STARTUP GRAVEYARD

Two 20-year-old University of Waterloo dropouts founded Embark to build self-driving trucks. They got a truck to drive itself from Los Angeles to Jacksonville, Florida. They went public via SPAC at a $5.2 billion valuation. Then reality hit. Autonomous trucking was further from commercialization than anyone admitted. The cash burned faster than the trucks could drive. Embark shut down in 2023 and returned remaining cash to shareholders. Two kids, a great demo, and five billion dollars in evaporated value. The graveyard of autonomous vehicles claims another one.

IRL raised $200 million — including a $170 million Series C that valued it at $1.17 billion — by claiming to be a social app with 12 million monthly active users. An internal investigation later found that approximately 95% of those users were bots. Not exaggerated metrics. Not generous counting. Bots. The SEC charged the founder with fraud. A unicorn built on fake people.

Lordstown Motors bought a massive GM factory in Ohio, promised to build an electric pickup truck for commercial fleets, went public via SPAC at a $1.6 billion valuation, and then delivered almost nothing. The CEO was forced out after a short seller revealed that the company had faked pre-orders. The truck that did eventually ship had a tendency to catch fire during testing. Lordstown filed for bankruptcy in 2023 having burned through $675 million and produced fewer than 50 trucks.

Adam Neumann convinced SoftBank that a company that subleases office space was actually a technology company worth $47 billion, then watched the whole thing collapse so spectacularly it became a Hulu series AND an Apple TV series. WeWork burned through $22 billion in funding, attempted an IPO that valued it at $47 billion, got slashed to $8 billion before pulling the IPO entirely, and eventually filed for bankruptcy in 2023. It shed $4 billion of debt in Chapter 11 and walked back out in June 2024, smaller and finally solvent. The craziest part? The coworking model actually works — WeWork just spent money like it was printing it.

Zilingo was supposed to be the fashion technology platform that connected Southeast Asian garment factories directly to global retailers. Raised $310 million. Hit a $970 million valuation. Then the board discovered accounting irregularities, suspended the 30-year-old CEO, and the whole thing unraveled in weeks. A startup that pitched itself as the operating system for fashion turned out to not have its own books in order.

Argo AI raised $3.6 billion from Ford and Volkswagen, hired over 2,000 employees across two continents, and shut down in October 2022 without ever launching a commercial product. The founders came from Google's self-driving car project and Carnegie Mellon. They had the pedigree, the money, and the backing of two of the world's largest automakers. What they did not have was enough time before their sponsors lost patience.

Fast raised $124.5 million to build one-click checkout. Stripe already had this. Shopify already had this. Apple Pay already had this. Fast launched in 2019, burned through its entire war chest in three years, and shut down in April 2022 with reportedly only $600,000 in annual revenue. That's a $124.5 million spend for $600K in revenue. Someone check the math on the "fast" part.

Sam Bankman-Fried built the third-largest crypto exchange in the world in three years, raised $1.8 billion from Sequoia, SoftBank, and BlackRock, put his name on the Miami Heat arena, and then it all collapsed in 72 hours when a CoinDesk article revealed his hedge fund was gambling with customer deposits. $8 billion in customer money vanished. He's now serving 25 years in federal prison. The arena already has a new name.

Katerra raised $2 billion to fix construction by building everything in factories — walls, bathrooms, entire apartments — and snapping them together on site like Lego. SoftBank's Vision Fund poured in $1.3 billion alone. The company hired 8,000 people, opened factories on three continents, and went bankrupt in 2021 having completed almost nothing on time or on budget. Turns out construction is hard for a reason.

MoviePass offered unlimited movie theater tickets for $9.95 a month. The average movie ticket costs $11. You don't need an MBA to see the math problem. At its peak, MoviePass had 3 million subscribers, was losing $20 million a month, and the parent company's stock dropped 99.9%. It was the most generous deal in entertainment history — generous to everyone except the investors.

Jeffrey Katzenberg and Meg Whitman raised $1.75 billion to build a mobile-only streaming service for short-form video. They launched in April 2020, six months before shutting down entirely. That's roughly $9.7 million burned per day of operation. The content wasn't bad. The idea wasn't insane. They just built a product that solved a problem nobody had — who needs a premium app for 10-minute videos when TikTok and YouTube exist and are free?

ScaleFactor raised $104 million by telling investors that AI was doing the bookkeeping. The AI wasn't doing the bookkeeping. Humans in the Philippines were doing the bookkeeping. The humans were making errors. Customers were getting incorrect financial statements. The company shut down in 2020 after customers started suing. A $104 million lesson in the difference between "AI-powered" and "actually AI."

