Skip to navigationSkip to content
STATE OF PLAY

Amazon reinvented retail, and everyone else is just trying to survive

David Huang for Quartz
By Marc Bain
Published Last updated This article is more than 2 years old.

Last July, a small group representing the giants of the tech industry gathered in the seat of US government, Washington DC. They probably didn’t want to be there.

Congress had summoned their employers—Apple, Facebook, Google, and Amazon—to answer questions about the command they hold over the markets they operate in. On Amazon’s behalf, associate general counsel Nate Sutton spoke in defense of his employer’s role in US retail. Throughout, he argued that Amazon isn’t so powerful as to be able to control prices and stifle competition. Amazon, he pointed out, makes up less than 1% of retail globally. In the US, it accounts for around 4% of retail. In fact, Walmart is much larger than Amazon, he said.

In terms of sales, Sutton is right. Walmart reported $510 billion in total sales across its US and international segments in the 2019 fiscal year, versus Amazon’s $233 billion in roughly the same period. (Amazon’s numbers include its sales of services too, but more on that later.) Other retailers, such as Target, may lag behind Amazon, but they’re still tough competitors who aren’t giving up without a fight.

On the other hand, if you look solely at US e-commerce, Amazon is an unrivaled behemoth. Currently, e-commerce only makes up about 11% of retail sales in the US, but its growth isn’t stopping. At this stage, Amazon owns an estimated 38% of online retail sales in the US, according to data from research provider eMarketer. In certain product categories, its market share approaches 50%, or far exceeds it in the case of books, music, and video. (Note: While eMarketer is a closely followed and widely cited source for e-commerce data, Quartz could not verify the accuracy of the company’s estimates because they share little about their methodology.)

But these may not even be the most telling figures about Amazon’s digital supremacy. Surveys consistently find around half of US shoppers interviewed 1 say Amazon is where they begin their search when looking for a product online. To a substantial share of Americans, the site isn’t just a portal for e-commerce, it is e-commerce.

Amazon launched in July 1995, within a year of what many consider the first legitimate internet transaction—a sale of Sting’s “Ten Summoner’s Tales” on CD by a website called Net Market. (“Attention Shoppers: The Internet Is Open,” the New York Times headline declared.) But the relationship is more than just a matter of timing. Because it was so early to e-commerce and so effective at satisfying its customers while trouncing its competition, Amazon was able to fundamentally shape what consumers would come to expect of online shopping. Concepts such as streamlined purchasing, fast and free shipping, and easy comparisons of reviews and prices across a vast selection of products are central features of e-commerce today in part because Amazon helped make them so.

That influence continues. In April, for instance, when Amazon announced it was upgrading from free two-day shipping to one-day shipping for members of Prime, its perk-loaded subscription deal, a Morgan Stanley analyst predicted it would lead consumers to eventually expect 1-day shipping across online retailers.

This sway over consumer expectations is why no company has been more disruptive to US retail in the past two decades. It’s why Amazon is part of a very small group of US businesses to recently reach a trillion-dollar valuation. And as e-commerce continues to grow, Amazon only looks set to grow with it.

But the same ingredients that make Amazon appealing to shoppers have also drawn skepticism from companies that feel their identities are lost in its focus on efficiency and its overwhelming flood of products. Shoppers might gain in choice but they lose the joy of discovering and falling in love with products. In the eyes of many retailers, Amazon has solved buying but killed shopping, as Emily Weiss, founder and CEO of beauty company Glossier, put it at the 2018 TechCrunch Disrupt conference. Meanwhile, its growing power keeps drawing the notice of regulators, and the bigger it gets, the harder it is to say whether it’s still following the consumer’s lead, or whether consumers are now just following Amazon’s.

Table of contents

The ingredients of Amazon’s retail dominance | The high costs of competing against Amazon | The relationship between Amazon and brands | For small businesses, success on Amazon can come at a price | Can Amazon be a retailer and a marketplace? | Voice, the next frontier

The ingredients of Amazon’s retail dominance

To navigate the Amazon age, it helps to understand what makes Amazon a success. Jeff Bezos, the company’s founder and CEO, wasn’t the first to figure out that customers like ease and convenience. Sears knew as much a century earlier when it was building its empire with mail-order catalogs. Nor was Bezos’s company the first to create an online marketplace where small sellers could list their goods. Ebay, for example, was on that game years before. Amazon wasn’t necessarily the first to do a lot of things.

