by Brad Stone
We all know that if you swing for the fences, you’re going to strike out a lot, but you’re also going to hit some home runs. The difference between baseball and business, however, is that baseball has a truncated outcome distribution. When you swing, no matter how well you connect with the ball, the most runs you can get is four. In business, every once in a while, when you step up to the plate, you can score 1,000 runs. This long-tailed distribution of returns is why it’s important to be bold.
Nearly a decade after it was conceived, it was still unclear whether the Go store would produce a 1,000-run windfall for Amazon. But it did lead in interesting new directions. Amazon began licensing the “Just Walk Out” system to several other retailers, such as convenience stores and airport kiosks. Amazon Books spawned a few dozen 4-star stores, where the company used its trove of data about people’s buying habits to tailor stores with an eclectic mix of locally popular items. And in 2020, Amazon started opening large Amazon Fresh grocery stores, without Go technology but with the long-gestating Amazon Dash Carts, which allowed shoppers to scan items as they walked the aisles and skip the checkout line.
Another significant outcome was the realization, in early 2016, that Amazon needed to get smarter about physical retail if it ever wanted to seriously compete against giants like Walmart and Kroger in the $700-billion-a-year U.S. grocery industry. Around that time, Steve Kessel joined a cabal of Amazonians that included senior vice president Doug Herrington and members of the Go and M&A teams, to answer a momentous question: whether Amazon should acquire a supermarket chain.
They looked at local grocers, regional chains, and the big national players. Among the calls they placed that year was to the Austin-based Whole Foods Market, the beleaguered organic food chain with cratering same-store sales, a reputation for high prices, and a stock price at a five-year low. But its iconoclastic founder, John Mackey, was confident in his turnaround plan and not ready to sell—yet.
CHAPTER 3 Cowboys and Killers
As Jeff Bezos chased Amazon’s next wave of growth by backing ambitious technology projects like the Go store, Alexa, and Fire Phone, he also opened an online store in India, the country of 1.3 billion people whose cosmopolitan cities were rapidly embracing smartphones and broadband internet access. Over the course of several years, Amazon would sink billions of dollars into the country. His bet there was a renewal of Amazon’s manifest destiny—to sell not only everything, but everywhere.
Bezos had missed an earlier opportunity to invest in India. In 2004, Amazon opened one of its first overseas software development centers in Bangalore, in a small office above an auto dealership. Employees working on its floundering search engine, A9, and its nascent cloud business, Amazon Web Services, repeatedly pitched plans to start a local online store. But as Amazon recovered from the dot-com bust and concentrated its energies on launching in China, India was practically an afterthought.
As a result, some of Amazon’s early employees in India quit to start their own firms. In 2007, two engineers, Sachin Bansal and Binny Bansal—unrelated friends and former classmates at the Indian Institute of Technology (IIT) in New Delhi—left Amazon to start their own company, Flipkart, to try to replicate Bezos’s original magic of selling books online. If Amazon wasn’t going to serve India’s increasingly connected and prosperous upper classes, they would do it themselves.
The Amazon executive who had helped start and run the Bangalore development center was a Bezos disciple and ardent workaholic named Amit Agarwal, also a graduate of IIT. From 2007 to 2009, Agarwal returned to Seattle to become the technical advisor to Bezos, preceding Greg Hart and Dilip Kumar in the crucial role of shadowing the CEO in all of his meetings. At the end of his tenure, he and Bezos had a critical discussion about what the TA would do next. Agarwal asked to join the international division and wrote a business plan to finally introduce Amazon into the country where he grew up.
At the time, Diego Piacentini, Amazon’s then senior vice president for its international consumer division, had mixed feelings about expanding into India. Though companies like IBM and Microsoft had large and successful operations in India, the country had complex laws in place to protect its vast, decentralized sector of mom-and-pop retail shops. These “foreign direct investment” regulations prohibited overseas companies from owning or directly operating a retail business. Piacentini, an Italian who had left Apple to join Amazon in early 2000, also felt that countries with larger gross domestic products should take priority. In 2010, he asked Agarwal to help him bring Amazon to his native country, Italy. A year later, they opened another foreign-language website, in Spain. Those successful introductions, Agarwal said, gave them confidence about “reigniting global expansion.”
