The Upstarts: How Uber, Airbnb, and the Killer Companies of the New Silicon Valley Are Changing the World

by Brad Stone

room and they are misinformed. They are thinking mostly about where their next campaign contribution comes from. There’s an Uber rep there, but he is basically alone, trying to describe an unfamiliar and strange technology to people who have no understanding of it. “The Uber guy has a lobbyist, but the lobbyist is also working for the other guys on the side. Finally, you have the big taxi guys there, and they have the city council locked up and paid for. “Then, meanwhile, you’d cut to the taxi guys at the airport. They are all waiting there for hours, playing cards or whatever, for the chance to pick up one fare. And there’s an Uber recruiter there, and he is surrounded, explaining this new system to the drivers…” Kalanick caught himself and trailed off. “Anyway, that’s how you’d start the movie.” On the street after dinner, he said again, “You have to inspire me,” as if I hadn’t just spent two hours laying out my best arguments. Then he and Plouffe took off on foot, back to the office. Six months passed, and despite my repeated entreaties, I didn’t hear anything. But then, after I’d talked to dozens of regulators, competitors, and current and former Uber employees, a new Uber PR executive somehow convinced Kalanick to cooperate. I eventually spoke with two dozen different Uber executives from all periods of the company’s brief history and had another few hours of Kalanick’s time to complement several interviews I had conducted with him over the course of my five years as ...

a pecha-kucha (Japanese for “chatter”), an event in which a series of designers present their new product ideas by showing twenty slides apiece and discussing each slide for twenty seconds.

“When you are a designer at school, especially an industrial designer, you just dream of getting something on the shelf,” Chesky says.

So a few weeks later, on September 22, 2007, with the World Design Congress coming to San Francisco and the city’s hotels either overbooked or overpriced, Gebbia sent Chesky the e-mail that would change their lives: Subject: subletter Brian, I thought of a way to make a few bucks—turning our place into a designer’s bed and breakfast—offering young designers who come into town a place to crash during the 4 day event, complete with wireless internet, a small desk space, sleeping mat, and breakfast each morning. Ha!—Joe. It took Chesky and Gebbia three days to put together the first Airbedandbreakfast.com website using the free tools on the blogging website WordPress.

How AIRBNB started

When you open up that app and you get that experience of, like, I am living in the future, like I pushed a frickin’ button and a car showed up and now I’m a pimp—Garrett is the guy who invented that shit! Like, I just want to clap and hug him at the same time. —Travis Kalanick1

In May 2007, eBay bought StumbleUpon for seventy-five million, turning it into one of the early successes of what became known as Web 2.0, the movement in which companies like Flickr and Facebook mined the social connections among internet users.2 For Camp, it seemed at the time like the highest possible level of success in Silicon Valley, and it was, by any reasonable standard—until the one that he achieved next.

For decades, San Francisco had deliberately kept the number of taxi medallions capped at around fifteen hundred. Medallions in the city were relatively inexpensive and couldn’t be resold, and owners could keep the permit as long as they liked if they logged a minimum number of hours on the road every year. So new permits usually became available only when drivers died, and anyone who applied for one had to wait years to receive it. Stories abounded about a driver waiting for three decades to get a medallion, only to die soon after. The system guaranteed a healthy availability of passengers for the taxi companies even during slow times and ensured that full-time drivers could earn a living wage. But demand for cars greatly exceeded supply and so taxi service in San Francisco, famously, sucked. Trying to hail a cab in the outer neighborhoods near the ocean, or even downtown on a weekend night, was an exercise in futility. Getting a cab to take you to the airport was a stomach-churning gamble that could easily result in a missed flight. (Even when a passenger arranged for a taxi via phone, he couldn’t be sure the cab would show up; the driver might decide to pick up a street hail instead.) Attempts to improve the situation were fruitless, since the fleets and their drivers were adamant about limiting competition. Over the years, whenever the mayor or the city’s board of supervisors tried to increase the number of medallions, angry drivers would fill city council chambers or s...

