World story incidents

In 2008, we were completely broke and drowning in thousands of dollars of credit card debt. VCs weren't biting on the idea of people renting out air mattresses to strangers. To fund the company, we noticed the Obama-McCain election was dominating the news cycle. We designed and printed 1,000 custom cardboard cereal boxes: "Obama O's" and "Cap'n McCains."

We bought cheap generic cereal at the grocery store, hand-glued the boxes together in our apartment, and sold them as limited-edition political memorabilia for $40 a box to collectors. We made $30,000, which literally kept our servers running.

Later, when we interviewed with Paul Graham at Y Combinator, he looked at a box and said, "If you can convince people to pay $40 for a $4 box of cereal, maybe you can convince them to sleep in each other's airbeds."

In the early days of Y Combinator, most founders would pitch their startup to their peers and say, "I'll email you a link to our beta." They rarely got sign-ups because everyone was busy building their own companies. We realized that friction was killing our early growth.

So, we developed a tactic that became known in the community as the "Collison Installation." When someone agreed to try our payments API, we wouldn't send a link. We would say, "Great, give me your laptop." We would literally grab their computer from them, open their terminal, and integrate Stripe into their application's codebase right there on the spot.

I needed to get early adopters for our private beta, but our target audience — the highly technical users on Digg and Hacker News — were incredibly cynical about new file-syncing tools. Instead of a slick marketing pitch, I recorded a simple screencast demonstrating the software.

But I secretly loaded the demo with internet inside jokes: a "Chocolate Rain" mp3, a text file named after the "09 F9" hex key controversy, and a folder of "TPS reports." The community realized I was one of them and not a corporate marketer. The video hit the top of Digg instantly, and our beta waitlist exploded from 5,000 to 75,000 overnight with zero marketing budget.

We read an article in Popular Electronics about the new Altair 8800 microcomputer. We immediately called the manufacturer, MITS, and confidently told them we had developed a BASIC interpreter for their machine. They asked for an in-person demo in a few weeks. The reality? We hadn't written a single line of code, and we didn't even own an Altair to test anything on.

We spent the next eight weeks in a mad sprint, writing the software on a PDP-10 mainframe using an emulator we built to mimic the Altair's hardware. When Paul flew to New Mexico for the demo, he realized he hadn't written a bootstrap loader to actually get the program onto the machine. He wrote it on the airplane.

By some miracle, when he fed the paper tape into the actual Altair for the very first time, it worked perfectly.

When we first launched, the site was completely empty. We knew that nobody wants to participate in a ghost town; people only contribute to communities that already look active. To solve the "cold start" problem and convince real users to stick around, we created dozens of fake accounts.

We spent all day submitting links to our own site and talking to ourselves in the comments under different pseudonyms. We even built an administrative interface just so we could quickly switch between fake usernames without logging out.

Eventually, the real users started outnumbering our fake ones, and we were able to abandon our digital ghosts.

Before we ever built computers, Woz read an article in Esquire about "phone phreaks" — hackers who discovered that a specific 2600 Hz tone could trick AT&T's network into routing long-distance phone calls for free. Woz spent weeks designing a digital "Blue Box" that could flawlessly replicate these tones.

I immediately saw the business potential. We scrounged up parts, built the illegal boxes, and I started selling them door-to-door in college dorms for $150 a piece. We eventually stopped after getting held up at gunpoint by a buyer, but it was our first real business venture.

I've always said, if it hadn't been for the Blue Boxes, there wouldn't have been an Apple.

I was selling fax machines door-to-door in Florida and hated the way pantyhose looked under white slacks. I cut the feet off a control-top pair and discovered I'd accidentally invented a product. I had $5,000 in savings and zero connections in fashion.

I drove unannounced to hosiery mills across North Carolina and pitched my idea to anyone who would meet me. Every owner said no — until one of them, days later, called back because his daughters had told him to take me seriously.

