Part IThe Story
There is really only one photograph of him in circulation. It was taken on the morning of March 11, 2021, outside the New York Stock Exchange. It shows a slim man of forty-two posed before the columned facade on the day his company sold 130 million shares at $35 apiece. A few hours later those shares opened at $63.50. In the version the Hankyoreh ran, the credit line reads courtesy of the NYSE; AP/Yonhap. Through the winter of 2025 and into 2026, Korean newspapers printed that picture above nearly every story about the man: the data leak, the hearing, the complaint, the criminal referral. They kept using it because they had nothing better. The founder of the company that delivers a very large share of what South Koreans buy was known to them mainly as a figure in a frame from lower Manhattan, and the frame had been supplied by the exchange.
On December 14, 2025, the man in the photograph wrote to the National Assembly. A former employee of his company had used a stolen signing key to reach into the records of 33.7 million customer accounts. The intrusion ran from June to November. Korean commentators kept noting that the number was roughly two-thirds of the population. The Science, ICT, Broadcasting and Communications Committee wanted the founder to explain it in person. His reply came down to one sentence:
— Bom Kim, statement to the National Assembly, December 14, 2025As I currently reside and work abroad and have official business appointments as the CEO of a global company operating in over 170 countries, I regret to inform you that I am unable to attend the hearing in person.
The hearing went ahead two days later anyway. Two other senior executives had also sent their regrets, so the witness table held only Harold Rogers, the company's general counsel and chief administrative officer, who had become interim chief executive of the Korean unit days earlier. Rogers told the lawmakers, in effect, that under American law the breach was not as large a matter as they believed. The lawmaker Choi Hyung-du pointed out that Mark Zuckerberg and Jeff Bezos had sat in front of legislators. The stock was already down 17 percent. On the 15th, the Democratic Party members of the committee had announced that they would pursue complaints under the Act on Testimony and Appraisal before the National Assembly. They called the absences "nothing but an organized attempt by the company to evade responsibility."
He had done this before. In 2015, summoned over unfair business practices, he declined and cited "back injuries caused after playing basketball." Very few people in Seoul believed that excuse, then or later.
Now consider what his company promises. Order before midnight and the package is at your door overnight, carried by the company's own drivers. Order fresh food before midnight and it is there by seven in the morning. The company has claimed that 99.3 percent of Rocket Delivery orders arrive within a day, and in 2020 it said seventy percent of Koreans lived within ten minutes of one of its logistics centers. Coupang is, before anything else, a machine for showing up. In its Chinese transliteration the name reads 庫房, kùfáng: storehouse. The man who built the storehouse declined, twice in a decade, to appear himself.
For ethnic Koreans who hold foreign passports and are believed to profit from Korea while evading its obligations, Korean has a phrase: 검은 머리 외국인, "black-haired foreigner." By January 2026 people were using it about him in print.
By the Numbers
The Storehouse and Its Keeper
$34.5BCoupang net revenue, 2025, up 14% year over year
33.7MCustomer accounts exposed in the 2025 breach
~74%Bom Kim's voting power in Coupang Inc. via dual-class shares, per Korean press accounts
$4.6BRaised in the March 11, 2021 NYSE IPO, the largest foreign listing since Alibaba
₩1.7TCumulative losses between 2015 and 2017 alone
2,500 → 30,000Jobs created by Coupang in 2014 versus 2019
>90%Share of Coupang revenue earned in South Korea
Everything I Am Not
Before the storehouse he made magazines, and to understand him you have to start there.
Kim Beom-seok was born in Seoul on October 7, 1978. His father worked for Hyundai Engineering & Construction, which in those years was sending Korean engineers and managers around the world to pour concrete for other countries' ambitions. The family moved to the United States. American sources put Bom at seven when they left; at least one Korean biographical sketch says he was in middle school. I could not resolve the discrepancy, and perhaps it doesn't need resolving. Each country seems to remember him leaving at an age that suits its own version of him.
He went to Deerfield Academy in western Massachusetts, class of 1996, and wrestled and ran track on the varsity teams. Wrestling is a sport of leverage and weight classes, and nobody else is responsible for what happens on the mat. From Deerfield he went to Harvard College, where he studied government. He spent the summer of 1997 at Brill's Content, Steven Brill's short-lived magazine about the press, and the summer of 1998 as a reporter-researcher at The New Republic, working alongside his classmate Daniel Loss. Loss would later finish Harvard Law School and keep coming back to work with Kim part-time. With Loss, Kim started a news magazine for college students called Current, which Newsweek bought in 2001. He was twenty-two or twenty-three. He had built a product, sold it to an incumbent, and gotten out. He would repeat that sequence, with variations, for the rest of his career.
After graduation came a short stint at the Boston Consulting Group. It may have mattered more than its length suggests, because the man who left BCG was always going to evaluate businesses as systems of costs and flows, even when they looked like media. Then came the venture that taught him more than any other, which was a failure.
He called it 02138, after the Harvard Square ZIP code. The concept was a glossy, independent magazine about Harvard and its alumni, and Kim, twenty-seven years old in June 2006, kept describing it to anyone who would listen as "the Vanity Fair for Harvard." The New York Observer noted that the university already published two alumni titles, Harvard Magazine and a newly merged house organ called The Yard, and sneered gently: not to be confused with the Vanity Fair, it wrote, "the Vanity Fair for the University of Ottawa!" Kim was untroubled. "We don't consider ourselves an alumni magazine in the traditional sense," he said. When the barebones website went up on June 8, with the first actual issue still three months away, he called it "a soft launch."
The money came from David Bradley, and Bradley is worth a paragraph, because he is the first of several men in this story who saw something in Kim that resembled a version of themselves.
Bradley was born in Washington in 1953 and liked to call himself a "five-dollar baby," after the price of the ID bracelet at Walter Reed. From the age of thirteen he wanted to be the young Republican senator from Maryland. Instead he founded the Advisory Board Company and the Corporate Executive Board and made more than $300 million selling research to the institutions he had hoped to run. Then, at forty, on a thirteen-hour flight to Vietnam while his friend slept beside him, he concluded that he "didn't have a leadership gift." So he bought National Journal, and two years later The Atlantic. "If you can't be in politics, you can be near it in journalism," he said. "If you can't take the course, you can audit it."
That was a man who had settled for being near power. He put $4 million into a young man who, as it turned out, would not settle. Bradley told the Observer that it was the first time he had invested a substantial sum in someone else's company, and he described what he was buying with a candor that reads, twenty years later, like a horoscope:
— David Bradley, on investing $4 million in 02138, The New York Observer, June 2006Among my concerns is how little I bring to the 02138 editorial sensibility. The magazine's market position is everything I am not: light, engaging, easy, original, sophisticated.
The more revealing material in the Observer piece is the list of rooms Kim had already gotten into. He had coffee with Kurt Andersen at Dean & DeLuca. ("I'm as self-obsessed as any Harvard alumni," Andersen said, agreeably.) He had breakfast with Walter Isaacson and James Fallows at the Hay-Adams, hosted by Bradley, on the morning of the White House Correspondents' Dinner. Steven Brill, his old summer boss, was an unofficial adviser. He had recruited Caroline Miller, a former editor of New York, as editorial director. Kim was explicit about what all of this was for. "These are people who are not only members of our community," he said, "but prominent members of the field. They've been very supportive."