The sparkling City by the Bay beckoned, but Camp had no reliable way to answer its call. Habitually restless, frustrated by inefficiencies, and armed with a willingness to challenge authority, Camp came up with his first attempt at a solution: he would call all the yellow-taxi companies when he needed a cab. Then he would take the first one that arrived. Not surprisingly, the cab fleets didn’t like that tactic. Though it is impossible to confirm, Garrett Camp believes his cell phone was blacklisted by the San Francisco taxi companies. “Eventually they wouldn’t take my calls,” he says. “I was banned from the San Francisco cab system.”

Birth of Uber idea?

To solve these mounting transportation challenges, Camp started to experiment with the city’s gypsy-cab fleet—the unmarked black sedans that would approach prospective passengers on the street and flash their headlights to solicit a fare. Most San Franciscans, particularly women, would stay away from those unmarked cars, fearing for their safety or worried by the ambiguity of a cab without a running meter. But Camp found that a majority of the cars were clean and that many of the drivers were friendly. The biggest problem for these drivers was filling in the dead time between rides, when they tended to wait outside hotels. So Camp started collecting the phone numbers of town-car drivers. “At one point, I had ten to fifteen numbers in my phone of all the best black-car drivers in San Francisco,” he says. Then he started gaming the system further: texting a favorite driver hours before he needed him and telling him to meet him at a restaurant or bar at an appointed time. On another night, he rented a town car and driver for himself and a group of friends for an entire evening. It was an indulgence that cost a thousand dollars, and zooming around the city at the end of the night dropping everyone off was a pain. And that is when the futuristic image from the James Bond movie Casino Royale popped into Garrett Camp’s head.

How uber got born

Both the passenger and the driver could have an app on their phones. The passenger could have a credit card on file and wouldn’t have to travel with any pesky cash. “I bounced the idea off of everyone,” Camp says. “All these ideas kept building and building.” The original idea was to buy cars, then share the fleet among his friends who were using the app. But Camp says that was only a starting point and that even back then he was considering the potential to use such a system to coordinate not just black taxis but eco-friendly Priuses and even yellow cabs. “I always thought it could become a more efficient cab system, particularly in San Francisco,” he says. He wasn’t sure it would work outside the city, though. If he could get it to work in just a hundred cities, he reasoned, it could be big enough for a company that generated about one hundred million dollars a year in service fees.

“I have this idea. In San Francisco it’s hard to get a taxi. I want to buy five Mercedes,” Camp said, taking out his phone and showing him a picture of a Mercedes-Benz S550, a high-end coupe that sold for around a hundred thousand dollars. “I’m going to buy the cars with some friends and we’re going to share drivers and the cost of parking.” He showed mock-ups of iPhone screens demonstrating how cars would move on maps and how passengers might see a town car coming toward them.

Basic concept of Uber

A few weeks into 2009, after their trip to Washington, DC, to see Barack Obama’s first inauguration as president, Camp called Kalanick. He was about to lease parking spaces in a garage near his home on Hawthorne Street in San Francisco for the fleet of Mercedes he was still determined to buy. Kalanick counseled him against it one last time: “Dude, dude! You don’t want to do that!” Camp finally gave in and ended the ongoing debate; he never signed the lease and never purchased the cars. Instead of buying a dozen flashy Mercedes, Camp, along with Kalanick, would pitch the app to owners and drivers of town cars. Kalanick would brag a few years later, in one of our first interviews: “Garrett brought the classy and I brought the efficiency. We don’t own cars and we don’t hire drivers. We work with companies and individuals who do that. It’s very straightforward. I want to push a button and get a ride. That’s what it’s about.”

Uber in a few paragraphs

Meanwhile, Graves, the CEO, and Kalanick, an adviser now spending about twenty hours a week on UberCab, cold-called and visited San Francisco town-car fleets and pitched the service to the owners. “It was old school dialing for dollars,” Kalanick would later say.6 “A third of the calls, you know, basically I got hung up on before I got to the core pitch. A third of the calls they heard for about a minute and a half and then I got hung up on. And a third were like, this is interesting.”

