation's description, a founder who wanted to "leverage his unique blend of experiences in e-commerce, modern search, and gaming," is accurate. Pinduoduo was not an idea that appeared all at once. It was a combination of three prior companies, each of which had taught Huang something the others had not. The $2.2 million sale of Ouku was not much of a payout, but it was valuable training.
Leqi stands out. Working as a paid operator inside your future competitor's platform lets you learn its weaknesses without having to attack it yet.
Tactic: If your next venture is a step toward a larger one, choose it for what it will teach you about the market you eventually want to enter, ideally from inside the incumbent's own system.
Principle 9
Buy trust with penalties the law doesn't require
After the 2018 counterfeiting scandal, Pinduoduo set compensation at ten times an item's value, three times the legal requirement, and froze ten times the trading volume of sellers caught selling fakes. The goal was to win trust faster than a reputation could build on its own. The cost was high: about a thousand sellers protested at headquarters, there were fights with security guards, and there were lawsuits, most of which Pinduoduo won.
The approach shows the trade-offs a two-sided marketplace faces. Huang protected buyers at the expense of sellers, because buyers were the scarcer resource and the one competitors were fighting over. Penalties above the legal standard let the company show seriousness without waiting for regulators. But penalties don't substitute for the underlying product. Pinduoduo remained on the USTR's Notorious Markets list for years, and the same issues followed Temu into the United States.
!
Pinduoduo's trust ledger
Interventions and their recorded outcomes, 2018–2024.
| Action | Scale | Result |
|---|---|---|
| One-week takedown, August 2018 | 1,128 stores; 4M+ listings | Regulator probe continued Risk |
| 10x buyer compensation | 3x the legal mandate | Buyer reassurance Scale |
| 10x seller volume freeze | Platform-wide | ~1,000-seller protest; most suits won Try |
| Mask counterfeit sweep, Feb 2020 | 500,715 items; 40+ stores | Pandemic credibility Scale |
| USTR Notorious Markets listing | From April 2019 | Still listed as of 2023 Risk |
Tactic: When a trust problem threatens your marketplace, set a penalty that is clearly above the legal minimum, apply it publicly, and decide ahead of time which side of the market will bear the cost.
Principle 10
Treat secrecy as an underdog's weapon, and know its price
PDD is unusually secretive. It discloses little in earnings, its staff reportedly use pseudonyms, its apps resist data scraping, and it enforces non-competes hard enough to use video evidence in court. Robert Wu's explanation is the standard defense: an underdog fighting an Alibaba or an Amazon gains nothing by showing its numbers. Temu's results were deliberately folded into consolidated figures.
The costs are real and can be counted. When the Financial Times published its March 2024 investigation, the lack of disclosure made the company look suspect, and billions of dollars in market value disappeared before analysts could respond. Investors were not told Huang was leaving as CEO, which Norris argued "dents confidence." The malware episode, followed two days later by reassigning the responsible team to Temu, showed what happens when a secretive organization has to explain itself under pressure. It can't do so convincingly.
Huang's own privacy, with no known public life and no appearances, sat on top of this corporate secrecy. The two reinforced each other. Silvers' phrase "a different, less visible approach" is accurate. The approach worked until the moment the company needed people to trust it, and then it became a liability.
Tactic: Keep competitively sensitive information confidential, but decide ahead of time what you will disclose when your credibility is questioned, because you won't have time to work it out during the crisis.
Principle 11
Leave the stage before the stage becomes a target
Huang's exit took place over about nine months, and the order of steps was deliberate.
→
The staged departure
How Huang removed himself from operating control while keeping his economic stake.
June 2020
Stake reduced to 29.4%: 2.37% to the Starry Night foundation, 7.74% to the Pinduoduo Partnership.
July 2020
Steps down as CEO a week after passing Jack Ma in wealth; CTO Lei Chen succeeds him.
Aug 2020
Resigns directorships at Hangzhou Weimi, Shanghai Xunmeng (Aug 6), and Pinduoduo Shanghai (Aug 19).
Mar 2021
Resigns as chairman; gives share voting rights to the board; announces food and life sciences research.
