تئوری آشوب

CHAPTER 6

Chaos Theory

Jeff Bezos liked to be counted on, but after Amazon reached profitability during the ebb of the dot-com bust in 2002, he discovered that he himself would need to count on someone. For while Amazon had quieted its most vociferous critics, Bezos needed help taming the growing chaos inside his company’s walls.

In every significant way, Amazon was becoming a larger and more complicated business. It had 2,100 employees at the end of 1998, and 9,000 at the end of 2004. And after it survived the worst effects of the dot-com crash, it resumed lurching into new categories, like sporting goods, apparel, and jewelry, and new countries, like Japan and China.

Size bred chaos. All companies hit this critical moment, when their internal structures, like a teenager’s old shoes, suddenly don’t fit anymore. But Amazon went through a severe form of this rite of passage. The larger and more ambitious it got, the more complicated it became structurally and the harder it was to keep everyone coordinated and moving quickly. Bezos wanted to execute several strategies simultaneously, but the company’s various interdependent divisions were wasting too much time coordinating with one other.

In the distribution centers, chaos wasn’t an ethereal thing but tangible, reflected in frequent system outages that could shut down facilities for hours and in omnipresent piles of products that sat on the floor, ignored by workers. During the early years of frenzied growth, new product categories had been plopped onto Amazon’s logistics network with little preparation. Employees remember that when the home and kitchen category was introduced in the fall of 1999, kitchen knives would fly down the conveyor chutes, free of protective packaging. Amazon’s internal logistics software didn’t properly account for new categories, so the computers would ask workers whether a new toy entering the warehouse was a hardcover or a paperback book.

Amazon once tried to conquer chaos by synchronizing its employees’ efforts with broad unifying themes like Get Big Fast and Get Our House in Order. That had gotten everyone paddling in the same direction, but now the company had become too big for that kind of transparent sloganeering.

During these years of its awkward adolescence, Bezos refused to slow down, doubling and tripling his bet on the Internet and on his grand vision for a store that sold everything. To guide his company through this transition, he created an unorthodox organizational structure with a peculiar name. And to quell the turmoil in the distribution centers, he started to rely on a young executive named Jeff Wilke, whose cerebral and occasionally impatient management style mirrored his own. “They fed off each other,” says Bruce Jones, a supply-chain vice president. “Bezos wanted to do it and Wilke knew how to do it. It was a hell of a lot of fun, in a very Machiavellian sort of way.”

Jeff Wilke’s job was to fix the mistakes of his predecessor. Jimmy Wright and his cowboy crew from Walmart had designed Amazon’s nationwide logistics network in the late nineties and were the best in the world at building large-scale retail distribution. But in moving quickly to satisfy Bezos’s open-ended goal to store and ship everything, they had created a system that was expensive, unreliable, and hungry for an emergency influx of employees from Seattle at the end of every year. “It was a mess,” says Bruce Jones. “It was pretty much how Walmart did all their distribution centers, which was great if you had to send out five thousand rolls of toilet paper. But it was not well suited to small orders.”

Wilke was from suburban Pittsburgh, the son of an attorney; his parents divorced when he was twelve. He learned he had a talent for mathematics in the sixth grade when to his surprise he placed second in a regional math tournament. When he was fifteen and visiting his grandparents in Las Vegas, he was enthralled by a casino’s video-poker machine. He went home and replicated it on his first-generation personal computer, called the Timex Sinclair 1000 (internal memory: 2 KB). Wilke got straight As throughout school but his guidance counselor told him not to apply to Princeton University because no one from Keystone Oaks High School had ever been admitted to the Ivy League. He applied anyway and got in.

Wilke graduated summa cum laude from Princeton in 1989, three years after Bezos. He earned an MBA and an MS from the Massachusetts Institute of Technology’s engineering/MBA dual-degree program. Called Leaders for Manufacturing (now it’s Leaders for Global Operations), the program is a novel alliance of MIT’s business school, its engineering school, and partner companies, like Boeing, created to address emerging global competition. Mark Mastandrea, an MIT classmate who would follow him to Amazon, says that Wilke “was one of the smartest people I had ever come across. He got to the answers faster than anyone else.”

Wilke began his career at Andersen Consulting and then joined AlliedSignal, the manufacturing giant, which was later acquired by Honeywell. He quickly climbed the ranks to vice president, reporting directly to CEO Larry Bossidy and running the company’s $200-million-a-year pharmaceutical business. In AlliedSignal’s headquarters in Morristown, New Jersey, Wilke was immersed in the corporate dogma of Six Sigma, a manufacturing and management philosophy that seeks to increase efficiency by identifying and eliminating defects.

Back in 1999, Scott Pitasky, an Amazon recruiter who later became the head of human resources at Microsoft, was put in charge of finding a replacement for Jimmy Wright. Pitasky had previously worked with Wilke at AlliedSignal, so he thought of his former colleague after concluding that Amazon needed someone who was smart enough to go toe to toe with Jeff Bezos, who delighted in questioning how everything was done.

Pitasky tracked Wilke down on a business trip in Switzerland and pitched him on taking over the critical distribution network at Amazon. He told Wilke that he would have the chance to build a unique distribution network and define a nascent industry, an opportunity that simply didn’t exist at AlliedSignal. Working quickly, Pitasky convinced then COO Joe Galli, visiting his children at the time on the East Coast, to meet Wilke at a hotel restaurant near Dulles International Airport as soon as Wilke returned to the States.

All of thirty-two years old at the time, with a toothy smile and unfashionable eyeglasses, Wilke did not immediately present the picture of a dynamic leader. “He was not a charismatic communicator,” Galli says. “He was an extremely smart and thoughtful supply-chain expert who relied on fact-based analysis and wanted to zero in and do the right thing.” Over dinner that night, and during a separate trip Galli made to visit Wilke and his wife, Liesl, at their home in New Jersey, the pair bonded. Wilke and Galli were both from Pittsburgh and had similar middle-class roots. Ever the salesman, Galli piqued Wilke’s interest in the massive logistics challenges Amazon faced. Wilke then visited Seattle to interview with Bezos and Joy Covey. He joined the company soon after as vice president and general manager of worldwide operations. After his last conversation with Larry Bossidy, who had just announced his retirement from AlliedSignal, the veteran CEO gave him a hug.

Immediately upon moving to Seattle, Wilke set about filling the ranks of Amazon’s logistics division with scientists and engineers rather than retail-distribution veterans. He wrote down a list of the ten smartest people he knew and hired them all, including Russell Allgor, a supply-chain engineer at Bayer AG. Wilke had attended Princeton with Allgor and had cribbed from his engineering problem sets. Allgor and his supply-chain algorithms team would become Amazon’s secret weapon, devising mathematical answers to questions such as where and when to stock particular products within Amazon’s distribution network and how to most efficiently combine various items in a customer’s order in a single box.1

Wilke recognized that Amazon had a unique problem in its distribution arm: it was extremely difficult for the company to plan ahead from one shipment to the next. The company didn’t store and ship a predictable number or type of orders. A customer might order one book, a DVD, some tools—perhaps gift-wrapped, perhaps not—and that exact combination might never again be repeated. There were an infinite number of permutations. “We were essentially assembling and fulfilling customer orders. The factory physics were a lot closer to manufacturing and assembly than they were to retail,” Wilke says. So in one of his first moves, Wilke renamed Amazon’s shipping facilities to more accurately represent what was happening there. They were no longer to be called warehouses (the original name) or distribution centers (Jimmy Wright’s name); forever after, they would be known as fulfillment centers, or FCs.