To get into Neiman Marcus, I cold-called the buyer for fourteen days straight. When she finally agreed to meet me in Dallas, I dragged her into the women's bathroom and showed her my own backside — once in cream slacks alone, once in slacks over Spanx. She placed an order on the spot. I wrote the press release for myself, on my own letterhead, the same week.

In 1962 I flew to Japan with no business and no plan, just an idea: import Onitsuka Tigers running shoes to America. When I walked into Onitsuka's Kobe office, the executives asked which company I represented. I had nothing. I blurted out "Blue Ribbon Sports" — a name I made up on the spot.

They sent me $50 worth of samples on the strength of that bluff. I sold them out of the trunk of my green Plymouth Valiant at high school track meets across Oregon. My stockroom was my parents' laundry room.

When Onitsuka eventually tried to cut me out, I had to start a brand of my own. I called it Nike.

I was the head of marketing at a small Seattle coffee bean retailer called Starbucks. The company sold beans, not drinks. In 1983, the owners sent me to Milan for a housewares trade show.

Walking from my hotel to the convention center I passed a small espresso bar, then another, then twenty more. I watched the baristas greet regulars by name, the steam, the theatrics, the intimacy of it. I came back to Seattle and told the owners we should make Starbucks an Italian-style café. They refused.

I quit, raised $400,000, started my own café. Two years later, when the original owners decided to sell, I scraped together $3.8 million and bought the name and the bean stores I had once worked for.

I was at Le Web in Paris in December 2011. UberCab existed in San Francisco only, and only as one product: black cars. On stage, a French host asked when we were launching internationally. With no infrastructure and no team in Europe, I said, "Tonight."

Then I and a handful of people I half-knew spent the next eighteen hours signing up four black-car drivers in Paris by phone, in broken French, paying off random local fixers in cash. By the time the conference ended, you could pull out your phone in Paris and a Mercedes would arrive.

I was a senior vice president at a hedge fund in New York with a stable career. I'd just read a statistic that the web was growing 2,300% a year, and the math wouldn't leave me alone. I told my boss. He took me on a long walk in Central Park and asked me to think about it for forty-eight hours.

After forty-eight hours I quit. We packed up our apartment, my wife MacKenzie drove the car west, and I sat in the passenger seat with a laptop, building Amazon's revenue projections in Excel. We didn't even know what city we were going to until we landed in Seattle.

I started Amazon on a desk made of a door. I still keep one in my office.

In the early 1950s every banker in Burbank told me a theme park was an absurd idea. The numbers said it would cost $17 million, and I had a quarter of that.

I borrowed against my own life insurance. I sold the family's vacation home. I made my brother Roy mortgage everything we had at the studio. To raise the rest, I struck a deal with ABC: they would put up $500,000 and a loan guarantee, and in return I would give them a weekly TV show called Disneyland that, conveniently, was also a year-long advertisement for the park I was building.

The day Disneyland opened, the asphalt was so hot that women's heels sank into it. We had bigger problems and they were all my fault.

In the early years of Wal-Mart, I had to figure out where to put the next store. The chains used market research firms; I couldn't afford one. So I taught myself to fly small planes.

I'd take off at dawn from Bentonville and crisscross the Ozarks at low altitude looking for the small towns big retailers were ignoring. I'd circle a town to count the cars in its parking lot, gauge the population, and pick a corner of land to buy.

I had visited every Kmart, Target, and J.C. Penney on the continent before I ever opened a store. People said I was paranoid — I just thought my competitors weren't paranoid enough.

In Pinterest's first nine months we had under 10,000 users. Investors told me image-bookmarking was a feature, not a company. I personally emailed the first 5,000 users to thank them and listen to them. I held mall meetups in California — for five people, sometimes three.

I'd also walk into Apple Stores and quietly change every demo Mac's homepage to Pinterest. By the time anyone noticed, the cycle would already be working: people pinned, friends saw the pin notifications, and the platform grew person by person, week by week.

