by Isaac, Mike
The CEO loved it. Kalanick’s approach to management was to let each department fight for control until a winner emerged. It was the fairest way, in his mind, to determine the most talented staff.
What Kalanick didn’t see—or chose to ignore—was all the politicking that went on behind his back. Everyone knew you couldn’t really challenge TK—the staff’s nickname for Kalanick—if you wanted to stay on his good side.
If you were bold enough to challenge Kalanick, you had better back up your arguments with cold, hard data. Kalanick wouldn’t listen to anything else. For years, general managers begged Kalanick to let them build a tipping function into the app so that riders could toss a few extra dollars to drivers at the end of a ride. It was a simple gesture that would earn the company significant goodwill with their driver base; besides, Lyft offered it. Yet Kalanick remained staunchly against tipping. Kalanick felt Uber worked so well because of the frictionless payment experience. A passenger could exit the vehicle without even thinking of money; adding a tipping function would require the rider to open the app again, needlessly, in Kalanick’s mind. He never understood—or never cared—how much tipping could improve Uber drivers’ livelihoods.
Sometimes, people pushed back. Kalanick once went head to head with Aaron Schildkrout, a tall, wiry product leader who would later rise to become the head of Uber’s driver product division. Schildkrout was sharp, a naturally argumentative hipster type. He almost always dressed in black, wore thick-rimmed dark eyeglasses and rarely combed his mop of dark brown tousled hair. Discussions with Schildkrout often ended up in the realm of the philosophical; he studied social theory at Harvard and the University of Chicago, and loved to think about why humans did things, not just what they did. Before Uber, Schildkrout had been a CEO of his own company, a dating startup, and he learned valuable lessons over his few years as a leader. One of the most important was knowing when a CEO needed challenging.
In one meeting, Kalanick had made a product decision, one of dozens he made throughout the day. But Schildkrout disagreed. He started rattling off a list of things that could go wrong with Kalanick’s plan, and how he’d do it differently. The two went back and forth until they were shouting at each other across the room, a table full of silent employees sitting awkwardly between them. Schildkrout was wearing a grey pullover sweatshirt when he walked in the room. By the time the two had stopped fighting, the front of Schildkrout’s sweatshirt was soaked through. But he had won the argument, and Kalanick respected him for it.
For young, promising engineers, winning Kalanick’s favor was worth it. He was a great motivator, something between a wartime general and a self-help guru. Kalanick always positioned Uber’s fight as “us against them.” If Kalanick thought you were a true believer, someone who stayed “super pumped” for the cause, he noticed. Maybe he’d reward you with some attention, a quick “jam sesh” together in Uber’s lobby, taking a lap and firing off ideas. Or a shout-out in an employee all-hands meeting. Whatever it was, employees loved being in his favor, and fought to stay there.
One of the highest honors for those in Kalanick’s inner circle was admittance to clandestine, 10 p.m. strategy meetings at Uber’s headquarters. In these evening sessions, Kalanick and a hand-picked crew would concoct new ways to spend the mountains of venture capital he had raised to battle competitors. Kalanick loved giving projects code names. He dubbed his late-night strategy meetings the “North American Championship Series,” or NACS, a nod to Uber’s competition with Lyft.
The luckiest employees got to work on “Black Gold,” the code name for Uber’s Asia strategy meetings. The name was special: “Black Gold” was a reference to political corruption, the “black” deeds carried out by gangs—the infamous Triads of organized crime in Taiwan.
For Uber, it meant playing dirty, because they were facing off against a Chinese competitor with a hell of a lot of money.
Chapter 15
EMPIRE BUILDING
For decades, Western technology executives have dreamed of successfully launching an American software business in mainland China. Very few have succeeded.
When Travis Kalanick looked at the country he saw a near-perfect market for startups. Home to nearly 1.4 billion people, China presented an untapped ocean of potential Uber users. Nearly one-third of that population were millennials: young, urban, upwardly mobile with growing disposable income, ardent students of technology and the sciences, and almost always connected online.