Before Wilke joined Amazon, the general managers of the fulfillment centers often improvised their strategies, talking on the telephone each morning and gauging which facility was fully operational or had excess capacity, then passing off orders to one another based on those snap judgments. Wilke’s algorithms seamlessly matched demand to the correct FC, leveling out backlogs and obviating the need for the morning phone call. He then applied the process-driven doctrine of Six Sigma that he’d learned at AlliedSignal and mixed it with Toyota’s lean manufacturing philosophy, which requires a company to rationalize every expense in terms of the value it creates for customers and allows workers (now called associates) to pull a red cord and stop all production on the floor if they find a defect (the manufacturing term for the system is andon).

In his first two years, Wilke and his team devised dozens of metrics, and he ordered his general managers to track them carefully, including how many shipments each FC received, how many orders were shipped out, and the per-unit cost of packing and shipping each item. He got rid of the older, sometimes frivolous names for mistakes—Amazon’s term to describe the delivery of the wrong product to a customer was switcheroo—and substituted more serious names. And he instilled some basic discipline in the FCs. “When I joined, I didn’t find time clocks,” Wilke says. “People came in when they felt like it in the morning and then went home when the work was done and the last truck was loaded. It wasn’t the kind of rigor I thought would scale.” Wilke promised Bezos that he would reliably generate cost savings each year just by reducing defects and increasing productivity.

Wilke elevated the visibility of his FC managers within Amazon. He brought them to Seattle as often as possible and highlighted the urgency of their technical issues. During the holiday season, in what remains today his personal signature, Wilke wore a flannel shirt every day as a gesture of solidarity with his blue-collar comrades in the field. Wilke “recognized that a general manager was a difficult job and he made you feel you were in a lifelong club,” says Bert Wegner, who ran the Fernley FC in those years.

Wilke had another tool at his disposal: like Bezos, he had an occasionally volcanic temper. Back in the fall of 2000, the software systems in Amazon’s FCs were still incapable of precisely tracking inventory and shipments. So that holiday, Wilke’s second one at the company, during the annual race to Christmas that the company internally referred to as the big push, Wilke started a series of daily conference calls with his general managers in the United States and Europe. He told his general managers that on each call, he wanted to know the facts on the ground: how many orders had shipped, how many had not, whether there was a backlog, and, if so, why. As that holiday season ramped up, Wilke also demanded that his managers be prepared to tell him “what was in their yard”—the exact number and contents of the trucks waiting outside the FCs to unload products and ferry orders to the post office or UPS.

One recurring trouble spot that year was the fulfillment center in McDonough, Georgia, a working-class city thirty miles south of Atlanta. In the heat of the tumultuous holiday season, McDonough—the source of the infamous Jigglypuff crisis of 1999—was regularly falling behind schedule. Its general manager, a once and future Walmart executive named Bob Duron, was already skating on thin ice when Wilke surveyed his managers on a conference call and asked them what they had in their yards. When he got to McDonough, Duron apparently hadn’t gotten the message and said: “Hold on a second, Jeff, I can see them outside my window.” Then he leaned back in his chair and started counting aloud on the phone. “I’ve got one, two, three, four…”

Wilke went off like a bomb. He was calling that day from his home office on Mercer Island, and he started screaming—an oral assault of such intensity and vulgarity that the handsets of the general managers on the call shrieked with feedback. And then, just as abruptly as the outburst began, there was quiet. Wilke had seemingly disappeared.

No one said anything for thirty seconds. Finally Arthur Valdez, the general manager in Campbellsville, said quietly, “I think he ate the phone.”

There were various interpretations of what actually happened. Some claimed that in his rage, Wilke had inadvertently yanked the phone cord out of the wall. Others speculated that he had thrown the receiver across the room in his fury. A decade later, over lunch at an Italian brasserie near Amazon’s offices, Wilke explains that he had actually still been on the line but was simply so angry that he could no longer speak. “We were just struggling to make it work on a whole host of different levels in McDonough,” he says. “We were struggling to recruit the right leaders, and struggling to get enough people to work there.”

That spring, with Amazon sprinting toward its profitability goal, Wilke shut down McDonough and fired four hundred and fifty full-time employees. Closing the facility wouldn’t solve Amazon’s problems; in fact, the reduction in capacity put even more pressure on Amazon’s other fulfillment centers. The company was already running at capacity over the holidays and sales were growing at more than 20 percent a year. Now Amazon had no choice but to master the complexity of its own systems and get more out of the investments it had already made.

Wilke had burned a boat in mid-voyage, and for the Amazon armada, there was no turning back. Along the way, he was exhibiting a style—leadership by example, augmented with a healthy dose of impatience—that was positively Bezosian in character. Perhaps not coincidentally, Wilke was promoted to senior vice president a little over a year after joining Amazon. Jeff Bezos had found his chief ally in the war against chaos.

At a management offsite in the late 1990s, a team of well-intentioned junior executives stood up before the company’s top brass and gave a presentation on a problem indigenous to all large organizations: the difficulty of coordinating far-flung divisions. The junior executives recommended a variety of different techniques to foster cross-group dialogue and afterward seemed proud of their own ingenuity. Then Jeff Bezos, his face red and the blood vessel in his forehead pulsing, spoke up.

“I understand what you’re saying, but you are completely wrong,” he said. “Communication is a sign of dysfunction. It means people aren’t working together in a close, organic way. We should be trying to figure out a way for teams to communicate less with each other, not more.”

That confrontation was widely remembered. “Jeff has these aha moments,” says David Risher. “All the blood in his entire body goes to his face. He’s incredibly passionate. If he was a table pounder, he would be pounding the table.”

At that meeting and in public speeches afterward, Bezos vowed to run Amazon with an emphasis on decentralization and independent decision-making. “A hierarchy isn’t responsive enough to change,” he said. “I’m still trying to get people to do occasionally what I ask. And if I was successful, maybe we wouldn’t have the right kind of company.”2

Bezos’s counterintuitive point was that coordination among employees wasted time, and that the people closest to problems were usually in the best position to solve them. That would come to represent something akin to the conventional wisdom in the high-tech industry over the next decade. The companies that embraced this philosophy, like Google, Amazon, and, later, Facebook, were in part drawing lessons from theories about lean and agile software development. In the seminal high-tech book The Mythical Man-Month, IBM veteran and computer science professor Frederick Brooks argued that adding manpower to complex software projects actually delayed progress. One reason was that the time and money spent on communication increased in proportion to the number of people on a project.