Bob, Ed and I were three engineers at Ampex who had read an obscure IBM research paper describing something called a relational database. The CIA wanted one. They wrote us a check before we had any code. The codename for their project was Oracle.

We built the software in our spare bedrooms, calling it Software Development Laboratories at first. When the prototype actually worked, we renamed the company after the project we'd snagged.

Years later we discovered we'd shipped Oracle Version 1, but we never sold a Version 1 to anyone. We labeled the first commercial release Version 2, because no enterprise buyer was going to bet on a 1.0.

After the war my piston ring factory was bombed flat. I had no factory, no clients, and a country that needed petrol but didn't have any.

I bought a job lot of 500 surplus military two-stroke radio generators — the kind army units had used to power field radios. I bolted them onto bicycles. They sounded like chainsaws and every neighbor wanted one to get to market.

When I ran out of generators, I started building my own engines. I was forty.

In 1955 I was running a small Tokyo electronics company that had just licensed transistor technology from Bell Labs. I built a portable transistor radio and flew to New York hoping for a distribution deal. Bulova, the watch company, offered to buy 100,000 units — the largest order I had ever seen, several times my company's annual revenue.

The catch: they wanted to put their name on it, not ours. The board demanded I take the deal. I refused. I told the Bulova buyer that in fifty years our brand would be just as famous as theirs. He laughed at me.

I went home with smaller orders and the brand intact. The brand was Sony.

I was a teenage rock climber in Burbank, California. The only pitons available to American climbers were soft European steel that bent on the first removal. I taught myself to use a coal-fired forge in my parents' backyard.

The steel I used came from a junkyard Ford Model A axle. I shaped each piton on a portable anvil that fit in the trunk of my car. I made them while traveling between climbing seasons in Yosemite, selling them for $1.50 each from the back of the car.

By the late 1960s my pitons were the standard for big-wall climbing. Chouinard Equipment became Patagonia.

In 1956 my photographer Gillis Lundgren was shooting a new IKEA table for a catalog. He couldn't fit it in his car to take it home. He pulled out a screwdriver and unscrewed the legs. The table fit.

So did the logic: customers could pick up boxed furniture from a warehouse, drive it home, and assemble it themselves — saving labor on our end and shipping on theirs. We were already a mail-order operation; we just hadn't realized that the box was the product.

The first flat-pack table was the Lövet, in 1956. Within a few years almost the entire catalog was flat-pack.

In 1995 we signed a $7 million contract with Sega to build a 3D graphics chip. Halfway through development we realized our architecture was wrong — Microsoft had standardized on something different, and our chip was going to be a dud. I flew to Tokyo and asked Sega to pay us anyway. They had every reason to say no. They didn't.

The money kept us alive long enough to build something else. Years later, when no one was buying our GPUs for anything but games, I made the same kind of bet on a programming language nobody asked for, called CUDA. It cost us a fortune, the stock cratered, and analysts told me to kill it.

Twenty years later, the entire AI industry runs on it.

My brother Kimbal and I started Zip2 — a sort of online Yellow Pages — out of a tiny office in Palo Alto. We rented one room and slept in it. We had one mattress and one couch. We showered at a local YMCA.

The internet was so new in 1995 that we had to physically take printed maps to potential newspaper customers and explain what "driving directions" were. When the rent went unpaid we covered it with a credit card. The first server I ever bought, I paid for with my own credit card and carried under my arm into the office.

I built AuctionWeb — what would become eBay — over Labor Day weekend, 1995, on my home Internet connection. I posted a broken laser pointer for sale as a joke, a test listing. Someone won the auction for $14.83.

I emailed the buyer and asked, are you sure you understand it's broken? He emailed back: "I'm a collector of broken laser pointers."

That was the moment I realized that for any object on Earth, there is a buyer somewhere who wants it more than the person currently holding it. I'd just built the marketplace that connected them.

I had a videotape of Apollo 13 that I was supposed to return to Blockbuster. I lost it. Six weeks later I dug it out from behind a shelf, returned it, and got handed a $40 late fee. I was angry the whole way home.