Of course, Amazon’s early start in e-commerce gave it a critical advantage, but just as important is how it put all its features together. Amazon is an ecosystem designed to remove all the barriers, small or large, customers face when purchasing a product, and keep them shopping and spending within Amazon rather than through its rivals. That behavior is clear when looking at how often shoppers order from the site.

Research firm Rakuten Intelligence uses anonymized data from more than a million US e-commerce shoppers to compile insights on Amazon and hundreds of other merchants. It found shoppers who buy on Amazon spend less per order than they do across other retailers, but they order from Amazon twice as frequently. The numbers suggest they’re turning to Amazon for even small purchases, the kind they might once have just waited to make in a store. (One note: Rakuten Intelligence says it doesn’t have insight into how many shoppers in its data pool are Prime members.)

When we win a Golden Globe, it helps us sell more shoes, and it does that in a very direct way

Amazon’s ecosystem has proved remarkably effective, thanks to years of work and small innovations that have contributed to a “flywheel” effect, a concept articulated by professor and author Jim Collins, where the disparate gears of a system all serve to propel the system as a whole and keep it accelerating. It’s evident in the businesses of companies such as Disney, where its movies, theme parks, merchandise, and now streaming service form an integrated, self-powering universe.

Amazon only discovered how its flywheel functioned in 2001, when Collins, who was working on Good to Great—the book where he would lay out his flywheel concept—gave a presentation to its executives. Brad Stone described the outcome in his 2013 book on Amazon, The Everything Store:

Drawing on Collins’s concept of a flywheel, or self-reinforcing loop, Bezos and his lieutenants sketched their own virtuous cycle, which they believed powered their business. It went something like this: Lower prices led to more customer visits. More customers increased the volume of sales and attracted more commission-paying third-party sellers to the site. That allowed Amazon to get more out of fixed costs like the fulfillment centers and the servers needed to run the website. This greater efficiency then enabled it to lower prices further. Feed any part of this flywheel, they reasoned, and it should accelerate the loop. Amazon executives were elated; according to several members of the S Team [a small group of top executives] at the time, they felt that, after five years, they finally understood their own business.

In fact, when Amazon started selling books online in 1995 it looked nothing like it does today. “There was little science to Amazon’s earliest distribution methods,” Stone wrote. “The company held no inventory itself at first. When a customer bought a book, Amazon ordered it, the book would arrive within a few days, and Amazon would store it in the basement and then ship it off to the customer. It took Amazon a week to deliver most items to customers, and it could take several weeks or more than a month for scarcer titles.” Its main early innovations were nascent features such as its system of customer ratings and reviews.

Still, knowingly or not, Amazon had already laid the groundwork to build the flywheel that keeps it charging ahead today through a mix of customer-facing features and a shrewd strategy that amplified the network effects they created.

Ease of ordering

In 1997, the same year as its IPO, Amazon introduced “1-click” ordering,” where it would store a customer’s payment and shipping information and allow them to check out with a single click. It removed friction for shoppers, eliminated the window of time in which they might reconsider purchases while typing in their billing and shipping information, and convinced customers to allow Amazon to store their data.

Low costs

To keep prices low, it leveraged its low-cost structure as an online retailer with no stores to maintain, leaning on wholesalers to get better terms any time it could, and sacrificing margins in order to gain market share. As it built out its logistics and became more operationally efficient, it used the savings to further cut prices and keep attracting shoppers.

The marketplace

In 2000 Amazon launched its marketplace for third-party sellers. Bezos’s initial ambition was to offer Amazon shoppers anything they might want to buy online. But under the traditional retail model, where the retailer bought wholesale from companies and sold the goods at a markup, it would be impossible to realize. The marketplace, where Amazon simply acted as a conduit for other vendors, allowed it to greatly expand its selection of items, including products from top brands that don’t sell on Amazon directly.