Bezos and other startup founders were reacting to lessons from previous technology giants. Microsoft took a top-down management approach with layers of middle managers, a system that ended up slowing decisions and stifling innovation. Looking at the muffled and unhappy hierarchy of the software giant across Lake Washington, Amazon executives saw a neon sign warning them exactly what to avoid.

The drive to cut costs also forced Bezos to eliminate any emerging layers of middle management from his company. After the stock market crash in 2000, Amazon went through two rounds of layoffs. But Bezos didn’t want to stop recruiting altogether; he just wanted to be more efficient. So he framed the kind of employees he wanted in simple terms. All new hires had to directly improve the outcome of the company. He wanted doers—engineers, developers, perhaps merchandise buyers, but not managers. “We didn’t want to be a monolithic army of program managers, à la Microsoft. We wanted independent teams to be entrepreneurial,” says Neil Roseman. Or, as Roseman also put it: “Autonomous working units are good. Things to manage working units are bad.”

But as was often the case, no one could anticipate just how far Bezos would venture into these organizational theories in his quest to distill them down to their core ideas. In early 2002, as part of a new personal ritual, he took time after the holidays to think and read. (In this respect, Microsoft’s Bill Gates, who also took such annual think weeks, served as a positive example.) Returning to the company after a few weeks, Bezos presented his next big idea to the S Team in the basement of his Medina, Washington, home.

The entire company, he said, would restructure itself around what he called “two-pizza teams.” Employees would be organized into autonomous groups of fewer than ten people—small enough that, when working late, the team members could be fed with two pizza pies. These teams would be independently set loose on Amazon’s biggest problems. They would likely compete with one another for resources and sometimes duplicate their efforts, replicating the Darwinian realities of surviving in nature. Freed from the constraints of intracompany communication, Bezos hoped, these loosely coupled teams could move faster and get features to customers quicker.

There were some head-scratching aspects to Bezos’s two-pizza-team concept. Each group was required to propose its own “fitness function”—a linear equation that it could use to measure its own impact without ambiguity. For example, a two-pizza team in charge of sending advertising e-mails to customers might choose for its fitness function the rate at which these messages were opened multiplied by the average order size those e-mails generated. A group writing software code for the fulfillment centers might home in on decreasing the cost of shipping each type of product and reducing the time that elapsed between a customer’s making a purchase and the item leaving the FC in a truck. Bezos wanted to personally approve each equation and track the results over time. It would be his way of guiding a team’s evolution.

Bezos was applying a kind of chaos theory to management, acknowledging the complexity of his organization by breaking it down to its most basic parts in the hopes that surprising results might emerge. That, at least, was the high-minded goal; the end result was somewhat disappointing. The two-pizza-team concept took root first in engineering, where it was backed by Rick Dalzell, and over the course of several years, it was somewhat inconsistently applied through the rest of the company. There was just no reason to organize some departments, such as legal and finance, in this way.

The idea of fitness functions in particular appeared to clash with some fundamental aspects of human nature—it’s uncomfortable to have to set the framework for your own evaluation when you might be judged harshly by the end result. Asking groups to define their own fitness functions was a little like asking a condemned man to decide how he’d like to be executed. Teams ended up spending too much time worrying over their formulas and making them ever more complex and abstract. “Being a two-pizza team was not exactly liberating,” says Kim Rachmeler. “It was actually kind of a pain in the ass. It did not help you get your job done and consequently the vast majority of engineers and teams flipped the bit on it.”

A year into Jeff Wilke’s tenure at Amazon, he called a former teacher of his, Stephen Graves, a professor of management science at MIT, and asked for help. Amazon operated an e-commerce distribution network of unrivaled scale but the company was still struggling to run it efficiently. Its seven fulfillment centers around the world were expensive, their output inconsistent. Bezos wanted the Amazon website to be able to tell customers precisely when their packages would be delivered. For example, a college student ordering a crucial book for a final exam should know that the book would be delivered the following Monday. But the fulfillment centers were not yet reliable enough to make that kind of specific prediction.

Wilke asked Graves if he might meet with Wilke and his colleagues later that month to take a fresh look at their problems. Bezos and Wilke were asking themselves a fundamental question that seems surprising today: Should Amazon even be in the business of storing and distributing its products? The alternative was to shift to the model used by rivals like Buy.com, which took orders online but had products drop-shipped from manufacturers and distributors like Ingram.

That St. Patrick’s Day, some of Amazon’s biggest brains descended on a drab meeting room at the Fernley, Nevada, fulfillment center. Jeff Bezos and Brewster Kahle, a supercomputer engineer and founder of Alexa Internet, a data-mining company Amazon had acquired, made the two-hour flight from Seattle on Bezos’s newly purchased private plane, a Dassault Falcon 900 EX. Stephen Graves flew from Massachusetts to Reno and then drove the dreary thirty-four miles through the desert to Fernley. A few other Amazon engineers were there, as was the facility’s senior manager at the time, Bert Wegner. In the morning, the group toured the fulfillment center and listened to a presentation by one of the company’s primary contractors, who listed the benefits of additional equipment and software that he could sell them, reflecting the same traditional thinking about distribution that wasn’t working in the first place. They then dismissed the surprised contractor for the day and spent the afternoon filling up whiteboards and tackling the question of how everything at the FC might be improved. For lunch, they brought in McDonald’s and snacked from the building’s vending machines.

For Wegner, the questions being asked that day carried personal resonance. “We had a key decision to make,” he says. “Was distribution a commodity or was it a core competency? If it’s a commodity, why invest in it? And when we grow, do we continue to do it on our own or do we outsource it?” If Amazon chose to outsource it, Wegner might be out of a job. “I basically saw my own career flash before my eyes,” he says.

Amazon’s problem boiled down to something called, in the esoteric lexicon of manufacturing, batches. The equipment in Amazon’s FCs had originally been acquired by Jimmy Wright, and, like the system in Walmart’s distribution centers, was designed by its manufacturers to operate in waves—moving from minimum capacity to maximum and then back again. At the start of a wave, a group of workers called pickers fanned out across the stacks of products, each in his or her own zone, to retrieve the items ordered by customers. At the time, Amazon used the common pick-to-light system. Various lights on the aisles and on individual shelves guided pickers to the right products, which they would then deposit into their totes—a cart of the picks from that wave. They then delivered their totes to conveyor belts that fed into the giant sorting machines, which rearranged products into customer orders and sent them off on a new set of conveyor belts to be packed and shipped.

The software required pickers to work individually, but, naturally, some took longer than others, which led to problems. For example, if ninety-nine pickers completed their batches within forty-five minutes but the one hundredth picker took an additional half an hour, those ninety-nine pickers had to sit idly and wait. Only when that final tote cleared the chute did the system come fully alive again, with a thunderous roar that rolled through the fulfillment center and indicated that it was again ready to start operating at peak capacity.

