How Amazon Is Using Whole Foods in a Bid for Total Retail Domination

Now they were meeting their new overlords for the first time. Whole Foods CEO John Mackey introduced Jeff Wilke of Amazon, who had flown in for the gathering. Wilke, the e-commerce giant’s CEO of Worldwide Consumer, decided to play to his foodie audience.

“I wanted to tell you just a little bit about how Whole Foods changed my life as a start,” he said. “As I was sitting this morning, eating breakfast, watching the sun rise over this beautiful city—by the way, quinoa, blueberry, and some other vegetables…”

That’s when Mackey, a vegan who avoids refined foods and travels with a rice cooker, lightheartedly corrected him. “Those aren’t vegetables,” he said. “That’s okay. We’re learning.”

In that moment, it was clear why Wilke and his team needed Whole Foods. His comment may have been just a slip of the tongue, but it reflected a persistent issue for the company: Amazon has expertise in many areas, but food is not one of them. For a decade, Amazon—a company with $178 billion in revenue and seemingly limitless resources—had not come close to breaking the billion-­dollar sales mark in its fresh food operation.

Rank 8

2017 Company Profile: Amazon.

[fortune-brightcove videoid=5786265530001]

The lack of progress is not entirely the retail giant’s fault. Grocery is a notoriously difficult business—and that’s before you start layering on the costs and challenges of delivery. Scott Galloway, a ­professor of marketing at New York University’s Stern School of Business, boils the problem down to this: A head of lettuce has a ­margin of less than a dollar and can survive outside the fridge for no more than a day. How can a retailer deliver it at peak ­quality—and make a profit?

But with the $13.7 billion acquisition, Amazon had bought itself a real shot at remaking the $800 billion U.S. grocery sector—the last frontier of e-commerce and a massive one at that. Some 20% of retail spending goes toward food, but only 2% of those sales take place on the Internet. “Grocery is the Wild West for online,” says Carrie Bienkowski, the chief marketing officer of online grocer Peapod. “The size of the prize is huge, and it’s growing.”

The very thing that makes grocery delivery hard—that food goes bad—is the reason it’s so desirable to a company like Amazon. Because cheese grows mold and meat goes rancid and milk sours, consumers can’t hoard it in their cupboards or refrigerators indefinitely as they might toilet paper or laundry detergent. As a result, the average family hits the supermarket at minimum once a week; there’s nothing else you purchase or consume so much or so often. For Amazon, getting in on that frequency is critical to further ingraining itself in our routines and behaviors. “Food is the platform for selling you everything else,” says Walter Robb, the former co-CEO of Whole Foods. “It’s an everyday way into your life. There’s nothing else that happens quite that way.” Amazon’s quest is therefore about much more than just food.

Amazon is perhaps the most disruptive and innovative company in retail, but using food as a lever for growth is nothing new. Walmart became the biggest retailer in the U.S. by turning itself into the nation’s largest grocer. In its fiscal 1998, 14% of Walmart’s U.S. sales came from grocery. This year, as the company hit $500 billion in revenue, that figure jumped to 56%. “Walmart pioneered this,” explains Wedbush analyst Michael Pachter. “Once you get them in the store for groceries, they walk up and down the aisle for everything else.” Says Jack Sinclair, who led Walmart’s U.S. grocery business for eight years, “The principle of what Amazon is doing is almost exactly the same.” Indeed, JPMorgan estimates that Amazon will match Walmart in U.S. sales by 2021.

Amazon declined to make any of its executives available to be interviewed for this story and has said little publicly about its long-term strategy for Whole Foods—or food more broadly, for that matter. But as a sign of how critical the sector is to ­Amazon’s future, Steve Kessel, part of CEO Jeff Bezos’s inner circle and a key figure on the Kindle team, has been brought in to run not only Whole Foods but also Amazon’s grocery delivery business, ­AmazonFresh, and Prime Now, its two-hour delivery offering.

For Amazon watchers, the company’s purchase of Whole Foods and its physical footprint—for more money than it had spent on all of its previous acquisitions combined—is a reflection of Bezos’s broader ambitions. Going as far back as a decade, Bezos has been explicit that grocery is essential to his long-term vision. In order to become a $200 billion company, he has said, Amazon must learn how to sell food.

More than 40¢ of every dollar consumers spend on the Internet already goes to Amazon—an astonishing sum. And yet it appears Bezos is not satisfied leaving behind the roughly 85% of retail that still happens in brick-and-mortar stores. For the CEO, owning our digital lives is not enough. By attempting to seamlessly link both realms, Amazon has the potential to be part of every single purchase we make.

