The Number That Broke the Spell
Forty-seven billion dollars. That was the figure — a private-market valuation assigned to a company that subleased desks — hovering over the entire proceeding like a hallucination that nobody could quite shake. In January 2019, SoftBank's Vision Fund poured another $2 billion into WeWork at that number, making it the most valuable startup in America, worth more than Ford Motor Company, worth double the
GDP of Iceland, worth more than the nation's largest office landlord, Boston Properties, despite the fact that WeWork owned almost no real estate at all. Eight months later, the company filed its S-1 with the Securities and Exchange Commission. Six weeks after that, the IPO was dead, the valuation had been slashed by 75%, the CEO had been defenestrated, and the company was weeks from running out of cash. By November 2023, WeWork would file for Chapter 11 bankruptcy with $18.65 billion in liabilities against $15.06 billion in assets. Its market capitalization: less than $50 million. The distance between $47 billion and $50 million is not merely a financial collapse. It is the complete evaporation of a consensual fiction — the moment when a story that thousands of sophisticated people had agreed to tell themselves was exposed as exactly that. WeWork is not, in the end, a story about office space. It is a story about what happens when the machinery of Silicon Valley venture capital — the narrative infrastructure, the valuation mechanics, the pattern-matching heuristics that reliably identify genuine disruption — is applied to a business that doesn't disrupt anything at all. It is a story about the difference between a technology company and a real estate company dressed in a hoodie.
By the Numbers
WeWork at the Apex and the Abyss
$47BPeak private valuation (January 2019)
$44MMarket cap at bankruptcy filing (November 2023)
~$17BTotal SoftBank investment
$1.8BRevenue in 2018 (with $1.6B in losses)
779Locations across 39 countries at peak
$18.65BTotal liabilities at Chapter 11 filing
12,500Employees at peak
$1.7BGolden parachute and stock sale for Adam Neumann
Green Desk, or the Archaeology of a Business Model
The founding mythology of WeWork is, like most founding mythologies, both true and misleading. In 2008, Adam Neumann — a 29-year-old Israeli entrepreneur who had already failed at a baby-clothes startup called Krawlers — was living in a building in Brooklyn owned by a landlord named Joshua Guttman. Neumann and his friend Miguel McKelvey, an architect from Oregon who had grown up in a five-mother commune and developed a deeply communitarian sensibility, noticed that the building had empty floors. They convinced Guttman to let them fill the space with a concept they called Green Desk — shared office space with an eco-friendly branding angle, marketed to freelancers and small businesses in a post-crash Brooklyn where the traditional employment contract was unraveling in real time.
Green Desk worked. Not spectacularly, but it worked: desks filled, revenue trickled in, the economics of buying long-term lease space and reselling it in smaller increments at higher per-square-foot rates to people who couldn't commit to ten-year leases were straightforward. Neumann and McKelvey sold their stake in Green Desk to Guttman in 2010 — reportedly for a few hundred thousand dollars — and used the proceeds to start something bigger. They called it WeWork.
McKelvey, six-foot-five and soft-spoken, was the temperamental inverse of his cofounder. Where Neumann sold visions, McKelvey designed spaces. He had studied architecture at the University of Oregon, worked at a small architecture firm in New York, and brought an obsessive eye for how physical environments shape human behavior — the right light, the right density, the right ratio of communal space to private space. He was the product designer in a company that would soon forget it was a product company at all.
Neumann was something else entirely. Born in Israel, raised partly on Kibbutz Nir Am near the Gaza border, he moved to the United States as a teenager with almost no money and very little English. He was dyslexic, charismatic in a way that people who met him describe with a kind of bewildered awe, and possessed of the unshakable conviction that he was destined to build something enormous. "He kind of looked like a rock star, and the whole room kind of bends to him," one journalist who interviewed him recalled. He walked around barefoot in public. He wanted to be president of the world. He wanted to live forever. He wanted to become humanity's first trillionaire. These were not jokes. Or rather, they were jokes in the way that a certain species of founder tells jokes — with just enough ironic distance to be deniable if pressed, but delivered with enough intensity that you understood the ambition was real.