At the time, Uber had been in the iOS App Store for all of two weeks. Ryan Graves and Travis Kalanick had succeeded in recruiting about ten town-car drivers in San Francisco. The service was facilitating ten rides a weekend, with most of those probably taken by Uber’s own employees, founders, and friends.

Now Uber had $1.3 million in the bank, a $5.3 million valuation, an office (small and crowded), and a product (buggy). It was finally looking like a real startup. Uber’s founders and investors told their influential and affluent San Francisco friends, and word began to spread. On July 5, the blog TechCrunch wrote its first story about the app: “UberCab Takes the Hassle Out of Booking a Car Service.”

By the fall of 2010, San Francisco was starting to notice Uber. The service was exceedingly viral; a user stepped out of a town car and walked into a bar, and suddenly his or her friends wanted to know everything about it. Limo and town-car drivers were also intrigued. They started showing up in the Uber office, one by one. Conrad Whelan recalls watching Graves give a driver the pitch and train him to use the app. Afterward, the driver laughed. “Oh, you guys are going to make a lot of money.” That’s when Whelan canceled his vague plans to return to scientific research.

That fall, complaints by taxi drivers and yellow-cab fleet owners about a new unlicensed competitor started to pour into the offices of city and state regulators. They claimed it was illegal and should be shut down. So on October 20, 2010, four months after the launch, when Graves was at a board meeting at First Round Capital with Travis Kalanick and Garrett Camp, four government enforcement officers walked into the tiny UberCab office. Two were from the California Public Utilities Commission, which regulated limousines and town cars, and two were from the San Francisco Municipal Transportation Agency, which regulated taxis. The plainclothes officers flashed badges, and then one of them held up a clipboard with a cease-and-desist letter and a large, glossy head shot of a smiling Ryan Graves. Waving the photograph around the room, he demanded: “Do you know this man?”

Uber's first regulatory crisis

Wolpert set up Cabulous in San Francisco at a startup incubator called Pivotal Labs, hired a developer, and started working on an app for smartphones. He also started cultivating drivers. The city’s largest taxi fleet, Yellow Cab, had a ten-year technology contract with one of the old-school dispatch companies, and another fleet, Luxor, was using Taxi Magic. But two other taxi companies, DeSoto and SF Green Cab, allowed the startup to pitch its drivers directly. Wolpert remembers long hours sitting in the front seats of cabs, learning the trade and coming to love the city’s grizzled taxi drivers, who had waited years for their medallions and were clinging to a reliable stream of income. “There were so many cool guys,” Wolpert says. “Old Harry Chapins, still driving. Just salt of the earth.” Wolpert was conceiving of a service that empowered those yellow-cab drivers. It would make the traditional taxi businesses more efficient and help drivers boost their earnings. This was his fatal mistake.

The app debuted on the App Store in the fall of 2009, more than six months before UberCab, and offered some of the elements that would later make Uber special. Unlike Taxi Magic, Cabulous showed the images of cabs on a map, and riders could either hail them electronically or call the fleet’s dispatch number. (They could also look for their favorite driver and summon him specifically.) There were a few bells and whistles too. When users pulled up the app, they heard the sound of a car door opening and shutting and a jet engine taking off. It was totally frivolous, but years later, Wolpert plays the old sound and smiles.

Unlike Uber, Cabulous did not automatically facilitate payments; riders still had to manually pay the driver based on the fare listed on the meter.

The biggest problem was that Cabulous didn’t control its supply of drivers or its fares and couldn’t grow its fleet to keep up with demand. So when cabbies got busy on weekend nights and had plenty of street hails, “they just wouldn’t turn on the app,” says Tal Flanchraych,

When he heard that San Francisco officials had served UberCab with that first cease-and-desist notice, adorned with a head shot of Ryan Graves, Wolpert believed it was justified. Regulation had a purpose. Taxi prices needed to be strictly controlled so that grandmothers could afford a ride home from the supermarket.

What he found even more objectionable was that Uber was using the iPhone as a taximeter to calculate fares. Meters are traditionally calibrated and closely monitored by cities’ weights and measures departments to protect customers from price gouging. He and Graves, who had been friendly since their first meeting and had attended some MTA meetings together, argued about it over the phone. “Hey, let’s completely disregard decades of regulation!” Wolpert shouted at him. “How is this a good idea?”