“This is a monumental reference point,” says Robb. “This is clearly a revolution in the world of commerce.”

Books, where it all started for Amazon, are the anti-food. They never spoil and are simple to ship and handle. Fifty Shades of Grey, one of the site’s bestselling and most reviewed books of all time, is the same whether you buy it at your local independent store, at Barnes & Noble, or on Amazon.com. Books are what Credit Suisse analyst Stephen Ju calls “homogeneous, easy-to-handle” products, and Amazon has excelled at selling things that fit this criteria. If it can be shipped in a box and the purchaser cares little about whom she’s buying it from, it’s the perfect product for Amazon to sell.

Nonperishable groceries like canned goods, crackers, and cookies, which the company launched in the mid-2000s, were an obvious fit. “A box of Cheerios and a book aren’t that different,” says Ian Clarkson, who spent 15 years at Amazon. “You don’t have to fundamentally rewrite or build an entirely new infrastructure.”

Fresh goods, however, are about as heterogeneous and challenging to handle as it gets. Bananas bruise easily, no two are the same, and they are constantly evolving within the supply chain, going from green to ripe to mush.

In 2007, Amazon decided to try fresh delivery anyway. Customers were asking for it, and Amazon was intrigued by how it might be able to leverage the frequency of grocery ordering to sell higher-­margin nonfood products. The company launched AmazonFresh with a pilot in ­Seattle in which users would place an order online, and the food would be delivered to their doorstep.

A DELUGE OF DELIVERY OPTIONS

Before the Amazon deal, the only legitimate way to order Whole Foods’ product online was through delivery service Instacart. Now the floodgates have opened.

Amazon’s website, however, was not well suited for how people shop for food. Most e-commerce transactions comprise two to four items, which buyers find through targeted search. But shoppers tend to browse grocery, and an order can average 50 goods. “Amazon’s business, I’d argue even today, is around a unit of one,” explains Clarkson, who was AmazonFresh’s first full-time employee. “You buy it and move on. People don’t shop to fill their fridge that way.” And in grocery, where about 85% of the items people purchase are the same week to week, what customers bought recently is highly relevant to what they’ll buy again. That’s not the case for books. One copy of Fifty Shades is enough.

To cater to this different way of shopping, Amazon tried designing a separate website for AmazonFresh. The stand-alone site was necessary for another reason as well: Fresh food’s short shelf life required Amazon to roll out the service city-by-city—then an unusual launch strategy for the company. Amazon couldn’t just put products into a national fulfillment network and then turn on the web experience for the entire U.S. “That’s how traditional Amazon categories get an advantage,” says Clarkson. “You have so many people coming to the site. That doesn’t work for grocery.”

Kevork Djansezian—Getty Images

After six years in Seattle, AmazonFresh expanded to California but never grew its U.S. service beyond 20 states. The results have been mixed, and “that’s putting it kindly,” says Morningstar analyst R.J. Hottovy. The service was too expensive, the delivery windows were inconvenient, and people didn’t really understand where their food was being sourced from, he says. In November, Amazon said that it was cutting service in several zip codes, a move it contended was unrelated to the Whole Foods deal.

And perhaps most critically, AmazonFresh hasn’t been able to reach the scale it needs to get the service to work. The big difference between good and bad strawberries isn’t where they’re sourced from, explains Aaron Cheris, who leads Bain’s Americas retail practice; it’s who’s able to sell them faster. He says, “AmazonFresh was never able to achieve that by itself.”

Amazon is partial to building businesses rather buying them. But after a decade of trying to grow its grocery operation on its own, it was time for the latter. The company started exploring the possibility of an acquisition, and spent the two years leading up to the Whole Foods deal “walking around to every grocer in the U.S. asking them to be its fresh supplier,” says Bain’s Cheris.

Then, in April 2017, Amazon got a call from a consultant working on behalf of Whole Foods. The grocer had seen a report that Amazon may have been interested in buying the chain in the past. Would there be any appeal in setting up a meeting?

That first rendezvous came during a tumultuous period for Whole Foods. Competition was fierce in natural and organics, the very category it had essentially created, and the grocer was struggling to shake its “Whole Paycheck” reputation. Facing slowing sales growth and a flagging share price, Whole Foods was now clearly in play. Just weeks earlier, activist investor Jana Partners disclosed an 8.8% stake in the company. In addition to Amazon, four private equity firms and reportedly supermarket chain Albertsons were among those who had expressed interest in a potential deal.

