The Way Amazon Uses Tech to Squeeze Performance Out of Workers Deserves Its Own Name: Bezosism
For Austin Morreale, working as a stower in an Amazon AMZN -0.43% warehouse was tough to the point of being physically unsustainable, but nonetheless rewarding. The hours were long and the work grueling. The night shift he took at Amazon on top of his day job as a case manager for a nonprofit group was plainly unsustainable, but he only planned to do it for a summer, anyway. He needed the money, the immediate access to health insurance, and the change of pace. He lasted six weeks.
Mr. Morreale, 50 years old, worked at the LGA9 fulfillment center in Edison, N.J., and says that while many people he trained alongside quit within their first two weeks on the job, he “actually had a good experience there.” But it was hard work—which in some ways reminded him of his days as a high-school athlete. “It was 10 hours of pretty much mind-numbingly boring work, pretty much standing in the same position for the whole shift,” he said. “But at the end of the shift, I was drenched in sweat and aching like I hadn’t ached since I was playing competitive soccer.”
Amazon warehouse workers who can’t ‘make rate’ don’t last. PHOTO: JEREMY M. LANGE FOR THE WALL STREET JOURNAL
Mr. Morreale was slow, he says, and kept messing up the patterns for efficiently putting items on robotic shelves—known as stowing—that he had been taught. He couldn’t “make rate”: Amazonese for keeping up with the pace of work. But, he adds, his managers were generous and “super-invested” in helping everyone on his team improve.
On the job, no one ever stood behind Mr. Morreale and barked at him to work faster. They didn’t have to. Twice a day at a stand-up meeting, his shift managers told the group how everyone was doing. They knew because Amazon’s software, and an assortment of sensors in the warehouse, tracked workers’ every move. “Those numbers are always in the back of your head somewhere,” he says.
Mr. Morreale’s story represented pretty much the median experience of the Amazon fulfillment center workers I’ve interviewed. On one end of the spectrum, there were those who found the work intolerable, and lasted less than two weeks. At the other end were those with an appetite for the work and a tolerance for the long hours of isolation and repetitive motion it entailed.
More than a century ago, Frederick Winslow Taylor and Henry Ford pioneered systems for speeding up work that we take for granted today. What Mr. Morreale experienced was Amazon’s 21st-century, algorithm-driven successor to Taylorism and Fordism. It’s a mix of surveillance, measurement, psychological tricks, targets, incentives, sloganeering, Jeff Bezos ’ trademark hard-charging attitude toward work, and an ever-growing array of clever and often proprietary technologies. Taken as a whole, this system is novel enough in the history of work that it deserves its own name: Bezosism.
At this very moment, Bezosism is diffusing through the world of work, rewriting the source code of the global industrial machine. If it proves as popular and durable as the systems of organization on which it builds—from Fordism to the Toyota Production System—it could be, along with the e-commerce and space companies he built, Mr. Bezos’ most important legacy.
Depending on how the company practicing Bezosism wields its power, this system of technologically supercharged management can be benevolent, or sinister, or both.
Take, for example, Amazon’s well-known metric for evaluating worker performance—the “rate” that Mr. Morreale was unable to hit.
In Amazon’s fulfillment centers, human productivity is measured by an overall pick or stow rate calculated for each worker at a robot-fed pick-and-stow station.
Imagine the delight of Taylor, who conceived “scientific management” in the early 20th century, or Ford, if they could know, to the millisecond, how long it took every worker to complete a task, every day, in every facility they owned. Imagine what early time-and-motion experts Frank and Lillian Gilbreth could have accomplished had they been able to discard their film cameras and replace them with millions of hours of video captured from the digital cameras that watch every station at Amazon’s fulfillment centers. Imagine how much additional just-in-time efficiency in inventory levels, capital allocation, and automated reordering Taiichi Ohno and Eiji Toyoda, creators of the Toyota Production System in postwar Japan, would be able to extract from a system that knew the precise moment a worker plucked an item from a shelf and sent it on its way.