The gap between the mundane nature of the business and the cosmic scale of the ambition is where the WeWork story lives.
The Arbitrage in the Middle
Strip away the mythology, the mission statements, the kombucha on tap, and WeWork's business model was an old idea executed at venture-capital velocity. The company signed long-term leases — typically ten to fifteen years — on office buildings in major cities. It spent significant capital renovating those spaces into open-plan, design-forward work environments with glass walls, communal kitchens, and a millennial-coded aesthetic that made traditional office parks look like Soviet institutions. It then rented those spaces out to tenants — freelancers, startups, and eventually larger enterprises — on short-term, flexible leases, sometimes as short as a month.
The unit economics, in theory, were elegant. WeWork could fit more desks per square foot than a traditional office tenant because it used open layouts and shared amenities. If a building's long-term lease cost $50 per square foot and WeWork could charge members $75 per square foot on short-term flex leases, the spread was the business. In a rising market with high occupancy, that spread was real.
The risk was equally obvious, and anyone who has ever taken an introductory finance course could identify it: a massive asset-liability mismatch. WeWork's obligations — the long-term leases — were fixed and long-dated. Its revenue — the short-term memberships — was variable and short-dated. In a boom, the model printed money. In a downturn, when tenants could walk away on thirty days' notice but WeWork was locked into decade-long lease obligations, the model would collapse. This was not a novel insight. It was the same structural fragility that has destroyed leveraged real estate businesses since the invention of the lease.
As Harvard Business School professors Vijay Govindarajan and Anup Srivastava noted in a 2019 analysis, the word "technology" appeared 110 times in WeWork's S-1 filing, but the company's primary Standard Industrial Classification code — 7380, "Miscellaneous Business Services" — told a different story. WeWork was purchasing long-term leases from landlords and renting them out as short-term leases to tenants. Full stop. The technology — an app for booking conference rooms, a member social network that nobody used, data analytics about space utilization — was a feature, not the product. It was not a platform. It was not a marketplace. It was not a network effect business. It was arbitrage.
But calling it arbitrage would have meant a real estate valuation multiple. And a real estate valuation multiple would have meant WeWork was worth, at most, a few billion dollars. The entire WeWork story — the $47 billion, the $12 billion in venture capital and debt, Adam Neumann's $60 million Gulfstream, the 111 cities across 29 countries — depended on maintaining the fiction that this was a technology company.
The Enchanter and the Kingdom
To understand how the fiction held for so long, you have to understand Adam Neumann's gift.
He was, by all accounts, one of the most effective fundraisers in the history of venture capital. Not because he had the best deck, or the most compelling metrics, or even a particularly original vision — the insight that freelancers and startups wanted cooler office space at flexible terms was obvious. He was effective because he possessed the prophetic charisma that Silicon Valley — for all its rationalist pretensions — has always been deeply susceptible to.
If you're twenty-two today and you're out of college, you can't go and work for corporate America in the old way, and you need a new solution.
Neumann understood instinctively that venture capitalists were not evaluating spreadsheets. They were evaluating narratives — and, more specifically, evaluating the person telling the narrative. He told investors that WeWork's total addressable market was $3 trillion, a number derived by counting every person who worked at a desk in any city where WeWork operated as a potential member. The math was absurd. But the number sounded like a technology platform's TAM, and that was the point.
The story worked. In the summer of 2012, Benchmark Capital led a $17 million Series A at a $97 million post-money valuation — an astonishing number for a company that was, at that point, operating a few shared office spaces in Manhattan. The investment was driven in part by Bruce Dunlevie, a Benchmark general partner who saw in the model a version of the marketplace dynamics that had made the firm famous. The implicit bet: WeWork could do for office space what Airbnb was doing for hospitality, or what Uber was doing for transportation — use technology and brand to create a platform that sat between supply and demand.
