by Brad Stone
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 manufacturer
s 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.
