The vast Illinois factory floor that will produce electric delivery vans for Amazon.com Inc. is starting to fill up. Battery and chassis assembly here. Tire storage there. A titanic, three-story metal press.
Sometime next year, the plant’s workers will complete the first van. A decade later, if all goes according to plan, van No. 100,000 will roll off the line, completing the biggest electric vehicle purchase in history. By the time Amazon has all these zero-emission vehicles on the road, around 2030, it will eliminate an estimated 4 million metric tons of CO2 now belched out each year by gasoline-powered vans and trucks.
That would wipe out a small company’s entire contribution to the emissions warming the planet. But for the largest online retailer, the savings represent roughly 8% of its carbon footprint. How Amazon hopes to tackle the other 92%—as well as any other emissions the fast-growing company tacks on in the coming years—is an open question.
Amazon last year committed to zeroing out its carbon footprint by 2040, or eliminating the greenhouse gas emissions caused by its activities. The challenge is daunting. The company is increasingly transporting its own goods from warehouses to customer doorsteps, acquiring the planes, long-haul trucks, and delivery vans that make its aggressive delivery promises possible. Its cloud computing division is building giant, electricity-hungry server farms from Sweden to South Africa. Through contract manufacturers, it’s a major builder of consumer electronics, household essentials, and apparel. Last year its emissions per dollar of merchandise declined, but its total emissions grew.
Amazon concedes there isn’t a clear path to that zero-carbon goal today. “We’re going to invest a lot, and we’re going to invent along the way,” says Kara Hurst, the company’s first sustainability director.
It’s a big admission about a big goal: There’s no specific plan for how to reach it. In some ways, it’s a perfect example of what happens when the build-the-plane-as-you’re-flying-it tech culture so celebrated in American business intersects with an unyielding scientific crisis. Amazon has been a pioneer again and again—and now it’s gunning to be the first company that figures out how to slow climate change at the same time it moves millions of items around the planet.
“It’s fine to set goals where you don’t know exactly what you’re going to do. I’m actually a big fan of that,” says Bill Weihl, a former sustainability executive at Facebook Inc. and Google, who runs ClimateVoice, an activist group focused on climate policy. “If I were to raise a concern in terms of what they’re planning, there’s not a lot of detail” in the company’s public plans so far, he says. “Amazon has put out a bold target, and now they need to show real progress.”
The company, acknowledging it can’t do it alone, has launched the Climate Pledge, an open invitation to other companies to match its own zero-carbon target and share their findings. (So far five companies have signed on, including Verizon Communications Inc. and the Indian software company Infosys Ltd.) It also established a $2 billion fund to back moonshots, pledging to invest in promising decarbonizing technologies from logistics to agriculture. But it’s going to take more than collaboration and money; it’s going to require an operational shift.
Amazon employees are trained to get customers to buy more sneakers, use more cloud services, and stream more movies. There’s a pattern that repeats in some of the company’s more carbon-intensive corners, whether it’s the vans racing to deliver orders within two hours, Amazon Air flights arranged to meet delivery-time goals, or the trailers rolling pallets of goods between warehouses: growth first, efficiency later.
“We have to go big first,” says one employee, who, like several current and former workers interviewed for this story, requested anonymity because of confidentiality agreements. In an endless effort to get more customers, this employee says, “We launch unsustainable—even financially unsustainable—endeavors. … It ties back to Amazon’s whole premise: The customer is always right.”
That leads to the bigger question hanging over the whole enterprise. Can a company geared toward satisfying shoppers above all else ever be truly sustainable?
For someone who hadn’t spent a lot of time inside corporations prior to this job, Hurst, 48, comes across as a classic Amazon executive. She’s tenacious and intense, colleagues say. And despite occupying a role that for years didn’t come with the authority to demand changes of other Amazon teams, she developed a reputation for holding her own inside a corporate culture whose guiding principles discourage compromise for the sake of comity.
For more than a decade, Hurst worked for organizations that advise companies such as General Electric Co. and Hyatt Hotels Corp. on social responsibility issues. When an executive recruiter working for Amazon contacted her in 2014, the company was known as a laggard on climate change action. Its initiatives at the time included backing renewable energy projects and sending items in smaller boxes, both of which helped cut costs.
For most of Amazon’s existence, topics not directly related to meeting customer needs were an afterthought. Jeff Bezos, founder and chief executive officer, said its contribution to the greater good came from its business activities, a refrain the company repeated for years when confronted with its absence from corporate social responsibility circles and its paltry philanthropic record. But when Hurst met with Amazon executives, she found they were receptive to her ideas, and to doing more.