As Amazon finally prepped for its incursion into India in 2012, the execs carefully considered some of the difficult lessons they were learning in China. Amazon had entered China propitiously in 2004, acquiring the bookselling startup Joyo.com for about $75 million with the belief that the same approach that had worked elsewhere could succeed in the world’s most populous country. Amazon planned to patiently invest, winning customers with wide selection, low prices, and reliable customer service.
But after a few years of steady progress, everything seemed to suddenly go wrong in China. A well-capitalized e-commerce competitor, Alibaba, opened a popular fixed-price online store for well-known brands, called Tmall, an offshoot of its eBay-like website, Taobao. A few years later, Alibaba expanded a tool called Alipay to let shoppers pay digitally for the products they ordered, while Amazon was still accepting cash from homebound buyers upon delivery. Alibaba and another mounting rival, Jingdong, or JD.com, had cluttered but arresting websites, which catered to the overall design tastes of Chinese internet users. The Amazon.cn website looked like Amazon’s other home pages around the world. Amazon’s China employees were dependent on technical support and other kinds of help from Seattle, and thus were slow to respond to these and other obvious market signals.
A year prior in 2011, Amazon dusted off another page from its global playbook to introduce a marketplace in China, allowing independent businesses to sell their products on the site. This was a key piece of the heralded flywheel: by adding outside vendors, the company drew in new shoppers and earned money from the fees it charged sellers. The extra revenue was then used to lower prices, which in turn attracted more buyers. But again, Amazon failed to adapt to the idiosyncrasies of the China internet—Chinese sellers were accustomed to paying about 2 to 5 percent of their sales to Alibaba, in addition to ads to make their listings more prominent. Amazon execs were skeptical of the advertising model so instead charged 10 to 15 percent of sales, which seemed unusually high to sellers. As a result, Alibaba raced further ahead.
Then a damaging report on China’s state-sponsored television, CCTV, drew attention to counterfeit goods, such as fake brand-name cosmetics, on Amazon’s third-party marketplace; any signs of progress in the country quickly evaporated. Amazon.cn executives from that time said that Bezos was totally uninterested in understanding the inner machinations of the Chinese government, cultivating ties with Chinese leaders, or using his budding fame to help Amazon’s cause in the country, as Elon Musk would do years later to set up a Tesla Gigafactory in Shanghai.
Without a closer relationship with the Chinese Communist Party, Amazon ended up losing even more ground. In an analysis of the struggling China business they delivered to the S-team in 2014, the international team estimated the company had lost a billion dollars in the decade since the Joyo acquisition. Wary of the gathering red ink, Bezos decided to curtail Amazon’s investment in China and set out on a plan to become profitable there instead of accepting the additional losses that would be required to stay competitive in the country.
An Amazon finance exec later described it as the equivalent of “shooting the business in the head.” Between 2011 and 2016, Amazon’s market share in China fell from 15 percent to less than 1 percent. “There was always the fear that if we in
vested a lot in China, we’d get screwed anyway and waste a lot of money,” Piacentini explained years later. “We were not bold enough to go head-on and compete. We always acted as a timid follower.”
On the eve of Amazon’s long-awaited incursion into India, Bezos could consider some of those hard-earned lessons: the company hadn’t invested or innovated boldly enough in China, didn’t cultivate ties with the government, and hadn’t set up operations with enough independence from Seattle. With his former shadow Amit Agarwal eager to bring Amazon to his native country, he wouldn’t make the same mistakes again.
* * *
One of Amazon’s first moves in India was to try to entice its two famous alumni back. Four years after striking out on their own, Binny Bansal and Sachin Bansal had built Flipkart into a nationally recognized brand that sold not only books but mobile phones, CDs, and DVDs. Amit Agarwal met his former employees at the upscale ITC Maurya hotel in central Delhi to discuss an acquisition. Feeling confident about their progress, the Bansals asked for $1 billion. Agarwal scoffed at the amount, and the talks fell through.