The tie-up solved a lot of problems for both parties. For Whole Foods, Amazon offered freedom from the relentless cycle of short-term quarterly pressures as it tried to fix the business. For Amazon, Whole Foods gave the company instant scale and the built-in demand it had lacked in fresh food. In a logistics operation, companies have a set of fixed costs and become more profitable by layering on incremental business. Thanks to Whole Foods, Amazon now had guaranteed and predictable volume for its grocery infrastructure.

Along with scale, Amazon was buying credibility. Most of the products consumers buy on Amazon are branded—a Sony TV, a Hot Wheels car, a S’well water bottle. But with the exception of a few products, such as Bolthouse Farms carrots or Cuties clementines, fresh goods don’t have brands, or at least not ones that the consumer knows. Instead, we decide where to buy our broccoli and tomatoes based on our trust in the retailer. That authority was something Amazon just didn’t have in fresh. Whole Foods supplied it—as well as providing Amazon shoppers with a more appealing story about where their food originated. “The idea of ordering groceries online is conceptual,” says Barnaby Montgomery, cofounder and CEO of Yummy.com, a Los Angeles–based online grocer with brick-and-mortar stores. “I don’t know where it comes from. I don’t see it. I don’t get it. It’s a barrier.” Having a physical place to shop turns “the conceptual offer into something more tangible.”

One question that remains unanswered is how Amazon will use Whole Foods’ 483 physical stores to help solve what’s known in the industry as the “last mile” problem—the final and most expensive step of the delivery process that takes a product from a central hub to its final destination. There’s some speculation that Amazon could use the stores as mini distribution centers, hiring people to pick product off the shelf to fulfill online orders that would then be sent out for delivery. “It makes sense in theory, but it’s a much harder problem than it appears,” says Juozas Kaziukenas, CEO and founder of e-commerce intelligence firm Marketplace Pulse. “The way you stock for the warehouse and retail are two different things.” Grocery stores are maximized to make customers wander through and buy more stuff. Warehouses, where random products are stacked next to each other to make the most of the space, are much more efficient. They’re also better for fresh goods because increased handling hurts quality. “Every time you touch a fresh product it degrades it,” says Matthew Hamory of consultancy AlixPartners.

Right now, Amazon seems to be trying it all. When I placed an order via AmazonFresh in New York City, the food came from a New Jersey warehouse and was delivered off a truck by a contractor. When I did the same with Prime Now, which offers delivery in two hours, the goods came from a distribution facility in Midtown Manhattan and were dropped off by an Amazon Flex driver—an Uber-type model in which the company pays drivers $18 to $25 an hour. In 10 other markets, Amazon is offering Prime Now delivery directly from Whole Foods stores. That model could be a good stepping-stone in smaller cities until the company reaches enough volume to build out a dedicated warehouse. “You have to scale your way up there,” says grocery consultant Neil Stern. “Picking from a store, while crappy, is a way to establish volume and a customer base.”

The future may be a hybrid approach: bigger stores that have a space carved out for putting together online deliveries. Or a store where the selection of packaged goods is automated but customers can pick out their own fresh goods.

“They’re going to experiment like crazy,” says Tom Furphy, formerly a VP of consumables and AmazonFresh and now CEO of a venture capital firm. “I would be completely surprised if they have it all figured out by now.”

OTHER INDUSTRIES AT RISK OF GETTING AMAZONED

Grocery isn’t the only business Amazon is turning upside down. Here are five sectors that have had better watch their backs

With the average supermarket operating on a 1% profit margin, the economics of the grocery industry are fragile at best. What makes Amazon such a threat is that the same math just does not apply. Investors have accepted that Bezos is uninterested in short-term profitability. “They are one of our first competitors who doesn’t really care about making money,” Peapod cofounder Thomas Parkinson said at a food conference last year, “so that’s a challenge for us.” Food entrepreneur Greg Steltenpohl puts it this way: “They’re not disruptive because they have all the answers, but because they don’t have to play by the same rules.”

Food delivery is a strange reversion to an old way of doing business. The supermarket was invented in the early 20th century to replace the full-service store of the past, where a clerk behind the counter would pull products off the shelves for the customer. The new supermarket model required shoppers to push the carts and do the heavy lifting themselves in exchange for lower prices. Today, a growing segment not only wants some version of that old-timey archetype in which someone else collects the items for them, they also want them delivered to their homes—services that Goldman Sachs estimates cost a retailer $22.68 per order. That’s just not something most consumers are willing to cough up. “There’s a cultlike avoidance of paying for delivery,” explains analyst Brendan Witcher of research firm Forrester.