As in America, this Chinese generation had grown up with ubiquitous access to the internet. Nearly 97 percent of Chinese internet users ages fourteen through forty-seven owned some sort of smartphone. Westerners had experienced the mass migration from desktop computer to smartphone. But China’s millennials skipped the desktop, advancing directly to their phones. Like Kalanick, the Chinese believed in technology, embracing it much faster than Westerners. Kalanick needed them to embrace Uber.
This was far easier said than done. Larry Page and Sergey Brin, Mark Zuckerberg, Jeff Bezos, Dick Costolo, Evan Spiegel—almost all of the most influential Silicon Valley leaders of the past twenty years have made overtures to China to operate inside the coveted country. And almost all of them have failed. Every company in Silicon Valley had—has—their own version of the China problem.
Kalanick projected confidence that he knew how to crack the market. The Chinese were still reliant on taxis, Kalanick’s most hated foe. He was convinced once they saw Uber’s better service they’d adopt Uber en masse. Moreover, he had another secret weapon: billions of dollars in free driver and rider subsidies, more than enough, he believed, to ignite Chinese consumer demand. China was going to be Uber’s toughest battle yet. Privately, he harbored doubts he would be able to beat the Chinese on their own turf. Even so, Kalanick relished the coming fight.
He feared the Chinese government would be hostile. The Communist Party took pride in promoting and ensuring the success of Chinese companies on Chinese soil. Under Xi Jinping, the government had invested hundreds of millions in state-backed venture funds, which seeded a wave of startups, giving China the fastest growing economic sector in history. It had created so-called “special economic zones” in cities like Shenzhen, fostering Chinese innovation and startup incubation. The West still maintained global tech dominance, but of the top twenty technology companies in the world measured by market cap, nine of them were Chinese.
Government control of the internet meant the Party could play kingmaker, choosing to regulate selectively based on what it felt was beneficial to the state. It was against their nature for the Chinese government to warm up to a foreign invader, especially one as brash and hard-charging as Kalanick. Xi read the news, and surely knew Kalanick’s reputation. Nevertheless Kalanick was confident he would prevail.
The only thing that worried Kalanick more than China’s Big Brother–esque government was another “brother,” this one a startup: Didi Chuxing. Roughly translated, it meant “Brother Travel.” Colloquially, it was “honk honk taxi”: “DiDi” mimicked the sound of a car’s horn in Chinese.
While the name sounded playful, the company and its leadership were anything but. Didi Chuxing was the preeminent ride-hailing startup in China, built on years of analysis of how China’s billion-plus citizens travel the country’s congested streets. Cheng Wei, Didi Chuxing’s CEO, was in his late twenties when he started the company, having only held a handful of jobs in sales prior. His bet on building a taxi-hailing business in 2012 ballooned into a multi-billion-dollar ride-hailing giant in just three years, backed by heavy-hitting venture investments from Tencent and Alibaba, two of China’s biggest and most popular technology companies.
Didi Chuxing had everything Uber needed to succeed: scale, recognition, and, most of all, support from the government. It also had incredibly deep pockets, having raised billions in capital from Chinese investors over just a few years of operation. The top brass there had mettle, too. Before becoming Didi Chuxing, Didi
Dache had been locked in a spending war with a major competitor, Kuaidi Dache. In 2015, the two companies came to a truce and merged, but not before burning millions of Yuan worth of free rides offered to users. By the time the two companies merged, app-based ride-hailing was woven into the fabric of everyday life.
Kalanick wasn’t fazed. He had overcome corrupt politicians and taxi unions in every major city in the United States. He had fought Lyft and outsmarted its leadership. He had charged into cities worldwide, outspent his opponents, outflanked governments, and won more customers with a better product. Barreling ahead had worked before, and it could work again.
“I get excited about doing things people think can’t be done,” Kalanick later said.