Our community. Kim was a Korean-born scholarship-circuit striver from a construction family, and the Harvard network was not a place he had been born into. It was something he had learned to build, much as a person learns a language.
02138 folded in 2008 as the financial crisis took the advertising market apart. One account has him selling it before the worst came, which spared his investors the full loss. The magazine itself is a footnote. The skill he took from it was not. He had raised $4 million from a stranger on the strength of a sentence, and he had found out that the people at the top of American institutions would take a meeting with him if a third party they trusted arranged it. Ten years later he would apply that knowledge to Wall Street, and twenty years later to Washington.
A Very Short Window
He went back to Harvard, this time to the Business School, and he went, by the most candid account, to buy time. He wanted somewhere to sit while he found the next idea.
"I had a belief when I was in grad school," he told CNBC in 2019, "that I had a very short window to really make something that had an impact." He doesn't say what would close the window: age, capital markets, his own nerve. Reading the line back now, you notice how much urgency it carries. A man who expects the window to close does not wait for anyone, including professors and, later, regulators.
At the Business School he heard Clayton Christensen lecture. Christensen was the professor who gave the business world the phrase "disruptive innovation." He had spent his career explaining why incumbents fail to notice cheap, crude, upstart products until those products eat them. Accounts in the Korean press describe Kim as inspired by the lectures, and Christensen did more than inspire him. When Kim founded Coupang in 2010, Christensen's firm, Rose Park Advisors, invested. At the IPO eleven years later, Rose Park owned 5.1 percent of the company.
The idea Kim settled on was not original, and he never pretended otherwise. He wondered whether some business that was working in America could be carried to Korea, where he understood the culture and the language better than most American founders and understood America better than most Korean ones. In early 2010 the obvious candidate was Groupon. Groupon was then the most celebrated startup in the country, and it was single-handedly reviving the post-crisis venture market. Its daily-deals model was simple to copy, so investors were funding copies everywhere.
He dropped out after roughly six months, in his second semester, against the advice of his professors. "I knew I wanted to start something in commerce," he later told Forbes, adding the words that mattered: in Korea.
He raised $2 million. One of the first checks came from Bill Ackman, the Harvard-educated activist investor at Pershing Square, whom the Korean financial press likes to call "Little Warren Buffett." Ackman held the position for more than a decade. After the IPO, he gave away Coupang shares worth more than a trillion won to charity. Ackman introduced Kim to David Frankel of Founder Collective, another Harvard man. After the 2021 listing Frankel posted a public message to Ackman: "Bill, please keep introducing me to good grad students." It reads like a joke between friends, which it was. It also describes the mechanism exactly: introductions moving from one trusted person to the next.
Founder Collective's own retrospective on that investment is useful because its authors admit how close they came to passing. By May 2010, they wrote, they had seen more than twenty daily-deal pitches, "and even though Groupon was ascendant, as a group we felt that the space was generally a house of cards." They invested anyway, because of "his energy, focus, smarts, and unique insight into media and the highly specific Seoul market." In other words, they backed the man and not the idea. They thought the idea was weak, and they were mostly right.
The Korean market for Groupon clones was not empty when Kim arrived. One business history puts him at roughly the thirtieth entrepreneur to try to build a Groupon for Korea. Part of his reason for going had been a belief that entrepreneurship in Korea was thinner than in America and that competition would be lighter. That belief, the same history notes drily, "couldn't have been further from the truth."
The name was a pun: coupon plus pang, the Korean onomatopoeia for a small burst, a pop of surprise or fun. The company that would one day face hearings over stolen data, dead workers and a criminal referral was named after the sound of something going off.
The Greece of Asia
Coupang was established in the summer of 2010. Wikipedia gives July 1; a Korean startup publication, writing two years later, says August, with seven employees and monthly turnover of 190 million won. It was registered as a Delaware limited liability company so it could take American money. That detail looked like paperwork at the time. Fifteen years later it was a constitutional argument.
The first year was absurd. Coupang did roughly $10 million in sales and signed up more than three million customers. Part of that was the novelty of the format and steep discounts. Part of it was something Kim noticed before his competitors did: Korean internet users were on Facebook, and Korean advertisers weren't yet. He became one of the first serious Facebook advertisers in the country and saturated it. By one count, Coupang at one point was showing the average Korean internet user seventy-two ads a month. By mid-2011 the company was at a $300 million annualized run rate, taking a commission of around fifteen percent. Because customers paid for their deals upfront, it was briefly cash-flow positive, which is rare for anything in this story.
That April, Business Insider ran a Q&A headlined "HOW TO BEAT GROUPON: an Exclusive Q&A With the Man Who's Doing It." The copy had become the cautionary example for its original.
In late 2012, Kim spoke at a conference in Half Moon Bay, California, hosted by Altos Ventures. Altos had invested in Coupang in August 2010, when the company had about $100,000 in revenue. By the time of the talk it had twelve million subscribers, 732 employees, 25,000 merchants, and was on track for perhaps $600 million in revenue that year. Twenty-two months after launch it had ranked fifth among Korean e-commerce companies, and it was profitable. Kim used the talk to explain why Korea was a special case, and the explanation is the closest thing to a manifesto he has given in public.
The country, he said, put fifty million people into a habitable area about the size of Massachusetts. That was the state of Deerfield and of Harvard, and of the rooms where he had learned how America's elite introduces itself. Seventy-five percent of Koreans lived in cities, so delivery was "cheap and easy." Broadband penetration was ninety-five percent. And there was a third point, the one people remember:
— Bom Kim, at an Altos Ventures conference in Half Moon Bay, California, 2012It's like the Greece of Asia. You can't invade it from the outside.
He had numbers to back it up. Korea, he said, was "pretty allergic to outside tech companies." Groupon had spent tens of millions of dollars in the country and was about a fifth of Coupang's size. Google had a few percent of search, and Naver held the rest. The lesson he drew was clear: the walls were real, and you had to come in through a gate. Coupang, he said, had done it "by taking the best practices of the East and West."
That sentence should be read twice. Its speaker was Korean by birth, American by passport and training, and fluent in both countries' codes of trust. He was telling a room of California investors that a foreign company could not conquer Korea, and he was telling them this in order to explain why they should give money to a company that, legally and financially, was itself foreign. The insider-outsider position was his advantage, and he understood that. What took longer, perhaps until December 2025, was finding out that the walls he described could also keep someone out who was trying to stand halfway in.