Digital marketplaces can also create their own self-sustaining loops: “The more merchants on a marketplace, the more customers are likely to find what they are looking for; the more customers on a digital marketplace, the more likely merchants will find buyers,” research firm Euler Hermes stated in a recent report.

Fast and free shipping

Under the leadership of Jeff Wilke, Amazon’s top executive focusing on the worldwide consumer experience, it reorganized its warehouses and sped up its shipping, initially making ultra-fast delivery an option for customers who paid extra. In 2002, it introduced “Free Super Saver Shipping” on purchases above $99, quickly reducing the number to $49 and ultimately to $25. It was a preamble to other projects, including Prime.

Growth over profits

When profits were basically nonexistent for its first decade, Bezos made clear to investors that he was playing a long game, betting they would come eventually if Amazon could keep undercutting competitors and attracting more customers. “The stronger our market leadership, the more powerful our economic model,” he would write in his now well-known 2004 letter to shareholders. “Market leadership can translate directly to higher revenue, higher profitability, greater capital velocity, and correspondingly stronger returns on invested capital.”

Prime, or how it all came together

All these disparate pieces began to coalesce in 2005, when Amazon launched Prime, the main motor driving Amazon’s flywheel today. The program combined the ease of one-click ordering with two-day delivery at no extra cost. Subscribers paid a flat fee of $79 a year. The name referred to the “prime position” those items had in Amazon’s warehouses.

“This was all about getting people into a mode where they thought about Amazon first,” Vijay Ravindran, a former Amazon executive who was part of the team that built Prime, said on Recode’s “Land of the Giants” podcast about Amazon. “They bought on Amazon because it was convenient, that they knew they were getting a reasonably good deal. But they were getting out of this mentality of penny pinching.”

The year after Prime’s launch, Amazon made it more attractive by introducing a program actually aimed at its third-party vendors. Called Fulfillment by Amazon, or FBA, it allowed them to store their products in Amazon’s warehouses and let the company handle logistics such as packing and shipping for a fee. Sellers in the program were eligible for Prime shipping, which gave Prime customers even more products to choose from. And because Prime members were more likely to buy products eligible for Prime shipping, it created an incentive for third-party vendors to pay for FBA. Steadily Amazon was making an ever-larger selection of items easily available to a growing number of shoppers with a single click.

The flywheel in action

Prime didn’t take off immediately, but over the years, the benefits Amazon added have compounded one another. Video streaming, which Amazon lumped into Prime in 2011, may seem distinct from the company’s core retail business, but it’s not. Each new perk strengthens Amazon’s argument to customers to sign up for Prime, and once they’re subscribed, they tend to spend much more on Amazon than non-subscribers. “When we win a Golden Globe, it helps us sell more shoes, and it does that in a very direct way,” Bezos said at Recode’s Code Conference in 2016.

It’s a perfect illustration of the flywheel. That’s what other retailers are competing against: not just fast shipping and low prices, but a network of benefits designed to lure shoppers and keep them spending within Amazon. Today Prime members get all sorts of perks, including:

  • Free delivery that can be as fast as same-day for certain items, and one-day or two-day for millions more
  • Access to Amazon’s streaming services and their libraries of TV shows, movies, original content, and music
  • Exclusive discounts, including in Whole Foods, the upscale grocery chain Amazon bought in 2017
  • Early access to deals before non-subscribers
  • Exclusive products
  • Free two-hour delivery on groceries in select cities
  • Free access to select books and magazines
  • Free access to videos games and other bonuses for users of the streaming platform Twitch it owns

Amazon has occasionally raised the price of Prime along the way. A membership now costs $119 annually. Still, more people continue signing up. Last year, it passed 150 million Prime members globally. This past holiday season, Amazon said in a single week more than 5 million customers began a free trial of Prime or started a paid subscription. It’s a captive audience that continues to grow, all the while generating revenue. For the year ended Dec. 31, 2019, it reported $19.2 billion in sales of what it calls subscription services, which includes Prime memberships, access to content including audiobooks, digital video, digital music, and e-books, and other services.