Despite the cost, grocery executives realize that’s where the market is heading. One Goldman estimate puts 20% of the industry online by 2027. That’s market share Amazon has in its sights—and that supermarket chains cannot afford to lose. The stakes are incredibly high. When Walmart reported in February that its e-commerce sales growth had slowed, the stock took its biggest one-day dive in three decades. Walmart has been leveraging its physical locations with click-and-collect, in which customers order online and pick up their groceries at the store, but during the earnings call the company said it would “accelerate” grocery delivery. (A Walmart spokesman said the lower fourth-quarter e‑commerce rate was largely planned, as the company met its full-year guidance.)

Stephen Brashear—Getty Images

Walmart’s ultimate weapon in the grocery wars is its massive store footprint; Amazon’s is Prime. In April, Amazon announced that the membership service had exceeded a whopping 100 million paid subscribers and that it would increase its annual fee. Morningstar’s Hottovy thinks that the market for new members may be plateauing, and that Amazon is now focused on increasing what current Prime users spend rather than recruiting new ones.

The Whole Foods deal was a case in point. According to 1010data, at the time of the acquisition, 81% of Whole Foods customers were already Amazon shoppers. So rather than capturing a new base, adding Whole Foods to the portfolio served as a tool to further ingrain Amazon customers within its ecosystem—offering Whole Foods’ popular 365 private label online, selling products like the Amazon Echo in its stores, and setting up Amazon lockers within Whole Foods for customers to pick up packages. Already, the symbiosis is at work. Since the deal closed, quick visits to Whole Foods were up 11% in stores with Amazon lockers.

Bezos says his aim is to make Prime so essential that it’s “irresponsible” not to join. And its value to the company is unmissable: By some estimates, Prime members spend an average $1,500 a year on Amazon, more than twice as much as non-Prime users. “Once you’re a Prime customer, you don’t go anywhere else to look,” explains AlixPartners’ Hamory. The goal, say insiders, is for Amazon to be the place where you start all of your searches. “Many people think our main competition is Bing or Yahoo,” Google’s onetime CEO Eric Schmidt has said. “But, really, our biggest search competitor is Amazon.”

Capturing search means capturing data. Amazon knows every product you look for, every category you choose, everything you put in your cart then abandon—what’s known as the “clickstream.” Now the company is trying to capture the real-life version of the clickstream with Amazon Go. The Seattle store uses overhead cameras and weight sensors in shelves to track shoppers so closely that they can simply leave the store when done shopping—no need to check out.

Might Amazon have something similar planned for Whole Foods stores? Perhaps in its lower-priced 365 locations, but experts are skeptical that the “just walk out” experience is coming to flagship outposts. One of Whole Foods’ assets is that it is high-touch—sensory, visual, and experiential. That’s a major selling point in an era where boring, mediocre retail is dead or dying.

The stores play one part in the all-encompassing, frictionless retail experience that Bezos has long aspired to create; Alexa, Amazon’s virtual assistant, and its Echo speaker devices play another. “The future of grocery shopping is most likely the ability to say, send me groceries, add garlic, remove tomatoes,” says Witcher. “Amazon has put the pieces in place to actually make it happen.” RBC Capital Markets predicts that by 2020, 128 million households will be using Alexa, driving a 10% increase in their Amazon spending.

“Once you’re a Prime customer, you don’t go anywhere else.”

Mathew Hamory of consultancy AlixPartners

Another aspect of Amazon’s end-game is what NYU’s Galloway calls a “zero-click environment” in which recurring purchases—the majority of retail—end up not just on your doorstep but in your closet or fridge. Prime members in some locations can already get packages delivered inside their homes with Amazon Key, which includes a security camera, smart lock, and app. Amazon’s April acquisition of smart doorbell company Ring makes Galloway’s vision even more likely. “It’s staggering to think about what they could do with that,” he says.

Whether this sounds utopian or dystopian is a matter of perspective. But there is no small risk to Amazon that people eventually begin to feel like the company knows too much or controls too much of their lives, that they become uncomfortable with the consequences of all of this convenience. Just ask Mark Zuckerberg, who recently testified on Capitol Hill over Facebook’s privacy and security issues.

Food could be part of that tipping point for Amazon. For the past few years, as the tech world has been making strides with artificial intelligence and machine learning, foodies have been moving in the opposite direction, pushing to deindustrialize the agricultural system. “Healthy and sustainable food is incoherent with the Amazon model,” says William Rosenzweig, who leads the Food Venture Lab program at UC–Berkeley. “I worry that the culture of Amazon doesn’t contain that gene set. Convenience, price, and speed—those are not the right values or core competencies of food.”

Can Amazon learn to walk that delicate line between convenience and care? That’s something that not even Alexa can answer yet.

[fortune-gallery id=”2274911″]

This article originally appeared in the June 1, 2018 issue of Fortune.