Wirecard was a German payments company that made it into the DAX 30 — Germany's top stock index — overtaking Deutsche Bank in market value. Then auditors discovered that $2.1 billion in cash on the balance sheet didn't exist. It had never existed. The COO fled to the Philippines. The CEO was arrested. Germany's financial regulator, which had spent years protecting Wirecard from short sellers, became a global embarrassment. Europe's biggest accounting fraud made Enron look almost honest.

Elizabeth Holmes dropped out of Stanford at 19, founded a blood-testing company, claimed it could run 200+ tests from a single finger prick, raised $945 million, and reached a $9 billion valuation. The technology never worked. Not "it had bugs" — it literally never produced reliable results. She is now serving 11 years in federal prison. The finger prick was real, though — investors felt it.

Beepi raised $150 million to sell used cars online. The founders spent $7 million on office space. The CEO leased a BMW with company money. They burned through the entire war chest in three years and shut down in 2017 without ever figuring out how to make the unit economics work. Carvana did essentially the same thing and is now worth $30 billion. The idea wasn't the problem.

Jawbone raised $930 million to make wearable fitness trackers. They burned through nearly all of it. Fitbit ate their lunch, their products kept breaking, and they spent more time in court than in R&D. The company that was once valued at $3.2 billion liquidated in 2017 without even the dignity of a bankruptcy filing.

Juicero raised $120 million to build a $700 WiFi-connected juicer that squeezed proprietary fruit packs. Then Bloomberg reporters discovered you could squeeze the packs with your bare hands and get the same juice. The company shut down four months later. It remains the single greatest symbol of Silicon Valley's inability to distinguish between innovation and overthinking a problem that didn't exist.

Vine invented the short-form video format that TikTok turned into a $200 billion empire. Twitter bought it for $30 million before it even launched, watched it become a cultural phenomenon with 200 million monthly users, and then killed it because they couldn't figure out how to make money from it. The most expensive fumble in social media history wasn't building Vine. It was shutting it down.

Zirtual offered dedicated virtual assistants for busy entrepreneurs at $399/month. Simple business, obvious demand. Then one Sunday night in August 2015, every employee and every customer received an email saying the company had shut down effective immediately. No warning. No two weeks notice. 400 employees found out they were unemployed via email at 11 PM. The founder blamed a "cash flow crisis" that apparently materialized overnight. The brand was bought out of the wreckage within a week.

Better Place raised $850 million to solve EV range anxiety with battery-swapping stations. Drive in with a dead battery, drive out with a full one in three minutes. The idea was ahead of its time by about a decade. The execution was behind it by about the same. They launched in Israel and Denmark, sold roughly 1,400 cars, and went bankrupt in 2013. Shai Agassi spent $850 million proving that the future of EVs was charging, not swapping. NIO in China later proved him sort of right — but at a fraction of the cost.

Solyndra was supposed to prove that American clean energy could compete with Chinese solar panels. The government gave them a $535 million loan guarantee. Private investors poured in another billion. Their cylindrical solar panel design was genuinely innovative. It was also genuinely more expensive than the flat panels China was mass-producing at a fraction of the cost. Solyndra filed for bankruptcy in 2011 and became a political weapon that set back U.S. clean energy policy by a decade.

Pets.com became the poster child of the dot-com bubble by doing something genuinely innovative: losing money on every single order, then spending $17 million on a Super Bowl ad to tell more people about it. They sold 50-pound bags of dog food online for less than it cost to ship them. The sock puppet mascot was more famous than most CEOs. The company lasted 268 days as a public company. Chewy later proved the model works — you just can't sell products for less than they cost.

AeroFarms raised $238 million to prove that growing lettuce indoors was the future of agriculture. It was not. The 70,000 square-foot Newark farm could produce 2 million pounds of greens per year using 95% less water — but it used so much electricity that each head of lettuce cost more to produce than one grown in a California field under free sunlight. The SPAC deal collapsed. The company filed for bankruptcy in 2023. AeroFarms is the most expensive proof that sometimes the old technology — dirt and sunshine — still works better than the new one.

AppHarvest built a 60-acre greenhouse in rural Kentucky, raised $632 million, went public at a $1 billion valuation via SPAC in 2021, and filed for bankruptcy in July 2023. The thesis was good — local food production, fewer pesticides, 90% less water. The execution was a struggle from day one. It joins Infarm and AeroFarms in the vertical farming graveyard, proving that growing plants indoors at scale is harder than it looks from a PowerPoint.