It was a reasonable bet in 2012, when WeWork was small. The problem was that the thesis never evolved even as the evidence accumulated that WeWork was not, in fact, building a platform. It was building a chain. Every new location required its own capital expenditure, its own lease negotiation, its own buildout. There were no network effects — a member in Manhattan derived no value from WeWork opening a location in São Paulo. There were no zero-marginal-cost dynamics — every new desk required physical space, furniture, and a barista. The company grew linearly, not exponentially. But it raised money exponentially.
The SoftBank Supernova
The inflection point — the moment when a fast-growing but ultimately conventional real estate business became a $47 billion unicorn — was Masayoshi Son.
Son, the founder and CEO of SoftBank Group, had raised a $100 billion Vision Fund in 2017, the largest venture fund ever assembled, backed primarily by Saudi Arabia's Public Investment Fund and Abu Dhabi's Mubadala. The fund's size created its own gravitational distortion: it needed to deploy capital at unprecedented scale, which meant writing checks of $1 billion or more into individual companies, which meant Son needed to find companies capable of absorbing that much capital, which meant he gravitated toward businesses with the most aggressive growth narratives and the most charismatic founders, regardless of underlying unit economics.
Neumann and Son met in late 2016. The meeting lasted approximately twelve minutes — Son later described being driven around WeWork's headquarters for that brief period before committing $4.4 billion. The speed of the decision became part of the legend. "In a fight," Son reportedly told Neumann, "who wins — the smart guy or the crazy guy? The crazy guy. You're crazy, but I think you're smart, too."
In a fight, who wins — the smart guy or the crazy guy? The crazy guy.
The SoftBank relationship transformed WeWork in three ways. First, it removed all capital discipline. Between 2017 and 2019, SoftBank and the Vision Fund pumped approximately $10.5 billion into WeWork, and the total would eventually exceed $17 billion. With unlimited capital, WeWork could sign leases in every major city on earth without worrying about occupancy rates or payback periods. Second, each SoftBank investment came at a higher valuation — $20 billion, $42 billion, $47 billion — creating a reflexive loop in which the size of the check validated the valuation, and the valuation justified the next check. Third, and most corrosively, the SoftBank relationship insulated Neumann from accountability. When your largest investor is writing billion-dollar checks after twelve-minute meetings, the incentive to impose financial discipline on yourself or your company evaporates entirely.
The money went everywhere. WeWork was, at various points, the largest tenant in Manhattan, London, and Washington, D.C. Neumann acquired a $60 million Gulfstream G650. He and his wife Rebekah spent $90 million on a collection of six homes. He employed a squadron of nannies, two personal assistants, and a personal chef. He smoked marijuana on the company jet. He drank tequila at company events. He told colleagues his descendants would be running WeWork in 300 years. He paid himself $5.9 million in consulting fees for the trademarking of the word "We," which the company later bought from him for use in its corporate rebrand to "The We Company." He took out a $500 million personal line of credit from UBS, JPMorgan, and Credit Suisse, secured by his WeWork stock. He leased buildings he partially owned back to WeWork. The self-dealing was not hidden — it was, in the parlance of Silicon Valley, "founder-friendly."
"It was Succession craziness," a colleague would later say.
SoftBank's progressive investment tranches in WeWork
2012Benchmark leads $17M Series A at $97M post-money valuation.
2014$355M Series D; valuation reaches ~$5 billion.
2017SoftBank commits $4.4 billion; valuation reaches $20 billion.
2018Revenue hits $1.8B — with $1.6B in losses. Valuation reaches $42 billion.
Jan 2019SoftBank invests another $2B; valuation hits $47 billion.
Aug 2019S-1 filed. IPO valuation cut to $20B, then $15B, then $10-12B. IPO pulled.
Sep 2019Neumann ousted. Receives ~$1.7 billion exit package.
The S-1 that WeWork filed on August 14, 2019, was a document of remarkable ambition and inadvertent self-revelation. Its second line read: "Our mission is to elevate the world's consciousness." The word "energy" appeared thirteen times. The word "technology" appeared 110 times. Nowhere in the filing could one find a plausible path to profitability.