Travis Kalanick was excited to accomplish the impossible. But Thuan Pham had to deal with the day-to-day reality of it. And by 2015, Pham was already swimming in problems.
Uber was growing faster than anyone could have anticipated. Pham had installed teams in several dozen cities across China—Chengdu, Beijing, Wuhan, and many others—putting up an ample fight against the aggressive DiDi. That made Travis happy. And when Pham’s boss was happy, Pham was happy.
As chief technology officer of Uber, Pham was responsible for the entire engineering corps inside of Uber, a sprawling organization of hundreds of brilliant young hackers. Pham’s team looked up to him; a disciplined executive with dark hair, bronzed skin, and square, gunmetal-colored glasses that cut a sharp contrast to his smile. To many of his engineers, he was a rare breed of CTO: empathetic to his staff and even emotional when dealing with tough company issues. His workers repaid him with their loyalty. Most of all, they respected his work ethic, especially his ability to respond to subordinates’ emails at all hours of the day. Even when he took a vacation, Pham had his laptop open at the airport, answering emails on the runway up until the moment the flight attendant told him it was time to stow his electronic devices.
But today, Pham wasn’t happy. Even as the number of trips people took in China were skyrocketing, so were Uber’s incentives.
Everyone inside of Uber knew it was going to take an enormous number of free rides to gain significant market share in China. To head off concerns about the burn rate, Kalanick wrote a letter to investors warning them Uber may be spending heavily to gain a stronger foothold in China.
As CTO, Pham saw something the investors couldn’t. Uber was spending $40 million to $50 million on subsidies in China every single week, an enormous sum just to convince riders and drivers to use Uber over DiDi.
When press started sniffing around, Kalanick sent Emil Michael to fend off nosy reporters, telling the press that Uber’s operations in China were “more efficient” than many believed. If press had any inkling of what the numbers were—something Michael refused to discuss in interviews—their collective jaw would hit the floor.
In a letter Kalanick wrote to investors, that then leaked to the press, he noted that in just nine months, the number of trips taken in Chengdu and Hangzhou were more than four hundred times the number of trips taken in New York, one of Uber’s largest cities, when those markets were the same age. “This kind of growth is remarkable and unprecedented,” Kalanick wrote in the letter. “To put it frankly, China represents one of the largest untapped opportunities for Uber, potentially larger than the US.”
What Kalanick left out was the fact that in many cities, more than half of those trips were fake, a complete waste of money brought in by investors.
Pham, who was responsible for dealing with the fraud, wasn’t a stranger to tough situations. Born in Vietnam in 1967, Pham was thrown into conflict as young as twelve years old, as his mother piled him and his brother into a rickety wooden fishing boat, sailing out amid the rough waters of the South China Sea to escape the violence of the ’79 Sino-Vietnamese War. Pham weathered deadly storms, was robbed by Thai pirates, and felt lucky to land at a refugee camp in Indonesia, even though he was soon shuttled off to an island for other Southeast Asian immigrants that lacked basic sanitation. After finally making it to the United States, Pham shared a small, roach-infested apartment in Maryland with another family while his mother worked multiple jobs to support them.
Pham studied hard in school and grew to love IBM PCs, which he discovered at a young age. An entry-level computing job at Hewlett-Packard led to the startup world and, eventually, a big break with a high-paying, high-level job at Uber as CTO. Like Kalanick, Pham worked hard, and never flinched from the pressure, intensity, or long hours as the company exploded.
The China challenge, though, was unprecedented. Pham’s job was to turn the China strategy, an economic nightmare, into something that actually made financial sense.
Pham spun up a crisis team, poaching top security and fraud detection talent from local Bay Area competitors to form a fifty-person fraud squad at Uber’s HQ in San Francisco. He ordered local managers in China to review new sign-ups more rigorously. They introduced identity verification features and other tactics to screen drivers and riders over time.