Simultaneously, the company is bringing in ever-larger amounts of cash through its sales of other services. Beyond its subscription services, these include third-party seller fees, such as commissions and fees related to fulfillment and shipping, advertising, and sales for Amazon Web Services (AWS), its lucrative cloud-computing segment. These other businesses enable Amazon to keep investing in its core retail segment—which still makes up 57% of total sales—and fund ambitious projects such as automating its warehouses and building out its own plane and truck fleets to help it ship more and do it faster.

At this stage, Amazon has become much more than just a retailer. If the company wants to maintain its current growth, experts say it may need to start thinking beyond retail as its core business. “They’ll only see so much growth from physical-goods retail,” says Sucharita Kodali, an e-commerce and retail analyst at Forrester research firm. Sales to shoppers of products such as books, batteries, and clothes have low margins compared to services geared toward businesses and could take less priority as Amazon increases its focus on technology, or logistics, or both. The company has also expanded into areas such as lending and healthcare. Meanwhile, the flywheel just keeps spinning.

The high costs of competing against Amazon

Amazon’s rise has pressured other retailers to make sizable investments in areas such as e-commerce and logistics to stay competitive. In 2016, Walmart spent $3.3 billion to acquire e-commerce startup Jet.com, hoping to jumpstart its online sales and gain ground against Amazon. The next year, Target announced it would invest more than $7 billion over the next three years to bolster its digital business and “adapt to rapidly evolving guest preferences.”

To be fair, these investments aren’t solely in response to Amazon. E-commerce has fundamentally altered retail, changing the value proposition for physical stores and reshaping how consumers shop. It’s contributed to the decline of US department stores and the closure of numerous local mom-and-pop shops. Companies that want to stay relevant have to adjust regardless. Target’s investment, for example, included upgrading existing stores and branching into more small-format locations.

The subtext, however, has been that the changes are necessary to compete again Amazon, on terms Amazon has set.

“Amazon sets the tables for everyone in terms of what consumer expectations are and they raised the game, and we’ve all got to rise to the challenge of what the expectations are they set broadly for the market,” Josh Silverman, CEO of Etsy, an online marketplace for sellers offering handmade and vintage goods, said at a conference in May. Silverman was talking about Amazon’s ongoing push into handmade items, a move Etsy and its investors have closely monitored. A few months after his remarks, Etsy announced a plan to make it easier for its sellers to provide free shipping to customers who spent at least $35. Free shipping is now practically a necessity for retailers who want to stay competitive online.

Amazon’s speed at delivering orders has also set a standard and pushed rivals such as eBay to rethink shipping. Many shoppers trust Amazon above its competitors to get last-minute orders to them on time. During this year’s holiday rush, Amazon even temporarily blocked its third-party vendors from using FedEx’s ground network for Prime orders because it said the courier’s delivery performance needed to improve. On average, the company delivers about three days faster than its competitors, counting from the moment a shopper completes checkout to the time their purchase arrives at their door, according to Rakuten Intelligence.

To stay competitive online, several of the company’s large challengers, including big box retailers and department stores, have had to pour money into e-commerce and logistics. Some have seen their sales responding, but their margins have been pinched.

It’s the chief tradeoff in fending off Amazon: It can be done, but it will cost you—particularly when Amazon has always emphasized investment over profit. “It’s a different model you’re competing with today,” Terry Lundgren, former CEO of Macy’s, told the New York Times, noting Amazon’s growing revenue streams outside of retail, such as AWS. “It’s the reality of competing against a retailer who is purposefully reducing the price and not as concerned about making money on each item.”

The flip side of rising e-commerce, of course, has been less shopping in physical stores and malls. It’s rendered them less productive in terms of sales per square foot than they once were, and many were already too big to start. Lower productivity coupled with the need for more investment has put many established retailers, large and small, in a profit bind, contributing to the surge in retail bankruptcies.

Meanwhile, Amazon isn’t letting up. It said, for example, it was putting $800 million behind its switch to standard one-day shipping for Prime. The move dinged its profits, but it’s set to hit those of others too. “This, in our view, is likely to cause other brands, manufacturers, retailers, and logistics companies to have to invest more aggressively to compete with Amazon and its differentiated delivery,” Morgan Stanley said in a research note after Amazon’s one-day shipping announcement. “The cost to compete within e-commerce continues to rise.”