The numbers were stark. For the first half of 2019, WeWork reported revenue of approximately $1.5 billion and a net loss of nearly $690 million. For all of 2018, revenue was $1.8 billion against losses of $1.6 billion. The company had accumulated billions in long-term lease obligations and was burning cash at a rate that, without IPO proceeds or further SoftBank intervention, would have left it insolvent before Thanksgiving 2019.
But the most revealing feature of the filing was the metric WeWork had invented to describe its economics: "community-adjusted EBITDA." This was EBITDA — itself an already-generous measure that strips out interest, taxes, depreciation, and amortization — further adjusted to exclude building- and community-level operating expenses, marketing costs, general and administrative expenses, and development and design costs. By this measure, WeWork looked healthy. By any other measure known to accounting, it was hemorrhaging.
Community-adjusted EBITDA became an instant punchline. But the joke concealed a more serious point. The metric was not merely aggressive accounting — it was a philosophical declaration. It said: The costs of running our business are not really the costs of running our business. It was a company telling itself a story about what it was so persuasively that it had confused the story for the balance sheet.
WeWork's valuation and size is much more based on our energy and spirituality than it is on a multiple of revenue.
The public market did not buy it. Within days of the S-1's release, institutional investors had universally rejected the offering. The proposed valuation was slashed — first from $47 billion to $20 billion, then to $15 billion, then to $10 to $12 billion. Even at the lower range, there were no takers. The IPO was pulled on September 16, 2019.
What had changed? Nothing about the business model was new information. The losses, the lease structure, the asset-liability mismatch, Neumann's self-dealing — all of it had been reported, in some form, as early as 2015, when BuzzFeed published internal WeWork investor documents. The more skeptical financial press had long had WeWork's number. The Wall Street Journal's Eliot Brown had been cataloging Neumann's excesses for years.
What the S-1 did was force a translation. Private-market valuations exist in a social space — they are consensus fictions among a small group of insiders who share incentives to maintain the fiction. The number goes up because everyone in the room benefits from the number going up. An IPO forces that private consensus into public markets, where the audience is broader, the incentives are misaligned, and the scrutiny is forensic. The S-1 was not a revelation. It was a language change — from the idiom of venture capital, where losses are "investment in growth" and addressable markets are aspirational, to the idiom of public equity, where losses are losses and a company that subleases desks is a real estate company.
The Ousting and the Golden Parachute
Adam Neumann's removal from WeWork happened with the disorienting speed of a controlled demolition. On the afternoon of September 18, 2019 — two days after the IPO was pulled — the Wall Street Journal published a sweeping account of his management. Heavy drinking. Marijuana use on the company jet. The proclamation that he wanted to be president of the world. The $60 million Gulfstream. The $90 million in real estate. Neumann was dyslexic and had aides brief him on the article's contents.
He was, characteristically, unconcerned. He controlled 65% of WeWork's voting stock through a dual-class share structure that gave him ten votes per share. He could fire the board if the board moved against him. He had once declared in a company meeting that his descendants would be running WeWork in 300 years.
This time, the reality distortion field failed. Within days, SoftBank — finally awake to the magnitude of the problem it had created — intervened. On September 24, 2019, Neumann agreed to step down as CEO. But the terms of his departure were themselves a kind of masterpiece. He received a $185 million consulting fee, $500 million in stock, a $500 million loan repayment, and additional stock and options that brought the total package to approximately $1.7 billion. To put that in proportion: Neumann walked away with more money from his failure at WeWork than many founders make from building successful, profitable, enduring companies. The person most enriched by the WeWork story was the person most responsible for its destruction.
"It was foolish of me," Son told SoftBank investors in a May 2020 earnings call. "I was wrong."
The Pandemic, the SPAC, and the Slow Bleed
After Neumann's departure, WeWork brought in Sandeep Mathrani as CEO in February 2020. Mathrani was the anti-Neumann — a veteran real estate executive who had led the turnaround of mall owner General Growth Properties, a man whose expertise was not in elevating consciousness but in renegotiating leases and cutting costs. He was exactly what WeWork needed.