She spent her first week at a Phoenix warehouse in a training session for logistics managers, followed soon after by a ride-along with a driver in the retailer’s then-new grocery delivery service—a grand emissions tour of a sort.
Amazon was changing quickly. Until then it had primarily been a middleman, linking shoppers to goods stored in its warehouses and handing packages off to carriers for delivery. To support its aggressive shipping promises, it had recently started building new types of warehouses closer to big cities and hiring small trucking and delivery companies to reduce dependence on such carriers as the U.S. Postal Service or United Parcel Service Inc. During Hurst’s first year at the company, Amazon took steps toward building its own air fleet, chartering flights with cargo carriers and staking a claim in the quickest—and most carbon-intensive—way to move goods.
While Hurst would have latitude to experiment, there were constraints. Projects could lose money, but they had to make the same kind of business case as other units, backed up with copious data. She started small, with projects like creating energy-efficiency standards for buildings and making sure warehouse workers stopped using so much tape.
“We really wanted to be pragmatic about it,” she says. “There are a lots of ways in which we can think about saving the company money and driving new ways of providing value to customers.”
Hurst quietly began building her team with sustainability experts from rivals, academia, and elsewhere at Amazon, mixing devoted green believers with career engineers and project managers. In 2016 her group numbered about 50 people; by mid-2020 it would grow to 150. Some of them came to feel the company wasn’t moving fast enough.
“There were times when there was a lot of friction,” says one former member of the team. “There was a moment when a lot of us had to kind of slow ourselves down in terms of our philanthropy, humanitarian agendas, and our bleeding hearts. Kara was just very honest. She said we work for a business, and we have to make sure that we do things according to the boss. I got the message loud and clear.”
While advising Amazon business units about recycling programs and opportunities to cut costs, Hurst was aiming higher. She introduced the idea of a potential zero-carbon goal at a planning meeting for operations executives in New York in June 2016, making the case that the company could meet aggressive climate targets without breaking its promises to customers.
“It’s not just ‘let’s come up with stuff that’s an easy win,’ ” says Adam Elman, who led Amazon’s sustainability efforts in Europe for three years and left the company in June. “The pressure was, ‘We know we can do that. What are the difficult conversations?’ ”
To help drive down emissions, Hurst hired a guy who played a key part in shaming Nike Inc. over its labor practices. Dara O’Rourke started his career consulting for global development groups, making a name for himself by publicizing a report showing that even as Nike pledged better working conditions, its consultants were raising concerns about carcinogens and illegal overtime at a factory in Vietnam. He later co-founded Good Guide, a now-defunct website that rated consumer products by environmental and social factors, a do-gooder contrast to the customer reviews Amazon made ubiquitous.
In 2016, O’Rourke took a two-year leave of absence from teaching at the University of California at Berkeley to build models to quantify the climate impact of Amazon’s operations, work that could eventually help make the business case for a more aggressive sustainability program.
“It took a lot longer than I expected,” says O’Rourke, who continues to oversee the sustainability data science team. “I didn’t understand the complexity of Amazon.”
Last year, Amazon’s emissions per dollar of merchandise sales dropped 5%, meaning the company was able to warehouse and deliver packages a bit more efficiently than in the prior year. While headed in the right direction, that didn’t swing the overall numbers: Its total emissions in 2019 rose 15%, to 51.1 million metric tons of carbon dioxide equivalent. That’s about three times the total of Google parent Alphabet Inc. and twice that of Apple Inc. It’s a footprint closer in size to the largest retailers and some of the biggest power companies in the U.S. Walmart Inc., Amazon’s biggest rival in U.S. retail, reports emissions from its direct operations that are 60% higher than Amazon’s. But comparisons between the total footprints of the two—including indirect, so-called scope 3 emissions—are imprecise, in part because while Walmart declares some of its scope 3 emissions, neither company reports the totality of those numbers, as many other companies do.
Apple is pushing the manufacturers who build its iPhones and other gadgets to power their operations with renewable energy. Facebook mostly has to worry about sourcing clean energy for its data centers because, unlike Amazon, it doesn’t make or move as many physical objects around. Microsoft Corp. says it can get to negative carbon emissions numbers by 2030.
Perhaps the biggest contrast between Amazon and other companies trying to tackle their carbon problem is how quickly Amazon is growing. Sales increased by an average 28% a year in the last decade, a breakneck pace for a company already among the largest in the world.