Everything in the fulfillment center happened in this episodic manner. For a company trying to maximize its capacity during the big push each holiday season, that was a huge problem. Wilke subscribed to the principles laid out in a seminal book about constraints in manufacturing, Eliyahu M. Goldratt’s The Goal, published in 1984. The book, cloaked in the guise of an entertaining novel, instructs manufacturers to focus on maximizing the efficiency of their biggest bottlenecks. For Amazon, that was the Crisplant sorting machines, where the products all ended up, but picking in batches limited how fast the sorters could be fed. As a result, the machines were operating at full capacity only during the brief few minutes at the peak of the batch. Wilke’s group had experimented with trying to run overlapping waves, but that tended to overload the Crisplant sorters and, in the dramatic terminology of the general managers, “blow up the building.” It would take hours to clean up that mess and get everything back on track.

In the meeting that day at Fernley, the executives and engineers questioned the prevailing orthodoxies of retail distribution. In the late afternoon, everyone headed back onto the facility floor and watched orders move haltingly through the facility. “I didn’t know Jeff Bezos but I just remember being blown away by the fact that he was there with his sleeves rolled up, climbing around the conveyors with all of us,” says Stephen Graves, the MIT professor. “We were thinking critically and throwing around some crazy ideas of how we can do this better.”

At the end of the day, Bezos, Wilke, and their colleagues reached a conclusion: the equipment and software from third-party vendors simply wasn’t designed for the task at hand. To escape from batches and move toward a continuous and predictable flow of orders through the facility, Amazon would have to rewrite all the software code. Instead of exiting the business of distribution, they had to reinvest in it.

Over the next few years, “one by one, we unplugged our vendors’ modems and we watched as their jaws hit the floor,” says Wegner. “They couldn’t believe we were engineering our own solutions.” When Amazon later opened small facilities in places like Seattle and Las Vegas to handle easily packable items and larger fulfillment centers in Indianapolis, Phoenix, and elsewhere, it would go even further, dispensing with the pick-to-light systems and big Crisplant sorting machines altogether and instead employing a less automated approach that favored invisible algorithms. Employees would bring their totes from the shelves right to the packing stations, their movements carefully coordinated by software. Slowly, Amazon would vanquish wave-based picking, elicit more productivity from its workers, and improve the accuracy and reliability of its fulfillment centers.

Wilke’s gradual success in making the logistics network more efficient would offer Amazon innumerable advantages in the years ahead. Tightly controlling distribution allowed the company to make specific promises to customers on when they could expect their purchases to arrive. Amazon’s operating all of its own technology, from the supply chain to the website, allowed Russell Allgor and his engineers to create algorithms that modeled countless scenarios for each order so systems could pick the one that would yield the quickest and cheapest delivery. Millions of those decisions could be made every hour, helping Amazon reduce its costs—and thus lower prices and increase volume of sales. The challenge was getting good enough to do this well.

“No matter how hard it is, the consolidation of products within fulfillment centers pays for the inventory and for pieces of the overhead,” says Jeff Wilke, who claims that he never worried that Bezos would abandon the FC model at the Fernley meeting. “The principles and math were on our side, and I realized early on that this was a company where you can carry the day when you have the principles and math on your side, and you are patient and tenacious.”

Whenever Jeff Bezos roamed a fulfillment center or his own Seattle headquarters, he looked for defects—flaws in the company’s systems or even its corporate culture. On an otherwise regular weekday morning in 2003, for example, Bezos walked into an Amazon conference room and was taken aback. Mounted on the wall, in a corner of the room, was a newly installed television meant for video presentations to employees. A TV in a conference room did not by itself seem controversial, yet Bezos was not pleased.

The installations, which he had not known about or authorized, represented to him both a clumsy attempt at interoffice communication and an extravagant expenditure. “How can anything good be communicated in this way,” he complained.

Bezos had all the new televisions in Amazon’s conference rooms immediately removed. But according to Matt Williams, a longtime Amazon manager, Bezos deliberately kept the metal mounts hanging in the conference rooms for many years, even some that were so low on the wall that employees were likely to stand up and hit them. Like a warlord leaving the decapitated heads of his enemies on stakes outside his village walls, he was using the mounts as a symbol, and as an admonition to employees about how not to behave.

The television episode was the foundation of another official award at Amazon, this one presented to an employee who identified an activity that was bureaucratic and wasteful. The suddenly superfluous televisions were given as the prize. When the supply ran out, that commendation morphed into the Door-Desk award, given to an employee who came up with “a well-built idea that helps us to deliver lower prices to customers”—the prize was a door-desk ornament. Bezos was once again looking for ways to reinforce his values within the company.

Around the same time he was ripping televisions off the walls, Bezos made two significant changes to the corporate culture. As part of his ongoing quest for a better allocation of his own time, he decreed that he would no longer have one-on-one meetings with his subordinates. These meetings tended to be filled with trivial updates and political distractions, rather than problem solving and brainstorming. Even today, Bezos rarely meets alone with an individual colleague.

The other change was also peculiar and perhaps unique in corporate history. Up until that time, Amazon employees had been using Microsoft’s PowerPoint and Excel spreadsheet software to present their ideas in meetings. Bezos believed that method concealed lazy thinking. “PowerPoint is a very imprecise communication mechanism,” says Jeff Holden, Bezos’s former D. E. Shaw colleague, who by that point had joined the S Team. “It is fantastically easy to hide between bullet points. You are never forced to express your thoughts completely.”

Bezos announced that employees could no longer use such corporate crutches and would have to write their presentations in prose, in what he called narratives. The S Team debated with him over the wisdom of scrapping PowerPoint but Bezos insisted. He wanted people thinking deeply and taking the time to express their thoughts cogently. “I don’t want this place to become a country club,” he was fond of saying as he pushed employees harder. “What we do is hard. This is not where people go to retire.”

There was a period of grumbling adjustment. Meetings no longer started with someone standing up and commanding the floor as they had previously at Amazon and everywhere else throughout the corporate land. Instead, the narratives were passed out and everyone sat quietly reading the document for fifteen minutes—or longer. At the beginning, there was no page limit, an omission that Diego Piacentini recalled as “painful” and that led to several weeks of employees churning out papers as long as sixty pages. Quickly there was a supplemental decree: a six-page limit on narratives, with additional room for footnotes.

Not everyone embraced the new format. Many employees felt the system was rigged to reward good writers but not necessarily efficient operators or innovative thinkers. Engineers in particular were unhappy to suddenly find themselves crafting essays as if they had been hurled back through time into ninth-grade English. “Putting everything into a narrative ended up sort of being like describing a spreadsheet,” says Lyn Blake, a vice president in charge of the company’s relationships with manufacturers at the time. Blake herself suspected the whole thing was a phase. (It wasn’t.)

Bezos refined the formula even further. Every time a new feature or product was proposed, he decreed that the narrative should take the shape of a mock press release. The goal was to get employees to distill a pitch into its purest essence, to start from something the customer might see—the public announcement—and work backward. Bezos didn’t believe anyone could make a good decision about a feature or a product without knowing precisely how it would be communicated to the world—and what the hallowed customer would make of it.