The Founder Collective partners noticed the density argument too, and wrote about it with the excitement of Americans watching a future arrive somewhere else first. The Seoul metropolitan area, they observed, was geographically smaller than New York City but held nearly three times as many people. Amazon had been built to cross a continent, so it started with small, easy-to-ship goods. Seoul's compactness let Coupang sell apparel, electronics and produce from the beginning. Its stated aim was to get diapers anywhere in Seoul in under two hours. Kim told them a story about the last World Cup: it had been nearly impossible to step onto a Seoul subway car without seeing riders glued to matches streaming in live HD on their phones. The Coupang app was two years old and had been downloaded by nearly a fifth of the country. It accounted for seventy percent of $2.2 billion in sales.
Ten years later Coupang would buy the Korean rights to La Liga, the Bundesliga, the NFL and the NBA, and would stream them to some of the same commuters. Nothing he noticed early seems to have gone unused.
Jaws That Did Not Drop
By 2013 the company was preparing an IPO on KOSDAQ, Korea's secondary exchange. It was three years old and had, by Kim's later account, crossed a billion dollars in sales. Bankers were involved. In startup terms this is the safe ending: list at home, cash out partially, let the public markets carry the company from here.
He pulled it at the last moment.
He has told the reason several times with little variation. "We had to ask ourselves," he told CNBC, "'Was the business we had built, were the services and experiences that we were providing … creating that kind of world where the customers we love, their jaws would drop?'" He paused, by the reporter's account, and then answered his own question. "The reality was no."
The phrase is worth looking at. Jaws would drop is consumer-marketing language, the vocabulary of slide decks and all-hands meetings, and Kim said it straight. The company's internal hashtag became #WowtheCustomer. Its stated mission, as later retold by investors, was to make customers ask, "How did I ever live without Coupang?" The language is not sophisticated. It is meant to be repeated and passed down a hierarchy until a warehouse worker in Incheon and a lawyer in Seattle can both recite it. The editor of 02138 knew how to write elegantly. As a company builder he chose to write slogans.
What replaced the daily-deals business, and then the eBay-style third-party marketplace it had briefly become, was something nobody in Korea was doing at scale. Coupang would buy inventory itself, hold it in its own warehouses, and deliver it with its own employees in its own trucks. Kim later compared the ambition to building a UPS-style logistics arm, partly because the Korean parcel system, he said, was "disjunctured." He was open about where the template came from:
— Bom Kim, to CNBC Make It, 2019The only models that we had seen were primarily like that of Amazon … And at that time, we really envied that.
Envy is an unusual thing for a founder to admit. Kim admits it plainly, which is characteristic of him. He has never claimed to have invented the model. He copied Groupon and said so. He copied Amazon and said so. What he claimed was the ability to execute the copy in a country that, as he had told the Half Moon Bay audience, foreigners could not invade.
Rocket Delivery launched in 2014. The Korea JoongAng Daily later ran a company photograph of Kim marking its first anniversary in 2015; it is one of the few images of him from that decade that isn't the NYSE photograph. The delivery workers were company employees, called Coupangmen. Founder Collective reported that some of them wrote thank-you notes on the packages by hand. In a market where couriers were subcontractors paid by the piece, Coupang put people on its own payroll and in its own uniforms at Korean doors. Two thousand five hundred jobs were created in 2014. In July 2020, because many of the people doing the work were women, the company renamed the role Coupangfriend, 쿠팡친구.
By 2016 Forbes was running a long Ryan Mac feature with the dek: Coupang's "blitz into same-day online retail is all but keeping Jeff Bezos out of South Korea — and making its prodigal merchant CEO, Bom Kim, a very rich man." Prodigal is a curious word for him. A prodigal son leaves home, spends his inheritance abroad, and comes back to a feast. This one came home and spent other people's inheritances, very large ones, building warehouses.
The Price of Overnight
The money came mostly from Masayoshi Son.
SoftBank put $1 billion into Coupang in 2015, at a valuation reported at $5 billion. In November 2018, the Vision Fund added $2 billion at a $9 billion valuation. That made Coupang Korea's most valuable startup and made Kim, at forty, the second-youngest billionaire in the country. He explained the bet to Forbes in terms Son would have approved of: "We don't want to grow just 5% or 10%, we want to grow exponentially."
Exponential growth cost a great deal. Between 2015 and 2017, Coupang lost 1.7 trillion won. In 2018, on $4.05 billion in revenue, it lost $1.1 billion; Bloomberg headlined it "Harvard Dropout's Startup Loses Billions in South Korea Battle." Korean business journalists had a phrase for the model: 팔면 팔수록 손해, the more you sell the more you lose. That was nearly true for a while. Labor costs went from 100 billion won in 2014 to 1.4 trillion won in 2019. The headcount the company had created went from 2,500 to 30,000. By 2024 it employed roughly 95,000 people.
At the Milken Institute's conference in September 2019 Kim sat on a panel with executives from three other Vision Fund companies: Grab, OYO and Auto1. The moderator's premise was that all four had built asset-heavy businesses filling infrastructure gaps in their regions, and all four had put off going public in order to do it. Coupang's run rate was then $11 billion. Kim was the only one of the four whose company would eventually turn a full-year profit, and that took until 2023, and even then with help from a one-time tax benefit. In October of the same year, after what Korean reporters later called persistent outreach, he brought Kevin Warsh onto the board. Warsh was a former Federal Reserve governor who had spent a decade working alongside Stanley Druckenmiller, and he too was a Harvard alumnus. Recruiting a central banker to an e-commerce board looks strange in 2019. It looks less strange in 2026, when Warsh was nominated to chair the Fed.
All of this meant the warehouses were full and the trucks were moving. By 2018 the company said 99.6 percent of its orders arrived within twenty-four hours. It operated around two hundred warehouses covering roughly twenty million square feet. It stored about 5.3 million distinct items using a system it called the Random Stow Algorithm, under which a product goes to whatever shelf is open rather than a fixed location, and roughly 1.7 million Rocket Delivery items left its centers every day. In 2019 it added Rocket Fresh: order by midnight, groceries at the door by seven. Subscribers could ask for their food in a reusable insulated bag, which a Coupangfriend would pick up on the next delivery.
Overnight delivery means people working overnight.
In 2017 a parliamentary audit criticized Coupang for using a blanket wage system that left delivery drivers unpaid for overtime. In 2018 a pilot of the early-morning delivery service, Korean reporters found, exposed brutal conditions and a hostile workplace. In April 2021, Kwon Young-gook, a labor lawyer and co-chair of the Committee for Coupang Workers' Human Rights and Health, said: "Seven Coupang employees and two subcontractors have died of cardiovascular disorders, such as heart attack, for the past year. [...] Out of the nine deaths, five are connected to overnight work as they passed away during or after night duty." That same month the company introduced Coupang Care, with paid breaks and health coaching. Company-friendly statistics noted that Coupang had recorded no work-accident deaths through the end of 2020. Heart attacks after night shifts do not usually get classified as work accidents, which is exactly what Kwon was arguing about. By November 2025, the couriers' Taekbae Union counted twenty-seven Coupang workers who had died on the job since 2020.
Merchants complained as well. The Item Winner system awarded a product listing to whichever seller offered the lowest price, sometimes as low as one won, and under Coupang's terms the images and content a seller uploaded could then serve the winning competitor. In 2024 the Korea Fair Trade Commission fined the company 140 billion won for manipulating search rankings in favor of its own products.