Wolpert left Cabulous in 2011 after Uber was starting to drive laps around it, ceding the fight to a more experienced CEO. Years later, when the company changed its name to Flywheel, the brand started appearing all over the exteriors of San Francisco taxis owned by DeSoto Cab, part of a co-marketing agreement.2 Wolpert now gets emotional when he sees it. “I may not be rich from this, but I changed the face of the city,” he says. “It makes me happy.” He concedes that it was probably a mistake to partner with yellow-cab drivers and taxi fleets, which were handcuffed by regulation and ill prepared to counter the disruptive threat posed by Uber. “It was like watching a shark devour a seal,” he says. “We are living in an era of robber barons. If you have enough money and can make the right phone call, you can disregard whatever rules are in place and then use that as a way of getting PR. And you can win.”

Logan Green, a young, introverted software engineer who grew up amid the transportation chaos that was Los Angeles in the 1990s. In high school, Green got a part-time job working for the celebrated video game entrepreneur Nolan Bushnell, the founder of Atari and one of the first bosses of Apple co-founder Steve Jobs. Green went to the hippie-ish New Roads High School in Santa Monica and for years navigated the gridlocked city streets in his beat-up 1989 Volvo 740, commuting to Bushnell’s gaming company, uWink, in Playa del Rey.

Green became the youngest member of the Santa Barbara Transit Board and got a full education in the grim economics and politics of the public bus system. Seventy percent of every bus ride had to be subsidized by the city. The quality of service was low but attempts to raise fares and add sales tax frequently buckled under a wall of local opposition.

During Logan Green’s senior year back in Santa Barbara in the fall of 2005, the pieces started coming together in his mind—the ridesharing channel on Craigslist; the crowded vans in Victoria Falls; the intractable flaws of the public-transit systems. He started working on a concept he called Zimrides (short for Zimbabwe rides). The idea was to use the internet to fill every open seat in every car.

Lyft began here

In December of 2006, Zimride’s first app, called Carpool, allowed university students to post on Facebook, specifying where they were traveling to and looking for rides with others going in the same direction. Across the country, a recent graduate of Cornell saw the app and was fascinated. A student at Cornell University’s school of hotel administration, John Zimmer had learned that the key to running a profitable hotel business was high occupancy and great hospitality. The transportation status quo offered neither. “If you take public transit and taxi and think of that as a hotel, those would be hotels you would not want to stay in,” he told me later. “They would be failing businesses.” Inspired by Zimride’s Carpool app, Zimmer (the similarity of the company’s name to his own was coincidental) got a friend to introduce him to Green and they decided to team up in a virtual cross-country partnership.

Internet marketplaces thrive when buyers and sellers are matched in ways that wouldn’t otherwise be possible, saving everyone time and money.

All the lethal mistakes of the other nonstarters were wrapped up in Zimride. The founders were too nice. They were idealistic. Their idea was too early—the great wave of smartphone ubiquity and social networking was just gathering momentum. But they were also pragmatic, and they believed in that Silicon Valley notion referred to as “the pivot.” As long as there is money in the bank, it’s never too late to change business models and seek more profitable pastures.

business, they got excited about a mobile version of Zimride that would let regular people share their vehicles not on long trips or daily commutes but every day, anytime, from one point to another within big cities. Inspired by a giant orange felt mustache that decorated the cubicle of an employee, John Zimmer decided to give every driver a pink mustache to put on his or her car fender; the mustache would make the car stand out and appear friendlier to those who might be wary of climbing into a car with a stranger. At first they referred to the new service as Zimride Instant, then changed the name to something a little catchier: Lyft.

Lyft was born from zimride

The four plainclothes enforcement officers who had served Uber with its first cease-and-desist on October 20, 2010, set off a mad scramble inside the company.

The letter threatened penalties of five thousand dollars per ride and ninety days in jail for each day the company remained in operation. But which laws had they broken, exactly?