Steve Jobs was known for the clarity of his insights about what customers wanted, but he was also known for his volatility with coworkers. Apple’s founder reportedly fired employees in the elevator and screamed at underperforming executives. Perhaps there is something endemic in the fast-paced technology business that causes this behavior, because such intensity is not exactly rare among its CEOs. Bill Gates used to throw epic tantrums. Steve Ballmer, his successor at Microsoft, had a propensity for throwing chairs. Andy Grove, the longtime CEO of Intel, was known to be so harsh and intimidating that a subordinate once fainted during a performance review.

Jeff Bezos fit comfortably into this mold. His manic drive and boldness trumped other conventional leadership ideals, such as building consensus and promoting civility. While he was charming and capable of great humor in public, in private, Bezos could bite an employee’s head right off.

Bezos was prone to melodramatic temper tantrums that some Amazon employees called, privately, nutters. A colleague failing to meet Bezos’s exacting standards would predictably set off a nutter. If an employee did not have the right answers, or tried to bluff the right answer, or took credit for someone else’s work, or exhibited a whiff of internal politics, or showed any kind of uncertainty or frailty in the heat of battle, the vessel in Bezos’s forehead popped out and his filter fell away. He was capable of both hyperbole and cruelty in these moments, and over the years he delivered some devastating rebukes to employees. Among his greatest hits, collected and relayed by Amazon veterans:

“If that’s our plan, I don’t like our plan.”

“I’m sorry, did I take my stupid pills today?”

“Do I need to go down and get the certificate that says I’m CEO of the company to get you to stop challenging me on this?”

“Are you trying to take credit for something you had nothing to do with?”

“Are you lazy or just incompetent?”

“I trust you to run world-class operations and this is another example of how you are letting me down.”

“If I hear that idea again, I’m gonna have to kill myself.”

“Does it surprise you that you don’t know the answer to that question?”

“Why are you ruining my life?”

[After someone presented a proposal.] “We need to apply some human intelligence to this problem.”

[After reviewing the annual plan from the supply-chain team.] “I guess supply chain isn’t doing anything interesting next year.”

[After reading a narrative.] “This document was clearly written by the B team. Can someone get me the A team document? I don’t want to waste my time with the B team document.”

Some Amazon employees currently advance the theory that Bezos, like Steve Jobs, Bill Gates, and Larry Ellison, lacks a certain degree of empathy and that as a result he treats workers like expendable resources without taking into account their contributions to the company. That in turn allows him to coldly allocate capital and manpower and make hyperrational business decisions while another executive might let emotion and personal relationships intrude. But they also acknowledge that Bezos is primarily consumed with improving the company’s performance and customer service, and that personnel issues are secondary. “This is not somebody who takes pleasure at tearing someone a new asshole. He is not that kind of person,” says Kim Rachmeler. “Jeff doesn’t tolerate stupidity, even accidental stupidity.”

Right or wrong, Bezos’s behavior was often easier to accept because he was so frequently on target with his criticisms, to the amazement and often irritation of employees. Bruce Jones, the former Amazon vice president, describes leading a five-engineer team working to create algorithms to optimize pickers’ movements in the fulfillment centers while the company was trying to solve the problem of batches. The group spent nine months on the task, then presented their work to Bezos and the S Team. “We had beautiful documents and everyone was really prepared,” Jones says. Bezos read the paper, said, “You’re all wrong,” stood up, and started writing on the whiteboard.

“He had no background in control theory, no background in operating systems,” Jones says. “He only had minimum experience in the distribution centers and never spent weeks and months out on the line.” But Bezos laid out his argument on the whiteboard and “every stinking thing he put down was correct and true,” Jones says. “It would be easier to stomach if we could prove he was wrong but we couldn’t. That was a typical interaction with Jeff. He had this unbelievable ability to be incredibly intelligent about things he had nothing to do with, and he was totally ruthless about communicating it.”

In 2002 Amazon changed the way it accounted for inventory, from a system called last-in first-out, or LIFO, to one called first-in first-out, or FIFO. The change allowed Amazon to better distinguish between its products and the products that were owned and stored in the FCs by partners like Toys “R” Us and Target.

Jones’s supply-chain team was in charge of this complicated effort, and its software, ridden by bugs, created a few difficult days during which Amazon’s systems were unable to formally recognize any revenue. On the third day, Jones was updating the S Team on the transition when Bezos tore into him. “He called me a ‘complete fucking idiot’ and said he had no idea why he hired idiots like me at the company, and said, ‘I need you to clean up your organization,’ ” Jones recalls, years later. “It was brutal. I almost quit. I was a resource of his that failed. An hour later he would have been the same guy as always and it would have been different. He can compartmentalize like no one I’ve ever seen.”

As Jones left the FIFO meeting, Jeff Wilke’s administrative assistant approached him with a telephone. Wilke was calling from vacation in Arizona, where he’d already heard about the confrontation. “He said, ‘Bruce, I want you to know I’m behind you one hundred percent. I’m completely confident in you. If you need anything, I’m on a golf course, and I’ll do whatever I can to help.’ ”

Jeff Wilke wasn’t always the softer counterbalance to Bezos. The pair visited each fulfillment center every fall in an annual ritual they called the whistle-stop tour. They spent a week on the road, one day in each FC, and used their commanding presence to focus attention on eliminating errors and improving processes. General managers, their palms sweaty and their pulses elevated, would present to the pair, laying out their emergency scenarios and the ways in which they had once again managed to wrangle thousands of temporary workers for the holidays. Wilke and Bezos dug into the details, asking their inhumanly prescient questions. It was both inspiring and terrifying. “Those guys could be brutal,” says Mark Mastandrea. “You had to be comfortable saying, ‘I don’t know; I’ll get back to you in a couple of hours,’ and then doing it. You could not ever bullshit or make stuff up. That would be the end.”

T. E. Mullane worked in Amazon’s logistics network for years, helping to open and manage new fulfillment centers. He opened a new FC in Chambersburg, Pennsylvania, and hosted Wilke on his first visit to the facility. Wilke, Mullane says, started the tour by quietly walking the inside perimeter of the building. In a corner of the building near the inbound docks, he encountered a disorganized pile of nonconveyable products—too heavy to move on the conveyor belts. For one reason or another, workers had not been able to match the merchandise with order slips and so had left them in a heap.

After the walk, Wilke looked at Mullane and initiated a typical exchange. “T.E., do you know why I walked the perimeter? Tell me why.”

“To look for errors,” Mullane said.

“So why do the operators leave piles?”

“Because the process isn’t correct right now. It’s not precise and predictable.”

“Right. So you are going to take care of this?”

“Yes.”

The whistle-stop visits usually occurred midway through the fourth quarter of the year, just as the holiday season was ramping up but before the big shopping days known as Black Friday and Cyber Monday. During the big push itself, Wilke would return to Seattle but stay in touch with subordinates via his grueling daily conference calls.

The pressure during the holidays could get so intense that Wilke instituted a new ritual as a form of therapeutic release: primal screams. When a logistics executive or his team accomplished something significant, Wilke would allow the person or even the entire group to lean back, close their eyes, and yell into the phone at the tops of their lungs. “It was clearly a great release but the first time it almost blew my phone speaker,” Wilke says.