Walmart understands these costs as well as anyone. Over the past few years it has made a number of moves to close its e-commerce gap with Amazon, adding free-shipping options, introducing more brands, and bolstering its online grocery business. It recently even started a program where it delivers groceries straight into customers’ refrigerators when they aren’t home. Its e-commerce sales have grown, but so has Walmart’s spending, 2 particularly in a category it lists as “eCommerce, technology, supply chain and other.”

Amazon, meanwhile, is finally starting to see its profits take off, helped along by its service sales.

Big retailers have found other ways to regroup. Rather than fight Amazon, Kohl’s has chosen to team up. As of May, Amazon customers can now make returns at Kohl’s stores across the US. Kohl’s says the arrangement increases foot traffic to its stores and brought in younger customers.

As the costs of e-commerce keep rising, others are finding their stores can be an advantage rather than a liability—provided they’ve made them places shoppers actually want to visit. Many are using in-store experiences or services to attract shoppers, such as the F1 racing simulator at Puma’s New York flagship or the tailor shops at Levi’s locations, where shoppers can alter or customize their jeans, and are finding ways to integrate their online and offline commerce into one seamless channel. A popular option is letting shoppers buy online and pick up in store, or using stores as fulfillment centers for online orders.

“The best retailers have innovated, so that the physical store has the upper hand again,” Ron Johnson, Apple’s former store chief and one-time CEO of JC Penney, recently told Fool.com.

The relationship between Amazon and brands

Amazon’s grip on US e-commerce can make it seem like any company with a product to sell needs a presence on the site to succeed, particularly as traditional retail partners such as department stores wither. It’s not really the case; lots of companies survive without Amazon. But Amazon’s reach and logistics network do make it a uniquely attractive place to sell online in the US. A number of big labels, such as Adidas, Levi’s, and Calvin Klein, have established partnerships with the company, and Amazon has continued to attract new partners or deepen ties with existing ones, such as shoe seller Aldo.

Lately, however, there’s also a potential backlash to Amazon as a handful of high-profile companies have stopped selling on the site. In recent months those have included Nike and Ikea, which follow past companies to sever ties such as Birkenstock and PopSockets, a maker of cellphone grips. Swatch, meanwhile, never quite signed up, having reportedly backed away from a deal in 2017.

Each company had its individual issues with Amazon—counterfeits on its marketplace being one shared by multiple companies—but the departures raised the question of how much brands do or don’t need the site. There’s no clear consensus. It can depend on what kind of company it is, its goals, and how strong its connection is to its own community of shoppers.3

For businesses competing on price and convenience, but which may not prioritize establishing their own identity or lasting relationships with consumers, selling on Amazon can be ideal. There are plenty of stories of small companies making millions on Amazon selling goods they bought in bulk or at discounts elsewhere.

Amazon can also be great for a specific product to find an audience. The so-called Amazon coat—a puffer jacket made by the Chinese company Orolay—found a niche among women in New York through Amazon, springing it to broader success in the US. Orolay told Reuters last February 70% of its sales come from the US, nearly all from Amazon.

But selling on Amazon is trickier for companies for whom the brand and the storytelling around it are critical. Since the advent of mass manufacturing more than a century ago, businesses have sought ways to distinguish themselves from competitors. A “brand” crafted from different associations—visual, emotional, etc.—helped companies stand out with personalities that went beyond the function of whatever they sold. Today, in the crowded e-commerce market, it can still be what makes a successful company.

Brand builder or killer?

A number of direct-to-consumer businesses have flourished by crafting identities that differentiate their products and allow them to build followings among like-minded shoppers. Many don’t sell on Amazon, including footwear maker Allbirds, beauty company Glossier, sock maker Bombas, and luggage maker Away.

A platform like Amazon “really devolves the brands into features and benefits and lacks the storytelling and lacks the depth of what we’ve put into our products,” Allbirds’ cofounder and co-CEO, Joey Zwillinger, told Quartz’s Japanese affiliate, NewsPicks, in a recent interview. Zwillinger noted some platforms have found ways to create positive experiences, pointing to Allbirds’ partner in China, Alibaba’s Tmall.