He had approximately three weeks before the world shut down.
COVID-19 was, for WeWork, a near-extinction event layered on top of an already-existing near-extinction event. The company whose entire value proposition was being in an office with other people watched its members vanish overnight. Occupancy plummeted. Revenue cratered. The entire cultural and economic logic of coworking — the informal networking, the free-flowing craft beer, the ambient energy of bathing in one another's respiratory droplets, as one writer acidly noted — was suddenly not just unappealing but illegal.
Mathrani stabilized the company through severe cuts. He slashed headcount. He renegotiated or exited hundreds of leases. He killed the most egregious Neumann-era excesses — the private school, the surf-wave pool company acquisition, the apartment-living venture called WeLive. Since the fourth quarter of 2019, WeWork would eventually cut $2.3 billion in recurring costs. It was a workmanlike triage operation, unglamorous and necessary.
WeWork eventually went public — not through the traditional IPO that had collapsed so spectacularly, but through a merger with a special purpose acquisition company (SPAC) called BowX Acquisition Corp, led by Vivek Ranadivé. The deal closed in October 2021, valuing WeWork at approximately $9 billion. The stock peaked at $13 per share.
It did not hold. By the second quarter of 2023, WeWork reported revenue of $844 million and a net loss of $397 million. The company issued a "going concern" warning — the accounting profession's most dreaded phrase — admitting that "losses and negative cash flows" had left management considering "all strategic alternatives," including bankruptcy. The stock fell to $0.13 per share. From its post-SPAC peak of $13 to $0.13 — a 99% decline — roughly $9 billion of value evaporated in less than two years.
On November 6, 2023, WeWork filed for Chapter 11 bankruptcy protection in federal court in New Jersey.
The Bankruptcy and the Resurrection
The filing listed approximately $15 billion in assets against $18.65 billion in liabilities. The company had roughly $100 million in unpaid rent. It had more than 700 locations and approximately 730,000 members. Ninety-two percent of its lenders agreed to a restructuring plan.
The bankruptcy proceeding was, in a sense, what WeWork should have been doing all along: rigorously evaluating each location's economics and keeping only the ones that made money. The company renegotiated more than 190 leases and exited more than 170 unprofitable locations. Annual rent and tenancy expenses were reduced by more than $800 million. The company secured $400 million of additional equity capital.
Yardi Systems, a property management software firm based in Santa Barbara, bought a majority stake. In June 2024, WeWork emerged from Chapter 11 with a dramatically smaller footprint — approximately 45 million square feet across 600 locations in 37 countries — and a new CEO: John Santora, a 47-year veteran of Cushman & Wakefield who had served as the commercial real estate firm's COO.
Santora was the fourth permanent CEO in five years. Where Neumann had spoken of elevating consciousness and Mathrani had spoken of adjusted EBITDA, Santora spoke the language of commercial real estate operations — vacancy rates, tenant improvements, lease optionality. Since taking over, he has invested more than $140 million in upgrading spaces and technology, and by early 2025, WeWork was reportedly profitable and cash-flow neutral.
Why make that long-term commitment, especially today, when you're not sure of how many people are coming back? We'll get you in 30, 60, 90 days, and you have the ability to walk away at certain points.
The irony is rich. The core product — flexible office space on short-term leases — turned out to be something the market actually wanted. Amazon, whose nearly 350,000 corporate employees were mandated to return to the office in early 2025, signed a lease with WeWork for 259,000 square feet at 1440 Broadway in Manhattan and operates a total of more than a million square feet with WeWork in New York. JPMorgan, Lyft, Pfizer, and Anthropic all use coworking spaces. WeWork now works with 40 of the Fortune 100. U.S. coworking space totals 158.3 million square feet across nearly 8,800 locations, accounting for more than 2% of office space — lower than pre-pandemic levels, but up 51.7% from 115.6 million square feet in about 5,800 locations three years ago.