None of this slowed the business. Revenue was $6.27 billion in 2019, $11.97 billion in 2020, $18.4 billion in 2021. During the pandemic Koreans stopped going out and Coupang's trucks came to them instead. Revenue per active customer rose from $161 to $256 in a year. The things that matter to a Delaware corporation and the things that matter to a Korean parliament were starting to diverge.
Why Do You Say It That Way?
In a meeting four or five years ago, according to a former executive interviewed by the JoongAng Ilbo, someone said the words "reasonable price." Kim frowned. The executive knew what would come next, because it always did. He would open with 왜 그렇게 말씀하시죠?, "Why do you say it that way?" He did.
"Why 'reasonable price'? If it's the cheapest price, it isn't a reasonable price. It's the lowest price. It's right to call it the lowest price."
The paper's account goes on to describe a founder who could not let an imprecise word pass, apparently because he believed an illogical phrase would eventually produce an illogical decision and the work would fail. The headline promised more: "Why?" — when he takes off his glasses, it's the first signal: Coupang's Kim Bom-suk and his "three stages of anger." Former employees, in the reporter's telling, all pointed to this "problematic figure" and his leadership as the reason for the company's rise. At Coupang, the paper said, taking your time deliberating before launching a service was simply not tolerated.
This is the magazine editor again. A man who once spent two years fussing over a magazine's sensibility now ran a logistics company the way an editor runs a copy desk, line by line and word by word. Every noun in a meeting got checked. It works as a management style. It is also a theory of truth: what you call a thing determines what you then do about it. That theory comes back later.
The product line grew the way a careful editor builds a magazine, with each section supporting the others. Rocket Wow, the membership, launched at 2,900 won a month and bundled free Rocket Delivery, thirty-day free returns and early-morning grocery delivery. On December 30, 2021, the price for new subscribers rose to 4,990 won. In July 2020 Coupang bought the assets of HOOQ, a failed Singaporean streamer, and on Christmas Eve that year launched Coupang Play on top of them, free to Wow members. Two years later Play had a hit when Saturday Night Live Korea, which it had brought over from tvN, ran a sketch mocking the presidential candidates that went viral. By January 2024 Play had 8.05 million monthly users, more than any other domestic streaming service in Korea. The streaming service mostly exists to make cancelling the Wow membership feel like a loss. Coupang Pay was spun out as its own subsidiary in 2020 and by year's end ranked as the second-most-used payment system in the country. Coupang Eats, a latecomer to a market led by Baemin and Yogiyo, went from 23.3 percent brand recognition to 72.4 percent in a year and topped a satisfaction survey at 74 percent.
Each of these pushed a Korean household's daily routine a little further inside Coupang. Each was also run from somewhere other than Korea.
The Seattle office opened in 2018. By 2022 it had 350 people and had become the company's headquarters. The company also has offices in Mountain View, Los Angeles, Beijing, Shanghai and Taipei. More than ninety percent of its revenue comes from Korea. Coupang is on the Fortune 500, ranked 132nd in 2026 with $34.5 billion in revenue, and almost no American shopper has heard of it. A Korean newspaper later described the situation in a sentence: Coupang earns 40 trillion won in Korea, but it's an "American company."
The quotation marks around "American company" in that headline are doing the same work Kim did when he asked why someone had said "reasonable price."
Founder, CEO of LLC
In October 2018, Coupang's lawyers started worrying about what might happen to Bom.
We know this because of a wrongful-termination lawsuit. Coupang's former chief privacy officer later submitted internal documents and messages from 2016 to 2020 to a Korean court, and the Hankyoreh reviewed them. In one exchange on the encrypted app Signal, Sam O'Brien, Coupang's vice president of legal and chief of staff to the chief administrative officer, flagged a problem: "the second issue," he wrote, was "that some guy from the Labor Office will question Bom about some CDM issue." CDM was internal shorthand for Coupang delivery men. "[A]pparently, we can't get out of this one," he added.
O'Brien then proposed a fix: give Kim a "'Founder' role and CEO of LLC," and appoint a Korean national as chief executive of the local entity. One of the candidates he mentioned was a man from the law firm Kim & Chang he had played golf with. The former privacy officer identified that man as Kang Han-seung.
Kang's background is a short history of Korean establishment power. He had been a partner at Kim & Chang, the country's most powerful law firm, and before that had served as secretary for legal affairs to President Lee Myung-bak. He came to Coupang first as a legal adviser and in October 2020 became representative director for business management. He was the kind of Korean executive whom Korean institutions recognize, and the documents suggest he was hired partly to be the person regulators could hold responsible.
In April 2020, according to the same documents, Harold Rogers, then the chief administrative officer, asked the head of people operations to hide the executives who reported directly to Kim from view in Outlook and Workday, the company's internal systems. The org chart was being altered so that the founder's name would not appear at its top.
Then, ahead of the 2021 listing, Kim stepped down as chairman and registered director of Coupang Corp., the Korean operating company, saying he wanted "to focus on global management." He remained chief executive and board chairman of Coupang Inc., the Delaware parent that owns all of Coupang Corp. Through Class B shares worth twenty-nine votes each, a structure available to a Delaware company listing in New York and not to one listing in Seoul, he kept roughly seventy-four percent of the votes. A Korean-born company that had once tried to list on KOSDAQ now had a corporate structure that placed its founder's control in America and his legal exposure, as far as the documents could arrange it, nowhere in particular.
The structure did what it was designed to do for four years. Its weak point turned out to be one sentence that American securities law requires.
In its quarterly report to the U.S. Securities and Exchange Commission, filed November 4, 2025, Coupang Inc. described its segments and named the person who ran them: "The Chief Operating Decision Maker ('CODM') is our Chief Executive Officer." That was Bom Kim. Every quarter, as American law requires, he had certified the filings with his own signature: I, Bom Kim, certify that: I have reviewed this report. The arrangement built in Seoul to separate him from the Korean business was contradicted by the disclosures filed in Washington, which said he made its decisions.
He knew the importance of what you call things. In America, his company had called him the decision maker.
The Morning in New York
Return to the photograph, March 11, 2021, now with what had been arranged behind it in view.
The IPO priced at $35, above the $32 to $34 range. It raised $4.6 billion, the largest American listing that year and the largest by a foreign company since Alibaba in 2014. The shares opened at $63.50, which put the company's value at around $114 billion for a while, more than nearly anything else listed in Korea. SoftBank owned about a third. Greenoaks Capital held 16.6 percent, Maverick 6.4 percent, Rose Park 5.1 percent, BlackRock 2.1 percent. Kim owned about ten percent of the equity and, through the super-voting shares, the company. His fortune briefly reached $8.9 billion. That morning he went on Squawk Box.
Coupang promised about $90 million in stock grants to staff and frontline workers. In Seoul the story went over well, at first. Bloomberg described him as the poster child of a new kind of Korean wealth: "self-made riches from the technology world instead of the dynastic hoards of decades-old conglomerates." That framing mattered in Korea, where the chaebol families had run the economy for three generations and where the head of Samsung had recently been jailed in an influence-peddling scandal. Kim was supposed to be the alternative: no dynasty, no succession fight, a Harvard dropout with venture money.