That Amazon has all this data—and can manage its workers, evolve its automated systems, and innovate new robots based on it—is one of the reasons it’s the most valuable retailer on earth.
Workplace-efficiency pioneer Frederick Winslow Taylor would have had a field day with the technology at Amazon’s disposal. PHOTO: BETTMANN ARCHIVE/GETTY IMAGES
The overall rate at which workers must complete a task in an Amazon warehouse, whether it’s putting items on shelves, taking them off, or putting them in boxes, is calculated based on the aggregate performance of everyone doing that task in a given facility, says an Amazon spokeswoman. This floating rate, Amazon argues, shows that none of its employees is being pushed beyond what’s reasonable, because that rate is something like an average of what everyone in a warehouse is already doing.
“We don’t set unreasonable performance goals,” Mr. Bezos, now Amazon’s chairman, wrote in an April letter to shareholders.
But this is not how many Amazon workers, even those who regularly exceed the rate at their facility, see things. Anyone can have a bad week—maybe they’re sick, or exhausted from taking care of a child or relative, or maybe they’re developing one of the repetitive stress injuries that are not uncommon when people have to perform the same task for an entire 10-hour shift, with only a half-hour for lunch and two 15-minute rest breaks.
On Wednesday, California legislators advanced a bill to regulate companies like Amazon that employ quotas and other algorithm-driven work practices at their warehouses.
Knowing that if you don’t make rate you’ll get a warning, triggered by an algorithm, and if it happens often enough your job is in danger, can be a powerful psychological spur to work harder, and possibly to exceed your physical limits, as Mr. Morreale discovered.
One day at the fulfillment center, he pushed himself too hard. Lightheaded and clammy, he sank to his knees, a no-no that Amazon’s performance algorithm treats as “time off task.” Associates aren’t allowed to sit down while on the job, unless it’s lunchtime or one of their 15-minute breaks.
“I don’t know if it was overexertion or what it was,” Mr. Morreale says. “My supervisors never themselves made me feel pressure. I put that pressure on myself: ‘Oh, I’ve gotta hit those numbers. Oh, I’m doing terribly.’”
In his six weeks at Amazon, he developed carpal tunnel syndrome, which abated only after he quit the job, Mr. Morreale says.
A floating rate also pits all workers at a facility against one another, says Tyler Hamilton, a worker at an Amazon fulfillment center in Shakopee, Minn., who was 22 years old when I first interviewed him in 2019.
“If there are people who cut corners, if there are people who take tons of coffee and tons of energy drinks to go faster, that raises the cumulative rate,” says Mr. Hamilton. “Meaning, if you want to keep up with the average, then you have to cut corners and drink coffee and energy drinks at every break.”
Cutting corners and getting juiced on caffeine isn’t just something people do when it’s Prime Day or peak season. For many, it’s what they do all the time. “I mean, the coffee is free out of the machines,” adds Mr. Hamilton. Another thing that is free at Amazon warehouses is aspirin, available from no-cost vending machines scattered throughout the warehouse.
It’s difficult to quantify the impact of Bezosism on workers, but some have tried. In 2019, the last year for which data are available, Amazon reported 5.6 injuries per 100 workers. The average rate for warehouses in the U.S. that same year was 4.8 per 100, according to company and federal workplace data.
Eiji Toyoda was a co-creator of the Toyota Production System in postwar Japan. PHOTO: THE ASAHI SHIMBUN/GETTY IMAGES
Amazon has argued that its injury rates only look high because the company’s safety culture means that it obsessively documents incidents in a way that its competitors do not.
Amazon has introduced a number of initiatives to reduce worker injuries in recent months. Those include its Working Well program, which has now been rolled out to 1,000 of Amazon’s approximately 2,000 facilities world-wide, says Heather MacDougall, vice president of workplace health and safety. (Amazon has more than 750,000 employees in positions that involve physical labor or management of people in those positions.) The company also added “strive to be the earth’s best employer” to its list of leadership principles and announced a partnership with the not-for-profit National Safety Council to find new ways to reduce the incidence of musculoskeletal disorders, which are the most common type of injury in warehousing and logistics. (These include, for example, repetitive stress injuries.) The company has also pledged to spend $300 million in 2021 to increase safety.