On my way to the gym, I realized: my gym lets me come twice or twenty times a month for the same fee. Nobody penalizes me for using it less, or more.

I went home and started Netflix that summer. The flat-rate, no-late-fees model — the part everyone remembers — came two years later, after I lost a gym membership and remembered Apollo 13.

At the 2003 Siebel User Group conference in San Francisco — the conference for the company we were trying to disrupt — I hired a small army of fake protesters. They marched outside Moscone Center holding signs that read "The end of software" and "Software is dead." I had paid actors. I had reporters. I had every Siebel customer in the world walking past the protest to get into the keynote.

By the time my own conference, Dreamforce, rolled around later that year, every customer in that crowd had been pre-conditioned to associate Siebel with "old software" and Salesforce with "the next thing." The cost of the protest was less than a quarter of one Super Bowl ad.

I started giving every new Zappos employee, after their full week of orientation training, a standing offer: take $2,000 in cash and quit, no questions asked. People thought I was crazy.

The math actually wasn't even close. We invested huge amounts in training and culture. If a new hire took the money and left, that just meant they didn't believe enough in what we were doing — the worst possible outcome would have been them staying.

Roughly two percent took the offer. The other 98 percent had publicly turned down free money to be there, and that was a kind of commitment money cannot buy.

I built two companies that started life as massively multiplayer online games and ended life as something else entirely.

The first was Game Neverending, which was failing in 2004. The internal photo-sharing tool we'd built for it worked beautifully, so we ditched the game and shipped only the tool. We called it Flickr.

The second was Glitch, which was failing in 2012. The internal chat tool we'd built for the team worked beautifully, so we ditched the game and shipped only the chat tool. We called it Slack. The same thing happened twice and I am not entirely sure I can explain why.

Burbn was a check-in app — half Foursquare, half photo-sharing, half gamified. The numbers were terrible. We had eight weeks of runway and a board meeting coming up.

Mike and I sat down with our user data and asked: which feature do people actually use? It was the photos. Just the photos. So we cut everything else. No check-ins, no points.

We rebuilt the app in eight weeks around three things: take a photo, apply a filter, share it. Eight weeks after launch, we had a million users.

In 2004 I tried to open an online snowboard shop called Snowdevil. I picked the leading e-commerce software of the day and lost a week trying to make it do anything well. I was a programmer; I gave up on the off-the-shelf software and built my own.

By the time the shop was running, I had a more interesting product than the snowboards: I had the platform that ran the store. Friends started asking to use it. I quit selling snowboards.

Today, more than a million stores run on the software I wrote because I couldn't tolerate someone else's checkout.

In 1999 we built OICQ — what would become QQ — as a Chinese clone of ICQ. Nobody used it because nobody was already on it. To make the rooms feel populated, we registered female user accounts in our own office and chatted with the male users who wandered in.

I personally posed as several different women across several different rooms. Within a year there were enough real users that we could quit our second jobs as imaginary girls.

Today, what's left of QQ has more than half a billion accounts.

I gathered seventeen friends in my apartment in Hangzhou and gave a speech I had been rehearsing for weeks. The pitch was that Chinese small businesses needed a way to sell to the world, and that we should build it before someone else did. Each of them put in part of their savings — about $60,000 between us.

We were eighteen people in a Hangzhou apartment trying to build a global business in a country that was barely online. A few months later I flew to California, met Jerry Yang at Yahoo, and asked him to invest. He told me to come back when I had something to show.

I went back to the apartment and built it.

I flew to Sunnyvale to meet Jerry Yang and David Filo of Yahoo. After a five-minute pitch I told them I wanted to give them $100 million for a third of the company. SoftBank did not have $100 million in cash; we had a fraction of it.

I went back to Tokyo, raised the money, sold pieces of the company to do it, and came back with the cheque.

I have made bigger bets since, but that one was the bet that proved to me I could underwrite an idea before I had the means to back it. Almost every major investment I've made since has started the same way.