Five months later, the same Bloomberg reporters published a piece headlined "He built an $8.9 billion fortune. Then the controversies began." There were the worker deaths. There was a fire in June 2021 at the Deokpyeong logistics center that killed a firefighter. There was the Item Winner investigation. And there was the question Koreans would keep asking in different forms for the next five years: who is responsible for this company?
Over that period Coupang had in practice become one of the conglomerates it was meant to replace. It now appears on the Korea Fair Trade Commission's list of large business groups, alongside Samsung, Hyundai, SK, Lotte and Shinsegae. The KFTC designates a natural-person "controlling entity," 동일인, for each such group, and it can extend obligations to that person's family. Coupang argued that because Kim's relatives played no role in management, the corporation itself should be named the controlling entity rather than him. In 2024, Kim submitted a written pledge to the commission to that effect.
He also expanded abroad. A trial operation in Japan launched on June 4, 2021, and was shut down less than two years later; Amazon and Rakuten were too entrenched. He had said himself, in 2012, that foreigners could not invade Greece. Taiwan went better: Rocket Delivery launched there in October 2022, and Coupang was the most-downloaded app in Taiwan in the second quarter of 2023. In January 2024 he bought Farfetch, the troubled London luxury marketplace, for about $500 million over the objections of some investors. In January 2025 he returned to Tokyo through Coupang Eats' "Rocket Now" food-delivery service.
The company reported its first annual profit in 2023, $1.36 billion including a one-time tax benefit, on $24.4 billion in revenue. In 2024 revenue reached about $30 billion and profit was $154 million. In 2025 revenue was $34.5 billion and profit was a little over $200 million. The Bloomberg headline had read loses billions. Those years were finished, and the business was working.
Exposure, Leak
The language problem started on November 30, 2025.
That day Coupang posted an apology on its homepage for a security incident. The text said customer information had been exposed, 노출, rather than leaked, 유출. In Korean, as in English, the difference is between a door left open and a house robbed. Two days later the apology came down from the front page. It took seventeen days for the company to change "exposure" to "leak." When it published a fuller statement on December 7, it attached a note saying the update was "a measure taken at the request of the Ministry of Science and ICT, the Personal Information Protection Commission and the National Police Agency."
Kim Kyung-wook, business and industry editor at the Hankyoreh, wrote a reporter's notebook about the episode. He took issue with the word request, 요청. The ministries had not requested anything, he pointed out. Under the relevant laws they had issued recommendations, 권고, which the recipient is legally obliged to make a good-faith effort to carry out. Then he made a joke about the company's own services that every Korean reader would get: Coupang used overnight delivery to hide inconvenient facts and overseas shipping to correct inaccurate ones.
The precision of a magazine editor, and of the man who asked "why do you say it that way," was now being turned against his company. Korean journalists were reading Coupang's statements the way Kim reads a colleague's sentences in a meeting, word by word.
The facts behind the words were these. A former employee, a Chinese national, police said, who had worked at Coupang for about two years, had kept an electronic signing key and used it from June to November to pull customer records: names, addresses, phone numbers, delivery histories. The company said the information of between 2,609 and 3,000 accounts had actually been saved and repeatedly accessed out of the 33.7 million touched. The government's joint investigation team later said it had found no secondary damage. None of this mattered much politically. Thirty-four million Koreans used Coupang every month, and nearly all of them understood that the company knew where they lived.
Park Dae-jun, CEO of the Korean unit, tried to take the blame himself. "I haven't met [Kim] in Korea this year," he told lawmakers on December 3. "That's because, as CEO, I'm responsible for our Korean business and am in charge of those operations." The day before, at another committee, he had said: "Since this happened at our Korean company, it's my responsibility." Park was a Korean national doing what Korean executives traditionally do, publicly absorbing the blame for the person above him. The next day the Hankyoreh published the SEC filing naming Kim as the Chief Operating Decision Maker. Park resigned on December 10. Harold Rogers, who five years earlier had asked that Kim's direct reports be hidden in Workday, became interim CEO.
Then came the letter about 170 countries. Then the hearing at which he did not appear.
The comparisons in the Korean press were unforgiving. The JoongAng Daily editorial board invoked Akio Toyoda, who in 2010, in the middle of a global recall, went in person before the U.S. Congress, apologized, and, in the board's telling, began Toyota's recovery from that moment. Other writers brought up Lee Jae-yong, the chairman of Samsung, who had appeared at parliamentary hearings and had been jailed in connection with the Park Geun-hye scandal. Korean readers didn't need the point spelled out. The dynastic heir had shown up and the self-made founder had not.
Kim did eventually apologize, on December 28, in a statement, while declining once more to appear. Coupang announced a compensation plan worth 1.69 trillion won. Police raided the Seoul headquarters. The tax authorities opened a special audit. Labor, financial and customs investigators each began their own inquiries. The government called this a "whole-of-government" response. Delivery workers and small sellers told The Straits Times that orders had dropped and that they, not Seattle, were absorbing the loss.
Kim has a Korean name and holds an American passport. Korean reporters noted that his Korean passport had been revoked when he naturalized, as the law requires. He goes by Bom in one country and Beom-seok in the other. Wikipedia says he lives in Seoul. The JoongAng Daily, in January 2026, said he lived in the United States and visited Korea "here and there." I could not confirm which is true. That may be the most accurate thing I can report.
The Long Arm of Delaware
On January 23, 2026, Vice President J.D. Vance met South Korea's prime minister, Kim Min-seok, in Washington. According to The Wall Street Journal, Coupang was one of the main subjects. Vance warned against measures that would disadvantage American technology companies, Coupang among them. A few days later President Trump announced tariffs on a range of Korean goods. Representative Jim Jordan, chairman of the House Judiciary Committee, cited Seoul's treatment of Coupang as one of the reasons. On February 24 the Judiciary Committee held a session on the matter and heard about seven hours of testimony from Harold Rogers. Rogers had not been treated as fully responsible by Korean legislators. The American Congress, by contrast, questioned him at length.
Greenoaks Capital and Altimeter Capital together held about $1.5 billion of Coupang stock. They petitioned the U.S. Trade Representative for a Section 301 investigation of South Korea, accused the Lee Jae-myung government of an "increasingly anti-U.S., pro-China stance," and said they would pursue international arbitration. They withdrew the petition in March after American officials signaled a broader trade inquiry was coming.
Korean industry observers were astonished by the speed of all this. The Seoul Economic Daily tried to explain it with a diagram of personal connections rather than lobbying. It ran from Christensen to Ackman to Frankel, through Warsh to Druckenmiller, and back to the Harvard network. The paper called it "Bom Kim's 16-year network." The point was not that Kim had hired lobbyists. It was that he belonged to the American elite and understood how it communicated. The young man who had arranged breakfast at the Hay-Adams on the morning of the Correspondents' Dinner had kept building that network for twenty years. David Bradley had settled for being near power. His former protégé's company had now been the subject of a vice president's meeting with a foreign prime minister.