After the Bansals thumbed their noses at Amazon, Agarwal started to build a team to compete against them. He prowled headquarters in South Lake Union, zealously pitching a “once-in-a-lifetime opportunity” to impact Amazon and “change the trajectory” of Indian democracy itself. His target was native Indian Amazonians who understood both the company and the cultural peculiarities and varied languages of the vast Indian market.
By 2012, an Amazon India team of a few dozen engineers occupied an office on the eighth floor of the grandiosely named “World Trade Center,” a curved glass high-rise in northern Bangalore. At first, they were uncertain how to proceed. India’s foreign direct investment rules seemed to preclude them from opening a standard Amazon web store, where the company bought products from manufacturers at wholesale prices and then sold them to online shoppers.
So in typical Amazon fashion, they tried to get clever, introducing a comparison-shopping website in February 2012 called Junglee.com. By scouring the web and listing all the prices and products for sale on other websites, Amazon could start collecting data and earning referral fees without brokering actual transactions and violating the law. But Flipkart recognized the move as a dangerous giant dipping its toe in their waters and declined to let Junglee trawl its site to gather information. After an initial burst of attention, Junglee failed to get any traction.
By 2013, Agarwal and his team had settled on another approach. They would deviate from the company’s playbook and operate Amazon India purely as a third-party marketplace. This would allow outside vendors to sell their wares on the newly christened Amazon.in, with Amazon brokering the transaction and collecting fees but never owning actual inventory. The glaring weakness was that, for the time being, Amazon couldn’t set prices or ensure the availability and quality of the most popular products.
After repeated delays, Amazon.in went live on June 5, 2013. A shaky handheld video of the launch posted on YouTube shows a conference room packed mostly with giddy young Indian men. After flipping the switch at 2 a.m., they burst into riotous applause. “Shop with Confidence” the new site blared.
Within a few weeks, Amazon India expanded from media products like books and DVDs to sell smartphones and digital cameras. Beauty products, kitchen gadgets, and Amazon’s Kindle Fire tablet soon followed. Agarwal wanted to introduce a new category each week—and, like his boss back in Seattle, he liked to set high expectations. “If there was a week when we didn’t launch selection in a category, we used to sit down and say this was a disappointing week,” he later recalled.
Although the new Amazon offshoot was eight thousand miles away from home base, Agarwal had managed to import key elements of Amazon’s culture. He asked movers to transport the door desk he had built for himself as a new employee in 1999—partly, he says, because he felt bad that his family didn’t have enough personal furniture for the shipping company to haul. He introduced Amazon practices like writing six-page narratives and correction of error or “COE” reports, to systematically address problems like delivery delays during the monsoon season. Like Bezos, Agarwal would regularly forward customer emails to his staff with a single question mark—they called them “Amit A. escalations” instead of “Jeff B. escalations”—to highlight problems to be immediately solved, or else.
A few months after launching, in the fall of 2013, Agarwal and his deputies returned to Seattle to present their annual road map to Bezos and the S-team as part of the company’s annual OP1 planning process. Their six-page narrative offered a range of conservative to more aggressive investment options for how to expand in India and catch up with six-year-old Flipkart in sales and other critical benchmarks. It also outlined prospects for an experimental advertising campaign, so that the company could test what resonated with Indian consumers.
By then, Amazon’s China bet was souring, so Bezos did not want to relinquish his shot at what seemed like the world’s next largest prize. In most OP1 sessions, he usually spoke last, not to sway the group with his formidable opinion. But this time, he interjected while Agarwal was still giving his presentation. “You guys are going to fail,” he bluntly told the Indian crew. “I don’t need computer scientists in India. I need cowboys.