Amazon promises an efficient shopping experience, but whatever product you’re buying, it can be the Amazon brand—convenient, methodical, transactional—that prevails. Companies seeking to make their brand a priority and create a genuine connection with shoppers haven’t always found Amazon the most hospitable environment.

“Revenue boosts make Amazon an attractive space, but in the long run, we found it’s not a suitable place to build a meaningful brand,” Sarah Ford, founder of a small company called Ranch Road Boots that was doing 50% of its sales on Amazon as a third-party seller, recently explained to Forbes. The company chose to pull its products from Amazon, also citing issues such as the high fees it paid and high rate of returns it experienced.

Amazon recognizes these issues. It created a brand registry to give its partner companies more control over their brands on the site and introduced features such as brand “stores,” which are like customizable mini-sites that allow a company to build its own branded space on Amazon. Despite past false starts getting image-conscious luxury labels to sell on the site, Amazon is also reportedly gearing up to launch a new platform for luxury labels, where the brands themselves will determine the look and feel of the shopping experience. It seems to be taking the lead of some of its big overseas competitors, such as Tmall and Germany’s Zalando, which have succeeded in luring prominent brands with customized pages on their platforms.

Yet even the brand-owned “stores” on Amazon have their issues. If you’re shopping for a pair of headphones or some shampoo and find an item from a brand you’re interested in, you have to click through to another page to get to the brand’s carefully crafted store. Traffic to those pages is “going to be significantly lower,” admitted Andy Thompson, director of marketplace services at Booyah Advertising, at a January conference on Amazon strategies for brands held by the site Digiday in New York. The advantages and challenges of branding on Amazon were a main topic of debate throughout the conference.

Is there a brand exodus afoot?

Even when well-known brands do sell directly on or to Amazon, they don’t do so indiscriminately. They may push their less-expensive mass-market items, as opposed to the more exclusive products they reserve for their own channels or specialty retailers. In 2018, Manny Chirico, CEO of PVH Corp—parent company to brands such as Tommy Hilfiger and Calvin Klein—said on an earnings call that about 70% to 80% of PVH’s business on Amazon was selling “core replenishment” items, typically meaning basics such as underwear and t-shirts.

Concerns also linger around issues on its marketplace such as unauthorized resellers and counterfeits, which Amazon recently acknowledged as a risk to its business and has said it’s working to address.

The Nike decision in particular prompted questions about whether an exodus from Amazon might be beginning. “Brands don’t need Amazon,” noted Randy Konik, an analyst at Jefferies, according to CNBC. “The move shows us that strong brands realize that traffic driven to their own site (e.g. NIKE.com) is self-sustaining, more profitable, and actually brand enhancing, while traffic and incremental revenue from Amazon.com is less profitable but also less brand enhancing.” He predicted other strong brands would avoid Amazon or reduce their presence on it in the future.

On the other hand, few companies are as powerful as Nike, which has spent years developing its own channels to sell straight to shoppers. “Nike has enormous reach and its products are in demand, so it can afford to be selective about where its products are distributed because customers will come find Nike where it is offered,” Neil Saunders, an analyst at GlobalData Retail, told Bloomberg. “I don’t think as many brands can be as selective as Nike.”

Perhaps both are right: Few can be as choosy as Nike, but Konik’s emphasis on “strong brands” was notable too. Not every company has one.

Amazon’s reach and the logistical benefits it offers to sellers will keep attracting brands. But it also seems likely many will remain skeptical about selling on the platform, particularly as other complaints continue to surface, such as the refrain that it can be difficult to get a quick response from Amazon when issues arise.

Some say the company has become domineering in its relationships. Charlie Cole, who runs e-commerce for the Samsonite luggage group, has spoken out about the issue to both the New York Times and Recode’s “Land of the Giants.” “It’s been, I’d say, for the most part a fairly carnivorous evolution on the Amazon side for the last two or three years,” he said on Recode’s podcast. He talked about Amazon preventing his company from offering lower prices elsewhere, launching private-label items that directly competed with Samsonite brands, and the mounting costs of buying ads on Amazon to stay competitive, though that’s a price many feel they have to pay.