Seoul reacted the way you would expect a country to react after being told it cannot be invaded from outside and then being invaded. On April 23 the Hankyoreh ran an editorial headlined "US has no business telling Korea whether it can arrest Coupang founder." On April 28, ninety-six National Assembly members from several parties sent a joint letter to the U.S. Embassy condemning what they called interference in Korea's judicial sovereignty. Korean wire services began using a phrase that would have been unimaginable when the company was making Groupon clones: the Coupang matter had become a national-security question in the alliance.
Meanwhile the Fair Trade Commission had been looking closely at Kim's family. Its investigation found that his younger brother, Kim Yoo-suk, had participated in management of the Korean business. According to Korean reports, he held a vice-president title and received compensation and stock. In April 2026 the commission designated Bom Kim personally as the controlling entity of the Coupang group, the status his 2024 pledge had been written to avoid. In late May the commission's chairman, Ju Biung-ghi, said at a briefing in Sejong: "Coupang wrote a pledge promising that members of the controlling family were not involved in management. Since then, we have found evidence showing that the pledge was violated." He said the commission had "no choice" but to investigate and seek sanctions. "When false statements are proven, the only options under current law are criminal referral and criminal sanctions." He also said that because criminal penalties had proved too difficult to obtain, he would ask for a new administrative fine for false submissions of up to 20 billion won.
That May, Korean reports said, the company faced a possible data-breach fine of up to 1.5 trillion won. In June a regulator fined it 500 million won for misleading advertising. That same month the new Fortune 500 list came out, with Coupang at number 132.
At 6:54 on the morning of July 18, 2026, a fire started at a Coupang logistics center in Seo-gu, Incheon. The Incheon Fire Department sent 148 firefighters and fifty-nine vehicles. It burned for about sixty-one hours before it was brought under control around eight in the evening on July 20. Officials said the cause, and any casualties or damage, would be determined later.
Across the country, every night, someone opens the app before midnight and orders eggs, milk, rice cakes, a bag of apples. Somewhere between midnight and dawn a Coupangfriend parks a truck, climbs the stairs of an apartment block, and leaves an insulated bag outside a door. It is reusable and made to be returned. On the next delivery the driver comes back for it.
Part IIThe Playbook
What follows are the operating principles that can be drawn from Bom Kim's career. Some are worth copying. Some are warnings. A few are both. Kim is an unusually useful case because he succeeded at almost everything an operator can control, including capital, logistics, product bundling, talent and corporate structure, and then ran into the one thing an operator cannot structure away: a country's expectation that the person in charge will show up in person. The principles are drawn from what he did, not from what was said about him.
The Principles
- 1.Arbitrage your own in-between status.
- 2.Copy openly, then abandon the copy before it hardens.
- 3.Walk away from the exit when the product doesn't astonish.
- 4.Own the infrastructure the market can't provide.
- 5.Treat density as a strategic asset.
- 6.Build a network in which each backer introduces the next.
- 7.Give your losses a story large enough to live in.
- 8.Bundle until cancellation feels like amputation.
- 9.Police language, because language is policy.
- 10.Corporate structure is strategy, and every shield has a seam.
- 11.The frontline workforce is both the moat and the exposure.
- 12.Presence cannot be delegated.
R
Three Coupangs and a Reckoning
The company changed its model twice before it found the one that worked, then spent a decade building the structure that would be tested in 2025.
2006
02138 launches its website on June 8, backed by $4 million from David Bradley. It folds in 2008.
2010
Kim leaves Harvard Business School after about six months, raises $2 million including from Bill Ackman, and founds Coupang in Seoul as a Delaware LLC.
2011
Daily-deals run rate reaches roughly $300 million. Coupang has moved toward an open marketplace.
2013
KOSDAQ IPO prepared and pulled at the last minute.
2014
Rocket Delivery launches with in-house drivers and warehouses. 2,500 jobs created.
2015
SoftBank invests $1 billion. Kim declines a National Assembly appearance, citing a basketball injury.
2018
SoftBank Vision Fund invests $2 billion at a $9 billion valuation. Seattle office opens. Internal discussion of separating Kim from Korean liability begins in October.
2019
Kevin Warsh joins the board. Rocket Fresh promises groceries by 7 a.m.
2020
Coupang Pay spun out. Coupang Play launches December 24. Coupangmen renamed Coupangfriends.
2021
NYSE IPO on March 11 raises $4.6 billion. Kim resigns Korean registered posts. Japan trial begins.
2023
First annual net profit of $1.36 billion, aided by a tax benefit.
2024
Farfetch acquired. Kim submits family non-involvement pledge to the KFTC.
2025
Breach of 33.7 million accounts disclosed. Kim declines to appear at the hearing.
2026
Vance raises Coupang with Korea's prime minister. KFTC names Kim the controlling entity and signals a criminal referral.
Principle 1
Arbitrage your own in-between status
Kim's founding advantage was not technology or capital, at least not at first. It was his position between two countries. He could explain Korea to Californians in terms they would fund. "It's like the Greece of Asia. You can't invade it from the outside." He could build in Korea with an understanding of its culture that no Groupon executive had. Founder Collective said it outright: they doubted the category and backed him because of his "unique insight into media and the highly specific Seoul market." Being both foreign and native, to different audiences, was what got him funded.
The lesson is that a mixed identity can work as a moat when the two sides it connects don't trust each other. American capital didn't trust Korean founders it couldn't evaluate. Korean consumers didn't trust American platforms they hadn't invited in. Kim was the one person both sides could deal with.
The same arbitrage eventually had a cost. A position between two worlds is valuable as long as each side thinks of you as partly theirs. When one side decides you are entirely the other's, the "black-haired foreigner," the advantage reverses. Founders who build on dual status should expect that at some point one of the two sides will ask them to choose.
Tactic: Find a market where you are trusted by the people who have capital and also by the people who have customers, and build a company that sits on the line between them.
Principle 2
Copy openly, then abandon the copy before it hardens
Coupang went through three business models in roughly four years: a Groupon clone, then an eBay-style marketplace, then an Amazon-style retailer with its own logistics. Kim never claimed originality. "The only models that we had seen were primarily like that of Amazon … And at that time, we really envied that." What sets him apart is that he was willing to drop a copy that was working. The daily-deals business was cash-flow positive and growing. He moved away from it because he saw before the market did that the model, as Founder Collective had said, was "a house of cards."
Most founders who copy a proven model treat its success as permanent. Kim treated each borrowed model as temporary scaffolding that would let him build something else. A copy gets you started and buys you time, but it doesn't defend you, because everything that made it easy for you to copy makes it easy for the next founder too. He had been roughly the thirtieth Groupon clone in Korea, so he knew how crowded a copyable market gets.
Tactic: When you launch on a borrowed model, decide in advance what signal will tell you the copy has stopped compounding, and be ready to drop it when that signal appears.