“Don’t come to me with a plan that assumes I will only make a certain level of investment,” Bezos continued, according to the recollection of two executives who were there. “Tell me how to win. Then tell me how much it costs.” Another Indian executive at the meeting, Amit Deshpande, says the message was: “Go big and take risks. Make it happen. We have your backs.”
Amit Agarwal, a computer scientist with degrees from IIT and Stanford, was momentarily taken aback. But upon his return to India, he turned that directive into a rallying cry. Bezos’s command became such a part of the core mythology of Amazon India that execs would occasionally dress up in cowboy outfits at all-hands meetings. They scotched their meek OP1 marketing plan and became one of the largest advertisers in India, promoting Amazon.in on the front page of newspapers like the Times of India and in catchy commercials during Indian Premier League cricket games. One of their new team goals, as Amazon India execs described it, was to grow so fast that it practically forced Bezos to come to India.
The next few months were a blur. Members of the team worked all day, every day, traveling often and taking the first plane in the morning and the last at night. When they weren’t flying around the country, they went to China to observe the tactics of Amazon, Alibaba, and JD.com in a similarly competitive environment. “I had a suitcase in my room and a suitcase in the office,” says Vinoth Poovalingam, an operations manager who was setting up Amazon warehouses across India. “A couple of folks used to joke, ‘Dude, we are working in a labor camp.’ ”
Amazon had to operate differently in India. Without critical infrastructure like the multilane highways and credit card networks the company enjoyed in the West, execs had to devise distinctive logistics and payment strategies for the country, like hiring bike messengers and accepting cash on delivery. Where the company usually employed a single code base for the retail website across all of its regions, in India Amazon engineers developed new code and a less memory-intensive smartphone app since customers predominantly accessed the site on their phones and via sluggish wireless networks. To move more nimbly, all departments reported into Agarwal, instead of to their peers in Seattle. “At a fundamental level, we questioned everything and asked, ‘Is this the right thing for India?’ ” said an Amazon India executive.
Agarwal, international chief Diego Piacentini, and Peter Krawiec, Amazon’s corporate development chief, also found a solution to operating a pure third-party marketplace without being able to have a retail arm that could set prices and ensure the availability of products. In mid-2014, with billionaire Narayana Murthy, cofounder of the Indian outsourcing giant Infosys, Amazon formed a joint venture called Prione Business Services, which Amazon would own 49 percent
of. Prione would then run a firm called Cloudtail that would sell popular items like the latest smartphones and consumer electronics. Cloudtail immediately became the largest vendor on Amazon.in, responsible for around 40 percent of sales.
Prione was a transparent hack of India’s somewhat murky foreign direct investment regulations. (“Test the Boundaries of what is allowed by law,” read an internal Amazon slide from the time, Reuters later reported.) The arrangement allowed Amazon to offer customers exclusives on the hottest new smartphones from companies like Samsung and India’s OnePlus. Flipkart, backed by overseas venture capitalists, had paved the way here, setting up its own proxy seller, called “WS Retail,” and offering exclusives on phones from Motorola, Xiaomi, and Huawei. The two companies would play this game for years—duking it out with discounts and exclusives, which the vast assortment of mom-and-pop stores around the country couldn’t possibly hope to match.
By mid-2014, traffic was exceeding both Amazon’s and Flipkart’s most optimistic projections. On July 29, a few months after it acquired fashion rival Myntra, Flipkart announced a fresh infusion of $1 billion in venture capital, valuing the company at $7 billion—more than the total of all other Indian internet startups put together. A day later, Amazon, approaching $1 billion in total sales in the country after just a year of operation, put out a rival press release, loudly announcing a $2 billion capital infusion into Amazon India. The e-commerce opportunity in India was a lucrative echo of the previous battle in China—and this time, Bezos was determined not to lose.
That September, he fulfilled his promise to Agarwal, visiting India to use his budding business fame to advance Amazon’s cause. Flipkart greeted his arrival with an advertising campaign of billboards outside the Bangalore airport and around Amazon’s offices, promoting a new Flipkart-minted online holiday, Big Billion Day, to mark the upcoming Diwali festival.