China wasn’t the only porous market. Scams were endemic in every market worldwide. In New York in 2014, insiders noted that nearly 20 percent of the company’s gross revenue went to fraudulent rides. Thieves skimmed the same amount in London. From Washington DC to Los Angeles, Uber was bleeding money by the millions in its most important markets.
Pham’s fraud specialists soon proved invaluable—and not just in China. In Brooklyn, the team watched as credit card thieves used stolen card numbers to run drug trafficking and prostitution rings using Uber vehicles. The ruse was simple: the dealers would buy stolen credit card numbers from the Dark Web, then plug those numbers into the app to charge Uber trips to the stolen accounts. Over hundreds of trips per week they delivered drugs and call girls throughout New York City—all paid by Uber incentives, or through chargebacks from credit card companies after the original card owners reported the fraud.
After monitoring the criminals for months, Uber eventually partnered with the New York Police Department to help take the scammers down in a complicated sting operation. Over the course of a single Uber ride, the police would obtain a report from a credit card company, call the driver of the vehicle and tell them to pull over, then arrest the rider on a number of charges, including credit card fraud, possession of narcotics, prostitution, and so on. Though they would never brag about it publicly, the fraud team helped the NYPD take out the entire operation.
Pham’s crew then built machine-learning models on past criminal behavior, training Uber’s systems to spot patterns of fraud as it occurred. After the team found its stride, fraud in markets like New York dropped to the low single-digit numbers; Pham was proud of his team, and so was Kalanick.
Fighting Chinese fraud, however, required another level of vigilance. Chinese scammers engaged Uber’s engineers in an aggressive arms race in which the anti-fraud team fought to overcome the fraudsters’ increasingly clever methods. In the United States, fraud was simpler; criminals usually either went for joyrides or used Uber to make illegal deliveries, all using stolen credit card numbers. But in China, drivers and riders colluded to scam Uber out of billions in incentives, divvying the rewards.
Most scammers found each other over text-based Chinese internet forums, a simple, anonymous way to match people who wanted to make a quick buck. They developed their own codified language; drivers seeking a fake ride would ask for “an injection,” a reference to the small, red digital pin that signaled a user’s location inside the Uber app. A “nurse,” or scammer, could respond in kind to give a “shot” to the original poster by creating a new fake account and going on a fake ride with the driver. The two parties would then split the bonus incentive payment from Uber. Repeated over and over across dozens of cities, small driver bonuses mushroomed into millions in squandered cash.
Kalanick couldn’t stop the incentives because he had to keep pace with DiDi, who were just as willing to bur
n money to attract ridership. To juice growth, Kalanick had made the new user sign-up process as simple as possible. Joining Uber only required a name, email address, phone number, and credit card number, all of which were easily replicable. Fraudsters simply entered fake names and emails. Then they used apps like “Burner” or “TextNow” to create thousands of fake telephone numbers to be matched with stolen credit card numbers. But requiring Chinese users to add other, more precise, forms of identification would add more friction to the process. And, as Kalanick’s data scientists found in their research, adding friction slowed growth. For Kalanick, putting a dent in growth was not an option.
Kalanick’s solution was to grow and rely upon the anti-fraud team. But scammers grew more shrewd over time. Eventually, hustlers found that searching forums for riders was inefficient and time-consuming, so they ended up creating “riders” themselves. Some drivers would purchase caseloads of cheap cell phones, creating multiple driver and rider accounts for each different phone. A scammer would request rides from his “passenger” phones, and use his “driver” phones to accept those rides. He would then drive around the streets of Chengdu with dozens of phones spread across the front and back seats of his car, racking up fares for each of the “trips” he was completing for his fake customers.
The fraud team eventually discovered the trickery. Little blips flickered along the complicated topography of the Chinese city maps on monitors lining the control rooms back at Uber headquarters in San Francisco. Each little blip represented a scammer’s vehicle, which was followed closely on the map by a trail of a dozen fake “passengers” who were taking the trip. It was as if the fraud team were watching dozens of digital centipedes skittering across their computer screens, each one getting fat on Uber incentives.