For small businesses, success on Amazon can come at a price

Not long ago, Amazon reached an important milestone. “In 2017, for the first time in our history, more than half of the units sold on Amazon worldwide were from our third-party sellers, including small and medium-sized businesses (SMBs),” Bezos wrote in his annual letter to shareholders in 2018.

The company likes to tout how its marketplace enables these SMBs to be successful. Unlike big companies, these don’t have easy access to a global audience of customers who know and want their products. Amazon offers them a platform to reach more shoppers than they would likely be able to otherwise. But the relationship can be complicated.

Last July, two days after an Amazon lawyer appeared before Congress to downplay Amazon’s power, a third-party seller on its marketplace named Molson Hart published a post on Medium about his company’s business. “We sell plush and construction toys on Amazon,” he wrote. “Well, technically, we sell toys on our website, on eBay, on Walmart.com, to brick-and-mortar stores, and we sell on Amazon. But, really, we only sell on Amazon. In 2018, we had about $4,000,000 in sales but Amazon.com accounted for over 98% of that.”

But according to Hart’s post, doing that business cost his company $1,950,000 in payments to Amazon. Hart broke down the fees his company would pay on the sale of a $15 plush toy. Beyond the 15% commission it owed Amazon on each sale, it also had a number of other costs, including those to be part of the FBA program and ads it bought on Amazon so shoppers could actually find its toys. To sell the $15 toy required $7.84 in various fees, more than half its price.

Hart’s payments to Amazon were likely higher than average. The financial research firm Instinet has estimated that $0.27 of every dollar merchants make selling on Amazon goes back to Amazon. Other estimates put it as high as one-third of every dollar.

Amazon is by far the biggest marketplace for sellers such as Hart. Many also sell on Walmart’s marketplace, which charges only a commission roughly in line with what Amazon charges. But sellers have to apply to be on Walmart’s marketplace, it doesn’t handle tasks such as shipping, and while it’s growing fast, it’s still smaller than Amazon’s marketplace. They could always try selling wholesale to stores, of course, but unless they can get into a national chain, their sales would probably be limited. Besides, local and independent stores are struggling as e-commerce takes more business. For many, Amazon is the best option.

“There’s really no other choice,” Scott Needham, whose company sells products such as sporting goods and toys on Amazon, told the New Yorker last year. “There’s a lot of things I don’t like about Amazon, but that’s where all the customers are.” He said the total his company paid in fees to Amazon last year was about $20 million.

The fees can feel like a necessity for sellers. Those who aren’t part of FBA can find it hard to compete against those who are, in part because their rivals can beat them on shipping. And the advertising sellers buy is to get higher placement in the results that appear when a shopper searches for a related keyword. So many products flood Amazon that sellers often feel their products will be impossible to find unless they buy ads.

Most of these fees are optional, of course. No company has to be part of FBA, and paying for advertising isn’t mandatory. “The vast majority of sellers have built and run their businesses in our store without advertising,” an Amazon spokesperson said in an emailed statement. “Just like elsewhere in the retail world, businesses decide if advertising makes sense for their company, try it, and determine if it provides a good [return on investment].”

Sellers, however, have also criticized Amazon over additional issues, such as penalizing them for offering lower prices elsewhere, and charging high costs for account managers who are supposedly there to address their problems but may take months to reply. In a separate statement, an Amazon spokesperson said support is constantly available to sellers via email, phone, or chat. It resolves more than 80% of issues raised by sellers in less than 24 hours, according to the spokesperson.

Even so, as Bezos noted in his letter to shareholders, sales by third-party vendors keep rising. A growing number are based outside the US, specifically in China, where Amazon has been aggressively recruiting. Slides from a presentation Amazon officials showed at a conference in Ningbo claimed a new product was listed on Amazon from China every 1/50th of a second, the Wall Street Journal reported. These sellers, who can be hard to get information on, sell products of all sorts, from electronics to toys—some of which the Journal also found to contain unsafe levels of lead.

While most sellers on the marketplace are based in the US, these sellers in China are thriving. Many are able to offer lower prices than their US counterparts by sourcing directly from factories in the country. It’s another way the marketplace can be tough: Competition is aggressive.

Can Amazon be a retailer and a marketplace?