The product was never the problem. The capital structure was the problem. The governance was the problem. The valuation was the problem. The story was the problem.
Neumann After the Fall
Adam Neumann has, with the resilience characteristic of those insulated by extreme wealth, moved on. He emerged from the WeWork wreckage with approximately $1.7 billion — the consulting fees, the stock sales, the loan repayments — and relocated to Israel, then to Miami. In 2022, he launched Flow, a startup focused on residential apartment living, which received $350 million from Andreessen Horowitz at a reported $1 billion valuation before the company had any tenants or operating history. The investment was widely viewed as one of the most audacious bets in recent venture history — a16z's Marc Andressen backing the man whose name had become a synonym for startup excess, at a billion-dollar valuation, on the thesis that Neumann's first attempt at reimagining physical space was a failure of execution, not of vision.
In October 2023, speaking at Saudi Arabia's Future Investment Initiative conference — the same event where SoftBank's billions had once flowed — Neumann addressed WeWork's bankruptcy with measured disappointment. "It's been challenging," he said, "to watch WeWork fail to take advantage of a product that is more relevant today than ever before." The self-awareness was, at best, partial.
He reportedly attempted to buy WeWork out of bankruptcy. The effort failed.
Miguel McKelvey, the quieter cofounder, had left WeWork before the bankruptcy. In 2025, he spoke publicly about "wallowing" in guilt over what had happened to the company. The two cofounders — the prophet and the architect — had built something that millions of people genuinely used and valued. They had also built a corporate governance structure so dysfunctional, a capital strategy so undisciplined, and a culture so untethered from financial reality that the product's actual merits became irrelevant.
Lessons in the Wreckage
The WeWork story has been told many times — in Reeves Wiedeman's
Billion Dollar Loser, in Maureen Farrell and Eliot Brown's
The Cult of We, in the Hulu documentary, in the Apple TV+ dramatization
WeCrashed. Each telling emphasizes a different villain: Neumann's narcissism, SoftBank's recklessness, the venture capital industry's suspension of disbelief, the Wall Street banks that fought to win the IPO mandate despite knowing the business couldn't sustain scrutiny.
All of these are true. None of them are sufficient.
The deeper lesson is structural. WeWork exposed the gap between two valuation regimes that had been allowed to diverge for more than a decade. Private markets, fueled by a $100 billion SoftBank Vision Fund that needed to deploy capital at scale, operated under a set of narrative rules — TAM as aspiration, losses as investment, charisma as a financial metric — that bore increasingly little relationship to the rules that govern public equity markets. The WeWork S-1 was the moment of translation, when the private dialect was forced into public syntax. The result was gibberish.
The company's trajectory after bankruptcy suggests something else, too. Stripped of the mythology, the self-dealing, the SoftBank billions, and the pressure to grow at all costs, the underlying business — selling flexible office space to companies that don't want to commit to ten-year leases — appears to be viable. Perhaps even attractive, in a post-pandemic world where office vacancy rates hit record highs (85.5 million square feet came up for renewal or vacancy in 2025 alone) and companies wrestling with return-to-office mandates and AI-driven workforce uncertainty need exactly the kind of flexibility WeWork always promised.
The product worked. The company didn't. The distinction matters.
The Building on Eighteenth Street
WeWork's former headquarters at 115 West 18th Street in Manhattan — the address listed on both its original S-1 and its Chapter 11 bankruptcy filing — sits in Chelsea, a neighborhood that has itself been transformed by the kind of capital-fueled gentrification that WeWork both benefited from and exemplified. The building is unremarkable from the outside. Inside, the open-plan floors that Neumann once roamed barefoot, trailing clouds of ambition and cannabis smoke, have been repurposed, re-leased, reorganized.
By early 2026, WeWork under Santora is profitable and cash-flow neutral, serving Fortune 100 companies from 600 locations in 37 countries. The coworking industry is growing. The business model is working.
The $47 billion is not coming back.