Our research project at Stanford in 1996 was called BackRub. The idea was to crawl the entire web and rank pages by how many other pages linked to them. The crawler ran on Stanford's bandwidth.

We did not exactly tell anyone how much bandwidth. It eventually consumed nearly half of the entire university's network capacity, brought parts of campus to a crawl for hours at a time, and forced the systems administrators to send us angry email.

We didn't stop. The web at the time was small enough to fit on a few hard drives in our dorm, and we needed those drives more than we needed campus to be happy with us.

In August 1991, I posted a short message to a Usenet group called comp.os.minix. The message said: "Hello everybody out there using minix — I'm doing a (free) operating system (just a hobby, won't be big and professional like gnu) for 386(486) AT clones." I asked what features people wanted.

I called my project Linux as a placeholder; I thought "Freax" was a better name, but my friend Ari Lemmke uploaded the first directory to the FTP server under "linux," and I never changed it.

Thirty years later, Linux runs the majority of all servers on Earth.

Building the streaming engine for Spotify in 2006 was easy. Convincing the four major record labels to license us their music in 2007 was the hard part.

I sat in the offices of Universal, Warner, Sony, and EMI for two years. I rewrote the same pitch fifty different ways. I made the engineers fly to London with me to run live demos when no one believed our streaming actually worked. I refused to launch in any market until all four had signed, because I knew if Spotify launched as a partial library it would just be another iTunes.

We finally went live in October 2008, two years after the company started. By the time we did, every major label had a license — and a small piece of equity in us they'd negotiated as part of the deals.

In May 2009 I was a 37-year-old engineer who had just been rejected for jobs at both Twitter and Facebook. I tweeted: "Got denied by Twitter HQ. That's ok. Would have been a long commute." Three months earlier I had tweeted, "Facebook turned me down. It was a great opportunity to connect with some fantastic people. Looking forward to life's next adventure."

I joined my old Yahoo coworker Jan Koum, who had this strange idea about a status app that you could see next to a contact's name in your phone book.

Five years later, the company that had turned me down bought our app for $19 billion.

I bought a small computer-graphics division of Lucasfilm in 1986 for $5 million and put another $5 million in cash on the balance sheet. By 1991 I'd sunk a total of around fifty million dollars of my own money into the company.

I sold custom hardware nobody wanted, and short animated films nobody bought. There were two years where I tried to sell Pixar to Microsoft, to Hallmark, to anyone with money. Nobody would take it. The only thing keeping us afloat was a little side contract with Disney to make a feature-length animated film. It got delayed twice.

When that film finally shipped — Toy Story, 1995 — I took the company public a week later. The IPO was worth more than NeXT, more than Apple at the time, more than I had any right to expect.

It took me five years and 5,127 prototypes to build a vacuum that didn't lose suction. Each prototype was made by hand in a workshop in the back of a coach house I rented in Bath. Each one was almost identical to the last but for one tiny variation.

By the time I had a working version, I had no money left. I tried to license the technology to Hoover and Black & Decker and every major appliance maker in Europe and the United States. They all refused — they made too much money selling vacuum bags every year to want a vacuum that didn't need bags.

So I started my own company in 1993. I was forty-six. We had to sell our family home to do it.

In 2006 I had a manuscript and twenty-six different working titles for it. I needed to know which title would actually sell. So I bought Google AdWords for each one. I ran identical ads with identical landing pages — the only thing that changed was the book title in the ad.

The cost-per-click on "The 4-Hour Workweek" was the lowest of any title; people were clicking it like crazy. I changed the manuscript title that week.

The book launched in April 2007. It sat on the New York Times bestseller list for over four years, and in retrospect every publisher in New York had rejected it more than once.

I was a master's student at the University of Calgary in 2001, supposed to be writing a thesis. Instead, I built a browser toolbar that had one button: Stumble. Click it and you'd be sent to a random web page that someone had thumbed up — and the more you stumbled, the better the algorithm got at picking pages you'd like.