After the big push was over and the last box had shipped, typically on December 23, “you got to enjoy Christmas Day more than anybody else on the planet, because you had worked so hard to get there,” says Bert Wegner. Then planning would begin all over again.

In 2002, Jeff Wilke led the first significant effort to use Amazon’s now impressive size to exact concessions from a major business partner: the United Parcel Service. That year, Amazon’s contract with UPS was up for renewal, and the package-delivery giant, embroiled in a separate standoff with the Teamsters Union, did not appear to be in the mood to grant more-favorable terms to the online upstart. Amazon wasn’t using Federal Express in any significant way at the time, and the primary alternative to UPS, the federally managed U.S. Postal Service, was not permitted to negotiate its rates. Amazon, it seemed, had no leverage.

But early that year, sensing an opportunity, Wilke approached Bruce Jones in Operations and asked him to begin cultivating FedEx. Over the course of six months, Jones and a team traveled frequently to FedEx’s headquarters in Memphis, integrating their systems and quietly ratcheting up the volume of packages. Amazon also increased its shipment injections with the U.S. Postal Service: company employees drove Amazon’s trucks to the post office and inserted packages directly into the flow of federal mail.

Wilke started his negotiations with UPS that summer in Louisville, ahead of a September 1 contract deadline. When UPS was predictably obstinate about deviating from its standard rate card, Wilke threatened to walk. UPS officials thought he was bluffing. Wilke called Jones in Seattle and said, “Bruce, turn them off.”

“In twelve hours, they went from millions of pieces [from Amazon] a day to a couple a day,” says Jones, who flew to Fernley to watch the fallout. The standoff lasted seventy-two hours and went unnoticed by customers and other outsiders. In Fernley, UPS representatives told Jones they knew Amazon couldn’t keep it up and predicted that FedEx would be overwhelmed. They were likely right. But before it came to that, UPS execs caved and gave Amazon discounted rates.

“Yes, we could have operated mostly without them,” Wilke says. “But it would have been very hard, very painful. They knew that. I didn’t want to leave them, I just wanted a fair price.” In the end, he got one, bringing home one of Amazon’s first bulk discounts and teaching the company an enduring lesson about the power of scale and the reality of Darwinian survival in the world of big business.

In 2003, Jeff Bezos came up with yet another way to frame his concept of Amazon. This time, it was for a group of buyers who were leading the company’s charge into the new hard-lines categories, a group of products that included hardware, sporting goods, and electronics. Amazon, Bezos said, was the unstore.

At the time, Bezos had selected jewelry as the company’s next big opportunity. It was a tempting target: the products were small, the prices were high, and shipping was relatively cheap. He tapped retail managers Eric Broussard and Randy Miller to lead the effort. As usual, the executives Bezos chose to head the product’s sales had no prior experience selling that product.

Though it seemed alluring, selling jewelry posed some challenges. Expensive baubles were difficult to display in full detail online; also, the products were valuable and tempting to pilfering workers in the company’s fulfillment centers. Another issue arose with pricing: The jewelry industry had a simplistic pricing model with generous margins. Retail markup was significant; stores doubled the wholesale cost (a practice known as keystone pricing) or even tripled it (known as triple-keystone pricing). Jewelry manufacturers and retailers clung tightly to that custom, which didn’t fit well with Bezos’s newly adamant resolve to offer the lowest prices anywhere.

The Amazon jewelry executives decided on an approach similar to the one the company had recently used for its cautious first foray into apparel. They would let other, more experienced retailers sell everything on the site via Amazon’s Marketplace, and Amazon would take a commission. Meanwhile, the company could watch and learn. “That was something we did quite well,” says Randy Miller. “If you don’t know anything about the business, launch it through the Marketplace, bring retailers in, watch what they do and what they sell, understand it, and then get into it.”

Bezos seemed amenable to that plan, at least at first. And then one day, in a meeting with the S Team and the hard-lines group, something set him off. They were discussing the margins in the jewelry business, and one of Randy Miller’s colleagues mentioned how the jewelry industry conducted business in the “traditional way.” “You’re not thinking about this right,” Bezos said, and he excused himself to get something from his office. He was gone a few minutes, then returned with a stack of photocopied documents and handed a page to everyone in the meeting. It had only one paragraph, about ten sentences long. It began with the words We are the “Unstore.”

The document, as Miller and other executives who were there remember it, defined how Bezos saw his own company—and explains why, even years later, so many businesses are unsettled by Amazon’s entrance into their markets.

Being an unstore meant, in Bezos’s view, that Amazon was not bound by the traditional rules of retail. It had limitless shelf space and personalized itself for every customer. It allowed negative reviews in addition to positive ones, and it placed used products directly next to new ones so that customers could make informed choices. In Bezos’s eyes, Amazon offered both everyday low prices and great customer service. It was Walmart and Nordstrom’s.

Being an unstore also meant that Amazon had to concern itself only with what was best for the customer. The conventions of the jewelry business allowed routine 100 or 200 percent markups, but, well, that just didn’t apply to Amazon.

In that meeting, Bezos decreed that Amazon was not in retail, and therefore did not have to kowtow to retail. He suggested that Amazon could ignore the conventions of pricing in the jewelry business and envisioned customers buying a bracelet on the site for $1,200 and then going to get an appraisal and finding out from the local jeweler that the item was actually worth $2,000. “I know you’re retailers and I hired you because you are retailers,” Bezos said. “But I want you to understand that from this day forward, you are not bound by the old rules.”

Amazon started selling jewelry in the spring of 2004; two-thirds of the selection came from its Marketplace and the other third came directly from Amazon. For months, Bezos was consumed by the design of the elegant wooden jewelry box that Amazon would use. “The box was everything to him,” says Randy Miller. “He wanted it to be as iconic as Tiffany’s.”

Amazon contracted with celebrity socialite Paris Hilton to sell her jewelry designs exclusively on the site, and the company spent considerable resources creating a tool to let customers design their own rings on the website. Amazon’s new staff jewelers would then craft the rings over an open flame on the mezzanine of the Lexington, Kentucky, fulfillment center. Additionally, Amazon introduced a feature called Diamond Search that let customers look for individual stones based on carat, shape, and color. And in a draconian tactic that further exposed his competitive streak, Bezos instructed Amazon’s communication staff to time public announcements in the jewelry category to coincide with the quarterly reports of Seattle-based rival Blue Nile, the leader in online jewelry sales.

Selling jewelry became a modestly profitable business for Amazon, according to employees who worked on the category, but the seeds clearly did not grow into the trees that Bezos had envisioned. Although Amazon’s watch business became robust, customers still wanted to go into actual stores to pick out engagement rings. After a while, the ring-designing tool and Diamond Search disappeared from the site. Amazon’s attention wandered to new battlefields, such as shoes and apparel. Employees who passed through jewelry later described a grueling experience, with shifting goals, rotating bosses, and endless disputes with suppliers who disliked Amazon’s pricing. Being an unstore was evidently not as easy as Bezos had thought. Amazon executives in the hard-lines business during these years had a running joke: “Why do you think they call them hard?”