Anyone who’s shopped online has seen the effect of Amazon’s seamless browsing experience in enabling an impulse buy. The company says it aims to be the most environmentally friendly way to buy anything, but it hasn’t publicly addressed its influence on shopper behavior or expectations for fast delivery. It offers consumers the ability to cluster deliveries on a single day of the week, but has so far stopped short of offering special shipping windows for customers willing to cut their carbon footprint, a step taken by some other retailers.
Elizabeth Sturcken, who leads the Environmental Defense Fund’s partnerships with companies, compares the evolution under way at Amazon to that of Walmart, which started to embrace environmentalism when it was under scrutiny for its anti-union stance, environmental record, and market power. (That’s a cyclone that Amazon, a target of antitrust regulators and labor organizers around the world, understands well.)
In the mid-2000s, Walmart, in partnership with the EDF and other environmental organizations, began adding parts of its emissions-reduction campaign to its low-price credo. “The same thing probably needs to happen at Amazon,” Sturcken says.
Amazon is “operating in a context where a lot of companies have been doing this for a decade or more,” says Weihl, the ClimateVoice chief. “So they are playing catch-up. You can cut them a little slack for that. But they need to move quickly.”
Some of that work has already begun—the electric delivery vans, for one. Amazon has also set up warehouses closer to customers, reducing emissions from delivery. But what comes next is harder.
Three-quarters of its emissions come from indirect sources, whether deliveries by contractors or the manufacturers who make Amazon Fire TV streaming devices or servers for Amazon data centers. Making that supply chain less of an environmental drag would require zero-carbon marine transportation and zero-carbon air freight, technologies that don’t exist today in a fashion compatible with Amazon’s speed and scale.
The company’s carbon accounting excludes emissions related to the manufacture and use of most goods sold on its website. Amazon says the goal is to avoid double-counting of emissions tracked and reported by the manufacturers, but sustainability experts worry the effect understates the company’s true challenge.
If Amazon narrowly focuses on its own activities, rather than seeking broad change in everything from its supply chain to government policy, “they will fail,” Sturcken says. “You can’t have an aggressive climate commitment without addressing the products that you’re sending to millions of people’s homes every day.”
Hurst says Amazon is working to address its impact beyond its operations through programs with suppliers and partners. She also doesn’t see a conflict between Amazon’s aim to satisfy shoppers and its carbon goals. “I don’t think there’s a trade-off between doing the right thing and implementing wins for customers,” she says.
By 2018 Hurst’s team was busy quantifying the carbon impact of different aspects of the business and expanding investments in rooftop solar installations on its buildings—but the work was invisible to most of their colleagues. Amazon hadn’t released a full accounting of emissions. That abstention from basic corporate sustainability practices had become untenable after Amazon’s meteoric rise made it arguably the most emulated company in business.
In late 2018, some employees launched a campaign to get Amazon to lay out public plans to address climate change. The group, Amazon Employees for Climate Justice, got more than 8,000 of their co-workers to back a shareholder resolution. Ultimately rejected by investors, it was an unprecedented employee movement at a company with little activist tradition.
As the employee initiative was brewing, Hurst’s team in February 2019 received approval to announce Amazon’s first major sustainability goals in years: commitments to making half of the company’s deliveries with zero-carbon emissions by 2030, and to release its carbon footprint for the first time.
Hurst says her team took the employee activism as a sign of growing passion for climate issues among Amazon’s customers, one outside signal among many as the company evaluated its plans. The employee group contends it’s clear Amazon reacted in response to worker pressure. Amazon would later fire two of the activists for what it says were violations of company policy.
By the beginning of 2019, the company had spent years evaluating batteries and electric delivery vehicles. The big problem for Amazon was that no one was manufacturing electric vans in large quantities.
Ross Rachey, Amazon’s delivery fleet director, was charged with fixing that. A month after the company announced its zero-carbon delivery goal, he had his first meeting with Rivian Automotive Inc. Amazon had recently become the largest investor in the startup, which was developing electric trucks and SUVs, a would-be Tesla Inc. for the outdoorsy crowd.
Rachey worked with Rivian to sketch out a delivery van that placed an Amazon-designed cab and storage space atop battery and automotive components Rivian was already making, and committed to buy 100,000 vehicles—more than twice as many vans as Amazon had at the time. Now, Rivian is retrofitting a former Mitsubishi factory in Normal, Ill., which will produce the electric vans expected to eliminate 8% of Amazon’s carbon footprint.
The companies declined to discuss terms of the deal. But internal Amazon estimates suggest that savings on fuel and maintenance costs, along with tax breaks, would likely make a deployment of electric vans cheaper than Amazon’s current fleet over the lifetime of the vehicles, according to two people who’ve seen the calculations.