A few blocks away, on the seventh floor of a former bank building at 1 South Van Ness Avenue, Christiane Hayashi was planning her next move. As director of Taxis and Accessible Services at the Metropolitan Taxi Agency, Hayashi was the most powerful figure in the city’s highly dysfunctional cab industry. She had worked as a deputy city attorney since graduating from UC Hastings College of the Law and had toiled away in environmental law and Y2K compliance. She was no stranger to San Francisco’s bare-knuckle politics, where rival democratic factions battled incessantly and corruption often bubbled indiscreetly under the surface. A stint in the Department of Elections, where she managed the transition away from punch-card ballots, was particularly grueling. Hayashi and two other lawyers were accused of mismanaging funds and signing false time sheets. A special counsel eventually investigated and cleared her.2 The experience “crushed my spirits for a while,” she says. After she was exonerated, she fled city politics and moved to San Cristóbal de las Casas, in the Mexican state of Chiapas, where for a few months she sang in the house band at a local disco. But performing six nights a week in a building she recalls as a hellacious firetrap couldn’t match a safe desk job and a comfortable government pension. In 2003 she randomly ran into a city supervisor while hiking in the Guatemalan jungle and, soon after, was lured back to the Bay, first representing the SFMTA in the city att...

She’s talking about this years later, at a barbecue in the backyard of a friend’s house in Berkeley. She’s in her early fifties and visiting from Las Vegas, where she clerks for a county court and lives on a farm with a view of the mountains. Though she’s recaptured her sense of humor, getting her to reminisce is difficult at first—those years at the MTA were the hardest of her

Hayashi was aware of the limits of her authority. Regulating limos and black cars was the responsibility of the state, not the city. But she saw an opening: this startup was calling itself UberCab and thus seemed to be marketing itself as a taxi company. She talked to the enforcement division at the Public Utilities Commission of California, which was tasked with regulating limos and town cars, and they orchestrated the joint cease-and-desist. After receiving the threatening letter, Uber promptly asked for a meeting.

Hayashi says that she was strident, not screaming, and remembers the Uber execs as “obnoxious” and Kalanick in particular as “arrogant.” “You can’t do this!” she told them. “You can’t just open a restaurant and say you are going to ignore the health department!” She says that nothing was decided at the meeting and calls it “totally pointless.” But that’s not entirely true. In fact, Uber’s first clash with city regulators likely changed the course of this tale.

A self-described C-Span junkie in high school, Kalanick was intrigued

in 2010, Uber was a tiny company. It had some half a dozen employees, a few dozen limo drivers in San Francisco using the platform, and little in the way of expansion plans. Its motto, “Everyone’s private driver,” conveyed luxury and exclusivity, not mass-market appeal.

Uber was turning out to be a company rooted in complex math. Its biggest challenge, and where he found himself already frustrated with the performance of the startup, was finding ways to attract more drivers during peak times and to route cars into the areas of highest demand. Uber had the data to make those kinds of prescient decisions. In fact, it was slowly dawning on the founders and board members that Uber was going to have more data about how people moved around cities than just about any other company in history. “I’m an engineer at heart and math moves the needle on this,” Kalanick told me a few years later. “My happy place is right in the midst of all that complexity.”

Uber having to resolve a transit problem

The company was exhibiting an elusive phenomenon called negative churn, in which users who joined the service were more likely to stay with it and gradually increase their frequency of use than they were to leave. In other words, once customers joined Uber, they turned into a sort of high-yield savings account.

When someone signed up, according to an early internal estimate, he or she was worth forty or fifty dollars a month in gross revenues and eight to ten dollars in gross profit—for the foreseeable future. “This is the equivalent of a perpetual-motion machine and cannot go on forever, but it means that rider spending is accelerating at a rate that is...

The California Public Utilities Commission wanted Uber to register itself as a limo company or, technically, as a “charter party carrier,” but Uber’s lawyers believed the company could make the case that it was merely an intermediary between drivers and riders, not an actual fleet operator. It was as much a limo company as Orbitz or Expedia were airlines, they argued. In a follow-up ruling in late 2010, the PUC agreed, and Uber never stopped operating. To the consternation of Christiane Hayashi, who was trying and failing to get the city attorney’s office to give her authority to regulate the startup, Uber had a winning argument.