Principle 3
Walk away from the exit when the product doesn't astonish
The 2013 KOSDAQ withdrawal is the most counterintuitive decision of Kim's career and possibly the most important. The company had reportedly crossed $1 billion in sales and was three years old. A domestic listing would have rewarded early investors and reduced the founder's risk. He asked whether customers' "jaws would drop," answered "no," and started over.
The test he used was not "are we growing?" or "will the market pay?" Both answers were yes. The test was whether the product delivered an experience so much better than the alternatives that customers would be changed by it. That is a higher standard than product-market fit, and it was useful because it was blunt. It let him reject a good result in favor of a possibly much better one.
The cost was nearly a decade of losses, about $1.1 billion in 2018 alone, and a dependence on SoftBank. The payoff came at the NYSE eight years later, and that IPO also gave him the dual-class control a KOSDAQ listing would not have allowed. The decision to walk away bought both scale and control.
Tactic: Before any liquidity event, ask whether customers would be shocked to lose your product, and if the honest answer is no, treat the IPO as premature rather than as the finish line.
Principle 4
Own the infrastructure the market can't provide
Kim decided that Korea's parcel system was too fragmented to deliver the experience he wanted, so he built his own: five distribution centers by 2014, about two hundred warehouses and twenty million square feet later, salaried drivers, and a Random Stow inventory system that handled 5.3 million distinct items. Founder Collective noted that Korean logistics providers "cannot cost effectively match what we take for granted from FedEx." The usual e-commerce approach is to stay asset-light and rent logistics. Kim did the opposite, deliberately and expensively.
The reason it worked is that the infrastructure became the product. Once Coupang could promise overnight delivery on millions of items, and groceries by 7 a.m., its competitors were not mainly competing on price or selection. They were competing with a physical network they would need years and trillions of won to replicate. At the Milken conference in 2019 the moderator grouped Coupang with three other Vision Fund companies that were filling regional infrastructure gaps. Coupang is the one whose infrastructure produced lasting profit.
The warning is that owning infrastructure also means owning everything that happens inside it: the night shifts, heart attacks, a warehouse fire that burned for sixty-one hours. Owning the network means you are accountable for it.
Tactic: If an outside provider's limitations define your customer experience, calculate what it would cost to own that layer outright, because that cost is also the size of the barrier your competitors would face.
Principle 5
Treat density as a strategic asset
In 2012 Kim told Californians that fifty million Koreans lived in a habitable area about the size of Massachusetts, which made delivery "cheap and easy." Founder Collective added that the Seoul metro area was smaller than New York City and held nearly three times as many people. Because of that density, Coupang could sell bulky and perishable goods such as diapers, produce and apparel from the start, whereas Amazon spent years limited to small, easy-to-ship items. By 2020 seventy percent of Koreans lived within ten minutes of a Coupang logistics center.
Density changes delivery costs, product selection and even what it's reasonable to promise customers. The insight is not simply that dense markets are good. It is that density determines which business models are possible at all. Overnight grocery delivery that loses money in suburban America can be a profitable habit in Seoul.
The failed Japan trial shows the limit. Japan is dense too, but it already had Amazon and Rakuten. Density helps whoever builds the network first, and Coupang was not first in Japan.
Tactic: Map your target market's population per square mile against your unit economics, and be first to build in the places where density makes an otherwise unprofitable promise profitable.
Principle 6
Build a network in which each backer introduces the next
The Seoul Economic Daily traced a sixteen-year chain. Christensen's lectures led to Rose Park's check, which came around the same time as Ackman's, and Ackman introduced Kim to Frankel. Warsh was recruited "after persistent outreach," and the chain eventually reached a vice president's meeting with a foreign prime minister. Earlier still, at 02138, Kim had used David Bradley's backing to get coffee with Kurt Andersen and breakfast at the Hay-Adams with Isaacson and Fallows. "Bill, please keep introducing me to good grad students," Frankel wrote to Ackman. That tweet describes the whole method.
Kim didn't collect investors so much as build a self-reinforcing reputation system. Each backer's endorsement lowered the risk for the next one. Each prominent board member raised the cost to a government of acting against the company. By 2026 the network was functioning as a geopolitical asset that Korean observers concluded no conventional lobbying effort could have built.
But a network that protects you also makes you more conspicuous. When Washington intervened on Coupang's behalf, Korean resentment intensified: ninety-six lawmakers signed a letter condemning interference in Korea's judicial sovereignty. Using his network to override a sovereign government turned a corporate dispute into a national one.
Tactic: Pick early backers partly for the quality of the introductions they will make, and treat each new endorsement as a step toward the next one.
Principle 7
Give your losses a story large enough to live in
Coupang lost 1.7 trillion won from 2015 to 2017 and $1.5 billion in 2021 alone. It did not report a full-year profit until 2023. Throughout, Kim kept capital coming by offering a story big enough to make the losses look like spending on the future: "We don't want to grow just 5% or 10%, we want to grow exponentially." SoftBank put in $3 billion. The public markets valued the company above $100 billion on its first day of trading.
This only works if the losses can be shown to be buying something durable, such as warehouses, trucks, a workforce and a membership base, rather than paying for customers who will leave. Revenue per active customer rising from $161 to $256 in a year was the evidence that made the story believable. Investors will tolerate losses when they can see what the losses are building.
Tactic: Tie every major loss to a specific durable asset you can point to, so that investors and employees understand the spending as investment rather than leakage.
Principle 8
Bundle until cancellation feels like amputation
Rocket Wow started at 2,900 won a month and included free delivery, free returns, morning groceries and, after December 2020, a streaming service built from the remains of a failed Singaporean company. By 2024 Coupang Play had 8.05 million monthly users and was buying rights to La Liga, the Bundesliga, the NBA and the NFL. Coupang Pay ranked second in the country, and Coupang Eats topped customer satisfaction surveys. On December 30, 2021, the subscription price for new members rose seventy-two percent, to 4,990 won.
The point is that no single component has to earn its keep alone. Coupang Play can lose money on sports rights if it keeps Wow members from leaving. Each added service raises the cost of cancelling. That is the defensive logic of Amazon Prime applied to a country with very high mobile adoption.
The breach showed both sides of this. A JoongAng Daily headline asked why users couldn't just leave if Coupang had botched the data leak. The bundle had made Coupang so much a part of daily life that even widespread anger didn't translate into mass cancellations. That stickiness is also exactly what makes regulators treat a company as a national utility.
Tactic: Add services to your core membership whenever doing so makes cancelling more costly to the customer than the service costs you to provide.
Principle 9
Police language, because language is policy
"Why 'reasonable price'? If it's the cheapest price, it isn't a reasonable price. It's the lowest price." Former Coupang executives told the JoongAng Ilbo that Kim corrected word choices in meetings because he believed imprecise language led to imprecise decisions. The habit goes back to the magazine editor who interned at The New Republic and Brill's Content. As a management practice it has real value: forcing exact nouns forces exact thinking, and a company that says "lowest price" commits to something it can measure.