Bird reached a $2.5 billion valuation faster than any company in history. It also crashed faster than almost any company in history. Travis VanderZanden put electric scooters on sidewalks without permission, raised $780 million from the best VCs in the world, went public via SPAC, and then watched the stock go from $2.3 billion to literally zero. The scooters broke constantly, cities hated them, people threw them in rivers, and the unit economics never worked — each scooter cost $500 and lasted a few months on streets where people treated them like disposable toys. Bird is the cautionary tale for what happens when Silicon Valley's move-fast-and-break-things culture meets physical infrastructure that actually breaks.

BYJU'S was the most valuable edtech company in the world at $22 billion. Tiger Global, BlackRock, Silver Lake, and Tencent all invested. Byju Raveendran was India's poster child for edtech. Then it all collapsed. Financials were not filed for years. Aggressive door-to-door salespeople pressured poor families into buying expensive courses they could not afford. A dozen acquisitions were overpaid. A $1.2 billion loan defaulted. Investors wrote their stakes to zero. BYJU'S went from $22 billion to insolvency in less than two years. It is the most expensive failure in Indian startup history and a reminder that growth funded by aggression and debt is not growth — it is a bubble.

Jeff Bezos invested personally. SoftBank put in $400 million. Convoy raised $900 million total to prove that software could fix the $800 billion US trucking industry. Then freight volumes collapsed, margins went negative, and Convoy shut down in October 2023 — laying off all 500 employees and selling its technology to Flexport for a fraction of what investors paid. Nine hundred million dollars spent to learn that trucking is a commodity business where the commodity (a truck) does not care about your algorithm when there is nothing to carry. The graveyard of logistics startups grows deeper.

Ten billion dollars. That is roughly what GM poured into Cruise before pulling the plug in December 2024. After the San Francisco pedestrian dragging incident, the regulatory coverup, and the CEO resignation, GM admitted what the market already knew — Cruise was burning $2 billion a year with no clear path to profitability. The most expensive "oops" in automotive history.

Ecom Express raised $350M to build India's second-largest e-commerce logistics network, then filed for insolvency in April 2024. It lost the game the same way many logistics companies do: dependent on a handful of large customers who either built their own networks or squeezed rates until margins disappeared. Delhivery survived. Ecom Express did not.

Getir invented ultrafast grocery delivery in Turkey in 2015, years before anyone else. Then COVID turned quick commerce into a global phenomenon and Tiger Global handed Getir $1.8 billion to conquer the world. They expanded to 9 countries. They acquired Gorillas. They hit an $11.8 billion valuation. Then interest rates rose, VCs stopped funding losses, and it turned out that delivering a bag of chips in 10 minutes for $2 is not a viable business model in most of the world. Getir retreated to Turkey, laid off thousands, and became the poster child for the quick commerce bubble. The lesson: being first does not matter if the math never works.

Infarm raised $604 million to prove you could grow fresh herbs inside supermarkets using modular LED farms. The technology worked. The economics did not. Energy costs were higher than anticipated, capital dried up in 2022, and Infarm went insolvent in 2023. Three years after SoftBank put in $200 million, the company was gone. It joins a graveyard of vertical farming companies that all faced the same problem: lettuce does not care how much your investors paid for it.

Olive AI raised $852 million to prove that artificial intelligence could replace hospital administrators. It could not. The "AI employee" that was supposed to automate insurance verification and claims processing produced enough errors that human employees had to check its work — defeating the entire purpose. Hospitals churned. Revenue stalled. Olive hit a $4 billion valuation in 2021 and shut down two years later. Eight hundred and fifty-two million dollars spent to learn that healthcare administration is messier than any AI demo suggests. The graveyard of AI-will-fix-healthcare startups gains another resident.

Yahoo paid $1.1 billion for Tumblr in 2013. Six years later, Automattic — the WordPress people — bought it for roughly $3 million. A $1.097 billion value destruction event, executed through a series of completely predictable decisions. David Karp got out early with his money. The users stayed. The advertisers never came.

Vroom raised $468 million in its IPO, sold used cars online for a few years, and then shut down the car business in January 2024 when used car prices normalized and the unit economics never closed. It is now a fintech platform. The used car e-commerce trade is buried between Carvana's near-death experience and Vroom's full pivot — proof that the category is harder than it looks.