2021
Named China's top philanthropist by Hurun after pledging $1.85B; $100M gift to Zhejiang University.
Aug 2024
Becomes China's richest person while holding no operating role.
Whatever his motives were, and he has explained them only briefly, the result was that during the most dangerous period for China's tech founders, Huang was not the face of a dominant platform. He was a philanthropist with a science foundation and no executive title. Leadership passed to an engineer co-founder in Lei Chen and later to a co-CEO in Jiazhen Zhao. Control was placed in a partnership and a board.
The approach holds up as a matter of design, separate from the political context. A founder who builds institutions able to operate without him (a written theory, a partnership, a trusted successor) makes the company more resilient and makes himself less of a target. He keeps the financial upside and lets go of the operating control.
Tactic: Plan your exit as a series of steps (title, board seats, votes, public profile), each of which leaves the company stronger without you, and begin before conditions force it.
Principle 12
Arbitrage rules written for another era, and plan for the day they change
Temu's American prices relied on the de minimis exemption, a 1930s rule meant for souvenirs that Congress raised to $800 in 2015 without imagining a billion parcels a year. NPR estimated that Temu paid no import taxes in 2022 while Gap paid about $700 million. As NPR pointed out, this was not illegal. It was a rule written for one era being used in another, and it was very profitable.
The broader pattern is Huang's. Pinduoduo grew in the space left by WeChat blocking Taobao. It moved its domicile from Shanghai to Dublin. Each time, the company found a rule or gap the incumbents had not noticed or could not use, and expanded quickly inside it.
The risk is that these openings can be closed by legislation. Blumenauer's bill had bipartisan support, and NPR asked whether Temu's model "only ever worked because of" the loophole. That is the real test of this kind of strategy: whether the business still works when the rules change. Pinduoduo's domestic results suggest the underlying model is sound. Temu's results will show how much depended on the rule.
Tactic: Identify the outdated rule your margins depend on, estimate what your business looks like once it is closed, and build that version alongside the current one.
Free playbook
Get The Business Model Playbook
58 business models, one visual page each: how the money flows, the metrics that matter, and who runs it. Free when you join the Faster Than Normal email.
Free. No spam. Unsubscribe anytime.
Part IIIQuotes and Maxims
In their words
— Colin Huang, on the lesson of the 2006 Buffett lunchA person's mind is easily corrupted. When you are making a judgement about something, you need to understand the background and the facts. After understanding them, what you need is not cleverness, but the courage to use reason and common sense when facing the facts.
— Colin Huang, "Reversing Capitalism," final WeChat blog post, 2017Warren Buffett is an admirable capitalist, a pure capitalist through and through. His entire career can be described as a relentless, focused, and rational effort to move money around to enjoy the fruits of compounding interest.
— Colin Huang, letter to shareholders, 2019 annual reportWhen this tiny virus was dropped into our world, it acted just like a catalyst in a test tube, accelerating the formation of a whole new world... It is the time of reestablishment.
— Colin Huang, on his school yearsI was too goal-oriented and wasted too much time striving to be number one in class and a good student.
— Colin Huang, letter to shareholders on resigning as chairman, March 2021After resigning as chairman, I will focus on my lifelong interests and devote myself to research in the fields of food science and life sciences.
Maxims
- Rearrange the horses. You don't need to be stronger overall; you need to be stronger at the point where the contest is decided.
- Demand first, factories second. Collected, committed demand is more valuable than any forecast.
- The customer is the channel. Build a product that only works when users recruit other users, and you won't have to buy them.
- Start with the fruit. The product incumbents look down on is often the best entry point.
- Sell the feeling. The goal, as Huang said, is not to be cheap but to make people feel they got a good deal.
- Write it down early. A theory published before success makes everything after it look consistent.
- Common sense takes nerve. Cleverness is easy to come by; stating the obvious when your interests argue otherwise is harder.
- Punish beyond the law when trust is at stake. But decide in advance which side of the market pays.
- Leave while it's still your choice. A founder who builds institutions that work without him can step away and still keep his stake.
- Every loophole closes eventually. Build the version of the business that survives once it does.