As the hard-lines teams were bringing Amazon into new categories, with varying degrees of success, Jeff Wilke and his group had nearly completed their job morphing Amazon’s fulfillment process from a network of haphazardly constructed facilities into something that could more accurately be considered a system of polynomial equations. A customer might place an order for a half a dozen products, and the company’s software would quickly examine factors like the address of the customer, the location of the merchandise in the FCs, and the cutoff times for shipping at the various facilities around the country. Then it would take all those variables and calculate both the fastest and the least expensive way to ship the items.

The complete software rewrite of the logistics network was having its desired effect. Cost per unit (the overall expense of fulfilling the order of a particular item) fell, while ship times (how quickly merchandise ordered on the website was loaded onto a truck) shortened. A year after the Fernley meeting, the click-to-ship time for most items in the company’s FCs was as minimal as four hours, down from the three days it had taken when Wilke first started at the company. The standard for the rest of the e-commerce industry at the time was twelve hours.

Amazon’s ability to ship products efficiently and offer precise delivery times to customers gave the company a competitive edge over its rivals, particularly eBay, which avoided this part of the business altogether. Fulfillment was a lever that Bezos had invested in, and he started using it to guide strategy.

By 2002, the company was offering customers the option, for an extra fee, of overnight, two-day, or three-day shipping. Wilke’s team called these fast-track or fast-lane orders and built a separate process around them. On the floor of the FCs, those items were accelerated through the Crisplant sorters and were the first to be delivered to the packers and the trucks waiting in the yard. The company refined this ability gradually, pushing the cutoff time for next-day delivery to forty-five minutes before the last trucks left its fulfillment centers. Expedited shipping was almost prohibitively expensive, for customers and for Amazon, but the website’s having the capability was to pay huge strategic dividends.

In 2004, an Amazon engineer named Charlie Ward used an employee-suggestion program called the Idea Tool to make a proposal. Super Saver Shipping, he reasoned, catered to price-conscious customers whose needs were not time sensitive—they were like the airline travelers who paid a lower rate because they stayed at their destinations over a Saturday night. Their orders got placed on the trucks whenever there was room for them, reducing the overall shipping cost. Why not create a service for the opposite type of customer, Ward suggested, a speedy shipping club for consumers whose needs were time sensitive and who weren’t price conscious? He suggested that it could work like a music club, with a monthly charge.

That fall, employees showed enough enthusiasm for Ward’s proposal that it came to the attention of Bezos. Immediately enchanted by the idea, Bezos asked a group that included Vijay Ravindran, the director of Amazon’s ordering systems, to meet him on a Saturday in the boathouse behind his home in Medina. Bezos conveyed a sense of urgency as he began the meeting, saying that the shipping club was now top priority. “This is a big idea,” he told the gathered engineers. He asked Ravindran and Jeff Holden to put together a SWAT team of a dozen of their best people and told them he wanted the program ready by the next earnings announcement, in February—just weeks away.

Bezos met with the group, which included Charlie Ward and Dorothy Nicholls, who would later go on to become a longtime Kindle executive, weekly over the next two months. They devised the two-day shipping offer, exploiting the ability of Wilke’s group to accelerate the handling of individual items in its fulfillment centers. The team proposed several names for the new feature, including Super Saver Platinum, which Bezos rejected because he didn’t want people to see the service as a money-saving program. Bing Gordon, Amazon board member and partner at Kleiner Perkins, claims he came up with the name Prime, though some members of the team believe the name was chosen because fast-track pallets were in prime positions in fulfillment centers. Focus groups were brought into Amazon’s offices to test the Prime sign-up process. The volunteers found the process confusing, so Holden proposed using a large orange button with the words Create my Prime account right inside the button.

Selecting the fee for the service was a challenge; there were no clear financial models because no one knew how many customers would join or how joining would affect their purchasing habits. The group considered several prices, including $49 and $99. Bezos decided on $79 per year, saying it needed to be large enough to matter to consumers but small enough that they would be willing to try it out. “It was never about the seventy-nine dollars. It was really about changing people’s mentality so they wouldn’t shop anywhere else,” says Ravindran, who later became chief digital officer for the Washington Post.

Bezos was adamant about the February launch date. When the Prime team reported that they needed more time, Bezos delayed the earnings announcement by a week. The team members finished mapping out the details for the service at three o’clock in the morning on the day of the deadline. It was a complex undertaking, but it was achievable because so many of the elements of the program already existed. Wilke’s organization had created a system for the expedited picking, packing, and shipping of prioritized items within the FCs. The company’s European operation had built a subscription-membership tool for its nascent DVD-by-mail business (a Netflix clone) in Germany and the United Kingdom, and that service, though rudimentary, was quickly improved and pushed into production in the United States to support Prime. “It was almost like Prime was already there, and we were putting the finishing touches on it,” Holden says.

In many ways, the introduction of Amazon Prime was an act of faith. The company had little concrete idea how the program would affect orders or customers’ likelihood to shop in other categories beyond media. If each expedited shipment cost the company $8, and if a shipping-club member placed twenty orders a year, it would cost the company $160 in shipping, far above the $79 fee. The service was expensive to run, and there was no clear way to break even. “We made this decision even though every single financial analysis said we were completely crazy to give two-day shipping for free,” says Diego Piacentini.

But Bezos was going on gut and experience. He knew that Super Saver Shipping had changed customers’ behavior, motivating them to place bigger orders and shop in new categories. He also knew from 1-Click ordering that when friction was removed from online shopping, customers spent more. That accelerated the company’s fabled flywheel—the virtuous cycle. When customers spent more, Amazon’s volumes increased, so it could lower shipping costs and negotiate new deals with vendors. That saved the company money, which would help pay for Prime and lead back to lower prices.

Prime would eventually justify its existence. The service turned customers into Amazon addicts who gorged on the almost instant gratification of having purchases reliably appear two days after they ordered them. Signing up for Amazon Prime, Jason Kilar said at the time, “was like going from a dial-up to a broadband Internet connection.” The shipping club also keyed off a faintly irrational human impulse to maximize the benefits of a membership club one has already joined. With the punitive cost of expedited shipping, Amazon lost money on Prime membership, at first. But gradually Wilke’s organization got better at combining multiple items from a customer’s order into a single box, which saved money and helped drive down Amazon’s transportation costs by double-digit percentages each year.

Prime wouldn’t reveal itself to the world as a huge success for another few years, and originally it was unpopular inside Amazon. One technology executive griped to Vijay Ravindran that he feared Bezos would now believe that he could commandeer engineers and ram his favorite projects through the system. Other execs were wary because of Prime’s estimated losses. Almost alone, Bezos believed fervently in Prime, closely tracking sign-ups each day and intervening every time the retail group dropped promotions for the shipping club from the home page.

But even back in February of 2005, Bezos suspected he had a winner. At Amazon’s all-hands meeting that month at the usual location, the classic Moore Theater on Second Avenue, Vijay Ravindran presented Prime to the company, and afterward Bezos led everyone in a round of applause.