I shipped it from a dorm room. Within five years it had ten million users. It made the procrastination of a generation of office workers feel slightly purposeful, and it eventually sold to eBay for $75 million.

I built it to avoid writing my thesis. I never finished the thesis.

To launch Spectacles in 2016 we did not build a website store, host a press event, or take pre-orders. We built a yellow vending machine roughly the height of a person, called it a Snapbot, and announced its location twenty-four hours before it appeared.

The first one was placed near Big Sur, California. The second appeared on Venice Beach. People drove for hours and stood in line for more hours to buy a single pair of camera-glasses they couldn't get any other way.

Lines around the block; press coverage that money could not have bought; a global product launch with zero marketing spend.

I was a Reno tailor in 1872 making heavy trousers for miners and ranch hands. They kept tearing at the pockets and the fly. A customer's wife asked me to make her husband's pants "so they wouldn't fall apart."

I tried something silly: I bought small copper rivets from a saddle-maker, and I hammered them through the corners of the pockets. The pants stopped tearing. Word got around. Within months I had more orders than I could fill.

I knew the design was patentable, but I didn't have the $68 to file. I wrote to my fabric supplier, Levi Strauss in San Francisco, and asked him to be my partner on the patent. He agreed. The patent was granted on May 20, 1873. The design we patented was blue jeans.

In 1946 the cosmetics business was almost entirely demonstration in fancy department stores. I had no money for advertising, and at first no department store would let me set up a counter.

I started with a single rule: if a woman walked past me, I'd offer her a sample of cream — for her hand, her face, her eye, anywhere. I gave samples by the thousands.

Eventually a buyer at Saks asked, "Why do my customers keep asking me for these jars by name?" Because of the samples. The buyer placed her first order; the lipsticks and creams sold out in forty-eight hours. The strategy of giving away the product to sell the product had no name in 1946. They call it "sampling" now.

In 1913 the turnover at the Highland Park assembly line was 370 percent a year. The work was repetitive, the wages were $2.34 a day, and we couldn't keep workers long enough to train them well.

On January 5, 1914, I announced that Ford would pay every assembly-line worker $5 a day — more than double the going rate. The press called me a socialist. Other car companies tried to have me indicted for upending the Detroit labor market.

Within months our turnover collapsed, productivity rose, and our workers — for the first time in industrial history — could afford to buy the thing they were building.

My husband was riding a horse from Buenos Aires to New York and I had two daughters and almost no money. I opened the first Body Shop in March 1976 in a small Brighton storefront.

The walls were damp, so I painted them dark green to hide the stains — that became the brand color. I couldn't afford new bottles, so I offered customers a refill discount and reused glass bottles from local pharmacies — that became "sustainable packaging" decades before anyone used the phrase. I couldn't afford an advertising budget, so I wrote my own labels longhand and told customers stories about where the ingredients came from.

By 1991 we had over 700 shops in 41 countries.

I left Tinder in 2014 in a way I'd rather not relive. Andrey Andreev — the founder of the European dating app Badoo — reached out about starting a women's social network with me. I told him no, but I had a different idea: a dating app where only women can send the first message.

He listened. He funded it. We called it Bumble. The premise solved a single, unglamorous problem in dating apps that nobody else had cared to fix.

When Bumble went public in February 2021, I was the youngest woman to take a U.S. company public, and the first to do it while holding my baby on my hip during the bell-ringing.

In 2015 I was sleeping on the floor of San Francisco International Airport between meetings, because I couldn't afford a hotel. I was a thirty-five-year-old gay Black woman who had been a music tour manager — every part of that biography was the wrong part for tech.

I read every book in the SFO bookstore on venture capital. I emailed strangers on LinkedIn and convinced them to take coffee meetings.

I started Backstage Capital with the explicit goal of investing only in founders who were Black, women, or LGBTQ — the founders who got less than ten percent of all venture funding. By 2018 I had backed over a hundred companies. The story was the strategy; the strategy was the story.