While it might seem as if the technology Amazon is using inevitably leads to a speedup in the pace and demands of work in its warehouses, former Amazon executives who designed these systems in the first place told me that their effects on workers are entirely up to the company’s leaders.
Kiva Systems, the robotics company Amazon acquired in 2012 and refashioned into Amazon Robotics, and which developed the robotic drive units that move shelves in Amazon’s fulfillment centers, used to serve customers other than Amazon. When Kiva’s engineers and managers first started rolling out their robots in warehouses belonging to companies like Walgreens, employees loved them, says Kiva founder Mick Mountz, who became an Amazon executive after the acquisition, and left the company in 2015. And why wouldn’t they? Employees went from walking 10 or more miles a day to retrieve items for delivery to walking almost none, because the inventory came to them, atop robots.
But imagining that a new technology that can make someone more productive will ultimately mean they have to do less work is a classic mistake. History shows that every time we automate a task, we tend to use more of the product or service requiring that task, in combination with others, to accomplish some other, more complicated or difficult end.
As Amazon itself puts it in public statements, “The fulfillment centers that have robots often have higher employment numbers because inventory is moved at a faster pace, which requires extra associates.”
Robots have made it possible for Amazon to increase its warehouses’ output, but it is decisions by management that have required workers to keep up a blistering pace. PHOTO: AMAZON
A worker using the Kiva system in its early incarnations would typically triple their output, say from an average of 100 picks an hour to 300, says Mr. Mountz. But it wasn’t as if the Kiva-using companies then reduced all their warehouse employees’ hours to a third of what they once were while paying them the same wage. Instead, Staples and Walgreens, both early customers of Kiva, used their workers’ increased productivity to increase the output capacity of their warehouses; store and ship a wider range of products; shorten the amount of time required to fulfill an order, and ultimately either lower the cost of their services, increase their profits, or both. All reasons Amazon, a customer of Kiva, decided to acquire it.
At Amazon, the “rate” is the purest expression of the company’s goals. Amazon’s leaders and spokespeople like to talk about how automation makes the job of an associate easier. But, until very recently, they seemed unable or unwilling to imagine that the increased demands of that automation on the associates could be grinding them down both physically and psychologically.
“We develop these [rate] targets across an extended period of time using actual employee performance,” says Ms. MacDougall, the health and safety executive. “We take into account a variety of factors, and everything is with the safety and well-being of employees front and center.”
I asked Mr. Mountz to comment on the injury rate at facilities with robots he and his engineers designed at Kiva. In the original design of the Kiva system, he answered, “We always pointed out the human is in control of the machine, not the other way around. We’d say, this is not the Lucille Ball episode where she’s on the chocolate line.”
SHARE YOUR THOUGHTS
Should companies use surveillance technology to maximize workers’ productivity? Join the conversation below.
In other words, in Amazon’s system the pace at which a worker picks, stows or packs goods is up to them, so the automation flexes to accommodate their pace. “Whether a customer, be that Amazon or Walgreens, says you have to pick 800 items an hour or 300 an hour, that’s a function of the type of inventory you’re handling, and management philosophy,” adds Mr. Mountz.
Current and former Amazon executives described its management philosophy to me as performance-driven and hard-charging, built on the idea that everyone should be pushed to their limits and underperformers should be cut. Amazon clearly wants the world to believe that that is changing. Whether or not those changes will have a meaningful impact on the lives of hundreds of thousands of its entry-level associates, whose work lives are ruled by sensors and algorithms, who do the physically demanding labor on which Amazon’s e-commerce empire depends, and for whom the pace and tenor of their work is a function of decisions made by company leaders, as much as technology, remains to be seen.
Adapted from “Arriving Today: From Factory to Front Door—Why Everything Has Changed About How and What We Buy,” by Wall Street Journal technology columnist Christopher Mims, to be published on Sept. 14 by HarperBusiness, a division of HarperCollins Publishers. Copyright © 2021 by Christopher Mims.