The breach showed the other side of that practice. Coupang's first apology said "exposure" rather than "leak," and it took the company seventeen days to change the word. Its December 7 statement described a legally binding government recommendation as a "request." Korean journalists read those statements with the same rigor Kim applies to his staff, and they concluded the company was using vague words to avoid responsibility. A founder known for demanding precise language had a company whose crisis statements were imprecise.
Tactic: Hold your crisis communications to the same standard of exact wording you demand in internal meetings, because outsiders will judge your honesty partly by whether you name the problem accurately.
Principle 10
Corporate structure is strategy, and every shield has a seam
Coupang was registered in Delaware from the beginning, which gave it access to American capital and later to American dual-class shares. Kim's Class B stock gave him about seventy-four percent of the votes on roughly a tenth of the equity. Internal records show that from October 2018 onward his lawyers discussed giving him a "'Founder' role and CEO of LLC" and installing a Korean CEO, a well-connected Kim & Chang lawyer, to bear liability for the Korean business. Before the IPO, Kim gave up his Korean board seats.
§
What the Structure Bought, and What It Cost
Each element of Coupang's governance design and the way it held up under pressure.
| Structural choice | What it delivered | Where it gave way |
|---|---|---|
| Delaware incorporation (2010) | Scale Access to Ackman, Rose Park, Altos, SoftBank | Risk Fed the "American company earning 40 trillion won in Korea" narrative |
| Dual-class NYSE listing (2021) | Scale $4.6B raised; roughly 74% founder voting control | Risk Lawmakers named the voting control as the reason he should answer personally |
| Kim exits Korean board posts (2021) | Try Formal distance from Korean labor and regulatory liability | Risk The SEC 10-Q named him "Chief Operating Decision Maker" |
| Korean-national local CEO | Try A familiar, accountable face for Korean regulators | Risk Park Dae-jun took responsibility and resigned within days |
| Family non-involvement pledge to KFTC (2024) | Try Avoided naming Kim personally as controlling entity | Risk Brother's role uncovered; KFTC signals criminal referral |
| Seattle headquarters (2022) | Scan U.S. corporate identity; Washington took up its case | Risk Escalated a breach into a dispute between two governments |
The general lesson is that legal structures are designed by lawyers in one jurisdiction and tested by regulators, journalists and legislators in another. Coupang's shield had a seam where American disclosure rules met Korean accountability norms. The same SEC filing that satisfied Washington supplied the evidence Seoul used against him. Every structure assumes a particular kind of attack, and the attack that actually comes is often a different one.
Tactic: Have a lawyer from each jurisdiction you operate in read your governance documents as a hostile regulator would, and fix any statement that one country's filings make true and another country's politics make damaging.
Principle 11
The frontline workforce is both the moat and the exposure
Coupang's in-house drivers were the core of its competitive advantage: salaried employees in branded uniforms, some writing thank-you notes on packages. Jobs created went from 2,500 in 2014 to 30,000 in 2019, and total headcount reached about 95,000 by 2024. The $90 million in stock promised to frontline workers at the IPO acknowledged that the people climbing apartment stairs at 3 a.m. were the company's competitive edge.
They were also its biggest liability. There were the unpaid overtime findings of 2017, the 2018 early-morning pilot, the nine cardiovascular deaths counted in the year before April 2021, and twenty-seven deaths on the job by the union's count since 2020. Coupang Care, the paid-break program introduced in April 2021, came after the criticism. And when the breach caused orders to fall in 2026, drivers and sellers were the ones who lost income first.
An operator who brings the workforce in-house gains control over the customer experience and takes on direct responsibility for those workers' wellbeing. Those two things come together. Treating a frontline workforce purely as a source of efficiency eventually makes it the place where public anger concentrates.
Tactic: Build welfare safeguards for night-shift and frontline workers before you scale the service that depends on them, not after the first deaths make the news.
Principle 12
Presence cannot be delegated
This is the principle Kim's career teaches most clearly, by its absence. In 2015 he skipped the National Assembly, citing a basketball injury. In December 2025 he sent a letter citing 170 countries. Each time, someone else appeared in his place: Rogers, Park, Kang. Each time, the absence became the story. Korean editorials kept making the same comparisons: Akio Toyoda before the U.S. Congress in 2010, and Lee Jae-yong at hearings and then in prison.
⚖
The Arithmetic of Showing Up
How comparable leaders handled a summons, as Korean commentators framed it.
| Leader | Crisis | Response | Read by the public as |
|---|---|---|---|
| Akio Toyoda | 2010 global recall | Testified before U.S. Congress, apologized | Accountability |
| Lee Jae-yong | Park Geun-hye scandal | Attended parliamentary hearings | Custom observed |
| Park Dae-jun | 2025 breach | Testified, claimed responsibility, resigned | Proxy |
| Bom Kim | 2015 business practices; 2025 breach | Declined twice; apologized by statement | Evasion |
Avoiding an appearance has a cost, and that cost grows with how central the company is to people's lives. A niche startup's founder can stay away from a hearing without much consequence. The founder of a company used by thirty-four million people a month cannot, because people experience his absence as a decision about them. Kim's operational philosophy was built around reliable delivery, overnight and on time. That made his own absence look worse, not better.
Showing up is not only a public-relations step. In a country where custom says the person in charge takes the blame, it is part of what holds the operation together, and no substitute, whether a general counsel, a local CEO or a statement, can do it for him.
Tactic: Before your company becomes central to a country's daily life, decide which summons you will answer in person, and keep to that decision even when lawyers tell you the corporate structure allows you to stay away.
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
— Bom Kim, to CNBC Make It, 2019I had a belief when I was in grad school that I had a very short window to really make something that had an impact.
— Bom Kim, at an Altos Ventures conference, 2012We've done this by taking the best practices of the East and West.
— Bom Kim, to Forbes, 2018We don't want to grow just 5% or 10%, we want to grow exponentially.
— Bom Kim, to CNBC Make It, 2019The shape of Coupang, the business model of Coupang, what Coupang looks like today, went through a lot of change.
— Bom Kim, in a Coupang meeting, as recalled by a former executive to the JoongAng Ilbo (translated)Why "reasonable price"? If it's the cheapest price, it isn't a reasonable price. It's the lowest price.
Maxims
- Stand between two worlds on purpose. Being half-foreign and half-native is an advantage only while both sides still think you belong partly to them.
- Copy without shame, then leave without sentiment. A borrowed model is scaffolding, and you take scaffolding down.
- Refuse the exit that rewards mediocrity. If customers' jaws don't drop, the IPO can wait.
- Own the layer that defines the experience. If someone else's trucks set your promise, the promise belongs to them.
- Density decides what is possible. Map population per square mile before you write a business plan.
- Each backer should bring the next. The best investors open doors as well as write checks.
- Name things exactly. "Lowest price" is a commitment you can measure, and "exposure" instead of "leak" is an evasion people will notice.
- Every shield has a seam. A disclosure that satisfies one country can be used as evidence in another.
- The people at the door are the company. If you scale overnight delivery, you are responsible for the overnight shift.
- Show up. Everything else can be delegated, but the founder's presence in a crisis cannot.