Prime opened up new doors, and the next year Amazon introduced a service called Fulfillment by Amazon, or FBA. The program allowed other merchants to have their products stored and shipped from Amazon’s fulfillment centers. As an added benefit, their products qualified for two-day shipping for Prime members, exposing the sellers to Amazon’s most active customers. For Wilke’s logistics group, it was a proud moment. “That is when it really hit home,” says Bert Wegner. “We had built such a good service that people were willing to pay us to use it.”

So when Bezos pulled Wilke out of an operating review in late 2006, Wilke wasn’t expecting to hear that that holiday season would be his last in the world of logistics. Bezos wanted Wilke to take over the entire North American retail division, and Wilke was charged with finding his own replacement. Wilke thought that Amazon’s progress in its FCs had plateaued, so instead of promoting from within the ranks of Amazon’s logistics executives, all of them molded, as he was, by the dogma of Six Sigma, Wilke went looking for someone with a fresh approach and additional international experience.

The search led him to Marc Onetto, a former General Electric executive with a thick French accent and a gift for animated storytelling. Under Onetto’s watch, engineers once again rewrote elements of Amazon’s logistics software and devised a computer system, called Mechanical Sensei, that simulated all the orders coursing through Amazon’s fulfillment centers and predicted where new FCs would most productively be located. Onetto also shifted Amazon’s focus toward lean manufacturing, another management philosophy that emanated from Toyota and was directed at eliminating waste and making practical changes on the shop floor. Japanese consultants occasionally came to work with Amazon, and they were so unimpressed and derogatory that Amazon employees gave them a nickname: the insultants.

Though Amazon was intensely focused on its software and systems, there was another key element of its distribution system—the low-wage laborers who actually worked in it. As Amazon grew throughout the decade, it hired tens of thousands of temporary employees each holiday season and usually kept on about 10 to 15 percent of them permanently. These generally low-skilled workers, toiling for ten to twelve dollars an hour in places where there were few other good jobs, could find Amazon to be a somewhat cruel master. Theft was a constant problem, as the FCs were stocked with easily concealable goodies like DVDs and jewelry, so the company outfitted all of its FCs with metal detectors and security cameras and eventually contracted with an outside security firm to patrol the facilities. “They definitely viewed everyone as someone who could potentially steal from them,” says Randall Krause, an associate who worked at the Fernley FC in 2010. “I didn’t really take it personally because probably a lot of people actually were stealing.”

Amazon tried to combat employee delinquency by using a point system to track how workers performed their jobs. Arriving late cost an employee half a point; failing to show up altogether was three points. Even calling in sick cost a point. An employee who collected six such demerits was let go. “They laid out their expectations and if you didn’t meet them, they had people waiting to take your job,” says Krause. “They wouldn’t give you a second chance.”

Over the years, unions like the Teamsters and the United Food and Commercial Workers tried to organize associates in Amazon’s U.S. FCs, passing out flyers in the parking lots and in some cases knocking on the doors of workers’ homes. Amazon’s logistics executives quickly met these campaigns by engaging with employees and listening to complaints while making it clear that unionizing efforts would not be tolerated. The sheer size of Amazon’s workforce and the fact that turnover is so high in the fulfillment centers make it extremely difficult for anyone to organize workers. Most recently, in 2013, workers at two Amazon FCs in Germany went on strike for four days, demanding better pay and benefits. The company refused to negotiate with the union.

The unions themselves say there’s another hurdle involved—employees’ fear of retribution. In January 2001, the company closed a Seattle customer-service call center, as part of a larger round of cost-cutting measures. Amazon said closing the facility was unrelated to recent union activity there, but the union involved was not so sure. “The number one thing standing in the way of Amazon unionization is fear,” says Rennie Sawade, a spokesman for the Washington Alliance of Technology Workers. Employees are “afraid they’ll fire you—even though it’s technically not legal. You’re the one who has to fight to get your job back if they do.”

Amazon often had to contend with something even more unpredictable than stealing, unionization, or truancy in its FCs: the weather. Company managers learned quickly that they had no choice but to install air-conditioning in their first fulfillment centers in Phoenix, where the summers were brutal, but they skimped on what they viewed as an unnecessary expense in colder climates. Instead, fulfillment-center managers developed protocols to deal with heat waves. If temperatures spiked above 100 degrees, which they often did over the summer in the Midwest, five minutes were added to morning and afternoon breaks, which were normally fifteen minutes long, and the company installed fans and handed out free Gatorade.

These moves sound almost comically insufficient, and they were. In 2011, the Morning Call, an Allentown newspaper, published an exposé about poor working conditions in Amazon’s two Lehigh Valley fulfillment centers during that summer’s brutal heat wave. Fifteen workers suffered heat-related symptoms and were taken to a local hospital. An emergency room doctor called federal regulators to report an unsafe working environment. In a detail that struck many readers and Amazon customers as downright cruel, the newspaper noted that Amazon paid a private ambulance company to have paramedics stationed outside the FCs during the heat wave—ready to deal with employees as they dropped.

Jeff Wilke argues that Amazon’s overall safety record, as reflected in the low number of incidents reported to the Occupational Safety and Health Administration, or OSHA, demonstrates that it is safer to work in the company’s warehouses than in department stores. (The low number of recorded complaints to OSHA regarding Amazon facilities backs up this contention.3) In terms of public perception, though, it didn’t matter. The report sent shock waves through the media, and the following year, battered by the negative publicity, Amazon announced it was paying $52 million to install air-conditioning in more of its fulfillment centers.4

Bezos and Wilke could battle chaos, they could try to out-engineer it, but they could never eradicate it completely. The capricious and unpredictable quirks of human nature always managed to emerge in unexpected ways, like in December of 2010, when a disgruntled employee set a fire in a supply room in Fernley. Employees were evacuated and had to stand out in the cold shivering for two hours before being sent home, according to two employees who were there. That same year in Fernley, a worker preparing to quit hoisted himself onto a conveyor belt and took a long joyride through the facility. He was subsequently escorted out the door.

Perhaps the best story stems from the busy holiday season of 2006. A temporary employee in the Coffeyville, Kansas, fulfillment center showed up at the start of his shift and left at the end of it, but strangely, he was not logging any actual work in the hours in between. Amazon’s time clocks were not yet linked to the system that tracked productivity, so the discrepancy went unnoticed for at least a week.

Finally someone uncovered the scheme. The worker had surreptitiously tunneled out a cavern inside an eight-foot-tall pile of empty wooden pallets in a far corner of the fulfillment center. Inside, completely blocked from view, he had created a cozy den and furnished it with items purloined from Amazon’s plentiful shelves. There was food, a comfortable bed, pictures ripped from books adorning the walls—and several pornographic calendars. Brian Calvin, the general manager of the Coffeyville FC, busted the worker in his hovel and marched him out the door. The man left without argument and walked to a nearby bus stop; sheepish, one might imagine, but perhaps also just a little bit triumphant.