Kalanick expressed some of the hard-charging enthusiasm and ambition that he had brought to Scour and Red Swoosh. “The bottom line is that I’m all in on Uber,” he wrote. The excitement and joy of being Uber is coming out my pores and I’ll stop at nothing to see Uber go to every major city in the U.S. and the world. So what’s next? Taxi frustration is going down. Reliability, Efficiency, Accountability, and Professionalism in urban transportation are going way up. Every city Uber rolls into is going to be a better place when we’re done with it and if you live in that city, your world of transportation is changing forever, and it will be oh so Uber when that change arrives.

Kalanick instructed Kochman to ignore New York’s Taxi and Limousine Commission and its rules, reasoning that its regulations, under the guise of consumer safety, were really there to protect entrenched taxi interests. Kochman didn’t necessarily disagree but he had successfully solicited approval from the Ithaca city council to operate MESS Express. He had a history of productive encounters with regulators. So he disregarded Kalanick’s orders and set up a meeting with a TLC deputy. “I wasn’t going to bust my ass to launch something the city was going to shut down immediately,” he says. Kalanick was furious when he heard about the meeting. “He was absolutely livid and said it was insubordinate,” Kochman remembers. After his fury passed, Kalanick flew to New York and they visited TLC headquarters together. The meeting—Uber’s first of many with the TLC—went fine. Stressing that Uber cars were not hailed or even electronically hailed like taxis, the pair emphasized that Uber cars fit the legal definition of livery cars and were prearranged; it just so happened that the prearrangement occurred five minutes ahead of time instead of sixty. The deputy commissioner Ashwini Chhabra, who would join Uber as head of policy planning three years later, initially asked Uber only to customize the app to display a driver’s permit number and base affiliation.

Austin Geidt had been adrift when she started working at Uber. The company’s first intern got the job after being rejected for a barista position, and she struggled to find a meaningful role. But during that first difficult year, she had a dawning realization: just about everyone else around her was making things up as they went along too. After the epiphany, Geidt gave herself permission to look at problems more constructively. When Ryan Graves made her head of driver operations that March after firing another early employee, Stefan Schmeisser, she had ample opportunity to make a difference. After training a driver in the office in San Francisco one day, she walked outside to get coffee and saw the man getting into a pink minivan. It occurred to her that the company should probably conduct vehicle checks to ensure cars were up to Uber’s high standards at the time. Later, she decided Uber should test whether drivers had at least a cursory knowledge of city landmarks (this was before Uber started having riders enter their destinations into the app). She asked Sofiane Ouali, the driver of the white 2003 Lincoln Town Car known as the unicorn, to obtain for her the city’s entrance exam for yellow-cab drivers. Then, with help from Ouali and other drivers, she changed some of the questions to match the expectations of Uber’s upscale, smartphone-using clientele. Instead of asking directions to the city jail, for example, Uber would ask drivers if they knew the location of the Rit...

Geidt and Graves based the Seattle operation on the three-person structure in New York. The general manager supervised the overall business in the city and was accountable for its growth. He or she needed to be entrepreneurial, scrappy, and aggressive in talks with regulators. An operations manager, usually an analytical type like a management consultant or investment banker, was in charge of signing up drivers and making sure there were cars for every passenger who opened the app. Finally, a community manager, a creative type with marketing chops, worked to stimulate demand among riders.

Uber entry into Seattle

Seattle, says Geidt, “became the first iteration of our playbook.”

by registering as a base, Uber was no longer asking drivers to bend the law by accepting rides through a smartphone app.

to really establish an equilibrium between supply and demand during such frenzied nights, Kalanick reasoned, Uber was going to have to remove the cap altogether and let the almighty hand of the market figure out the price. All internal objections were overruled. The company’s overarching goal was to have cars always available, at any time of day or night, and surge pricing could help Uber achieve that.

Why Uber surge pricing