That’s part of the point. Amazon, with its size and ability to invest without expectation of an immediate return, can help make out-of-reach technologies economically feasible. It’s perhaps the biggest sign of how the company can reach its aggressive climate goals while still watching the bottom line.
“If you want to be a large corporation and actually bend your carbon curve, you might have to take an active role in designing technologies that help do that,” Rachey says.
Bezos praised that work in September 2019. At a rare press conference, held one day before thousands of his employees joined the Global Climate Strike, he announced Amazon’s ambition to become a net-zero emitter of carbon by 2040. (The “net” acknowledges that Amazon will likely purchase carbon offsets, which typically underwrite preservation or growth of forestland.) He also announced the Rivian order and the Climate Pledge. The pledge was a last-minute idea from Bezos, employees and partners say, a characteristic flourish from a CEO who likes to package Amazon’s plans into big gestures.
Backed by that mandate from the top, environmental goals are filtering down through the company. Models built by O’Rourke’s team are in the hands of many groups across Amazon, allowing managers to determine how many grams of carbon dioxide were emitted over the course of a particular package’s trip, or the lifetime electricity use of a Kindle e-reader. Some employees say they haven’t yet seen any difference in their jobs to reflect sustainability goals. Hurst and O’Rourke say that will change. Amazon’s planning cycle for 2021 asks each team what they plan to do to help the company meet its zero-carbon goal.
“It’s not negotiable anymore, and it has to be a part of how we think,” Hurst says. “It might not always rank No. 1, but I want it to be on the list.”
For now, customers—and the ultrafast package delivery times Amazon promises them—are still winning many debates.
Next year, Amazon will start flying its 80th leased aircraft, one of a fleet cobbled together from older planes retired by airlines in favor of newer, more fuel-efficient models. Amazon has been among the largest buyers of 767 jets converted from passenger to freight use in the last five years. It hasn’t announced any purchases of new aircraft from Boeing Co. or Airbus SE.
That’s a big addition to Amazon’s carbon bill. “It’s this blind fealty to convenience and a complete unwillingness to recognize that convenience comes at a cost,” says Sucharita Kodali, who studies e-commerce at Forrester Research Inc. “In Amazon’s case, they’ve won so much on convenience.” Executives are likely wary of what might happen to customer loyalty if they ever stop pushing for the kind of faster delivery speeds enabled by air cargo, she says.
And Amazon’s stake in air freight is no temporary investment. The company is spending $1.5 billion to build a cargo terminal at Cincinnati/Northern Kentucky International Airport capable of supporting more than 100 aircraft. In May it announced a western hub in California, confirming plans to occupy a cargo facility being built in San Bernardino.
The company earlier this year announced the acquisition of 6 million gallons of jet fuel derived from biofuels. Dave Bozeman, the vice president who oversees Amazon’s transportation group, says the deal signals an ambition to “go big” on sustainable aviation fuel. Yet based on the size of Amazon’s fleet, that purchase likely accounts for less than 1% of the company’s annual fuel use.
For now, Bozeman says his team is working to meet Amazon’s delivery guarantees while being mindful of climate goals. That means eliminating the wasteful trips that can sometimes come with scrappy new lines of business. “My job is to push back on that and say, no, I think there’s a bigger purpose here,” he says.
There are other areas that have attracted the ire of critics. Like many large companies, Amazon doesn’t make clean energy policies a priority for its lobbyists, people who track climate policy say. Amazon continues to sell technology services to oil and gas drillers even as some peers, such as Google, back away from the industry. And after Bezos announced that the Climate Pledge would lend its name to a renovated Seattle sports arena, a pair of sustainability researchers at the University of Washington raised concerns that the pledge could end up as a public relations stunt instead of a way to rally others to fight climate change.
The company is famously secretive, with a reputation for incubating many long-shot projects for years before disclosing them. During a recent telephone interview, O’Rourke hinted at the work being done by his team, conferring several times with a public-relations representative on what he could disclose. (The names of Amazon’s carbon emissions models? Off-limits.) “At some point, hopefully we’ll be able to talk about some of the other stuff we’re working on,” he says.
Right now, he’s reluctant to predict a date when Amazon’s emissions might start to actually decline.
“It will tick up for a while,” he says, and progress won’t come in a straight line. “There are going to be hockey stick moments, and there are going to be tough years where we’re slogging through, trying to figure stuff out.” —With Siddharth Philip and Akshat Rathi
To contact the author of this story:
Matt Day in Seattle at firstname.lastname@example.org