Since Amazon quietly stepped into the private-label business with its own batteries in 2009, the company has dramatically increased the number of products it offers under its own brands or through Amazon-exclusive brands made by others. Across these brands it now sells more than 23,000 different items, according to Marketplace Pulse, an e-commerce research firm.

Private labels, of course, are an exceedingly common phenomenon in retail, found everywhere from CVS to Costco to Kohl’s. But unlike many retailers, Amazon also operates a giant third-party marketplace, which is why regulators and others are scrutinizing its practices so closely.

Through its marketplace Amazon has access to a huge trove of data about what customers are buying, not just from large corporations but also from small third-party vendors. In theory Amazon could use this information to find top-selling products from these vendors and go straight to the manufacturers to start buying and selling their products itself. Or if the third-party vendor is succeeding with its own brands, Amazon could start making private-label knockoffs to sell at lower prices, as some companies say it has done.

It’s debatable how great a concern the situation represents. First off, Amazon insists it does not use data in this way and that its policies strictly forbid employees from doing so, though former employees have said publicly they did. But walk into any national retail chain and you’ll probably find this scenario already, with the retailer’s cheaper private-label version sitting beside the brand-name best sellers.

Amazon’s critics argue the company’s size, its ability to control what shoppers see on its site, and whom it may be undercutting with its products make the situation different. David Cicilline, a Democratic representative from Rhode Island, grilled Amazon’s Sutton over the issue during the July Congressional hearing. When Sutton pointed out that retailers frequently create their own private-label items, Cicilline shot back, “But Mr. Sutton, the difference is that Amazon is a trillion-dollar company that runs an online platform with real-time data on millions of purchases and billions in commerce and can manipulate algorithms on its platform and favor its own products.”

A chain such as Walgreens is no pipsqueak operation, but when it puts its own mouthwash on the shelf, a shopper still sees all the other options too. The concern with Amazon is that it controls the search algorithm that picks which products rise to the top. A big retail chain might also be trying to undercut a corporate giant like Johnson & Johnson with its version, not necessarily a small seller that may rely on it almost exclusively for their livelihood. In December, one of Amazon’s early employees told Recode he thought the company should have to separate its main retail business from its marketplace because of the power it already has over third-party sellers.

Despite these fears, most of Amazon’s products don’t appear to be crushing the competition. In a report last year, Marketplace Pulse said while some items, such as batteries, have apparently done well for the company, its private-label sales aren’t overwhelming. “The popular narrative has been that by utilizing internal data, Amazon can launch its brands in many categories and capture most of the category’s sales. So far there is no evidence of this working,” it said. It noted many of Amazon’s brands struggle and none of the successful products are “obviously data-driven.” Amazon’s refrain has been that private-label products make up only 1% of its total sales.

That’s not to say the situation couldn’t change. A chief concern is that Amazon has the power to promote its products over the competition in various ways. But this ability, too, has drawn scrutiny. Lately Amazon has appeared to pull back on promoting its own items. It evidently realizes any move it makes to give its own items an advantage will be noticed.

Can it be a retailer and a marketplace at the same time? For now, yes. But this may be one case where Amazon isn’t able to set all the terms as it likes.

Voice, the next frontier

One of Amazon’s future-shaping ambitions is to make voice the next vital interface for the internet and commerce. The company has been pushing its cloud-based voice assistant, Alexa, and its voice-enabled Echo devices since first introducing them in 2014. Initially these were niche products with bugs to work out. But now, Amazon believes, voice interaction is at a critical moment on its path to becoming mainstream.

“Consumers are ready to make the jump,” said Patrick Gauthier, vice president of Amazon Pay, in a presentation at the NRF’s big annual conference in January.

A survey by Amazon Pay in April reported 30% of respondents said they were likely to use voice services to purchase products in the next three years and 44% expected to use voice services at some point in a shopping journey, from discovering a product to tracking it after purchase.

The numbers may not sound groundbreaking, Gauthier admitted. But when he sees signals like these, he said, “it’s telling me something about the beginning of the adoption curve, and it’s telling me something about what we need to do to start to connect with this, because at the beginning of the adoption curve is when your customers are willing to learn with you.” Now, he pointed out, is the time when companies can shape consumer expectations of voice commerce—much as Amazon did for e-commerce generally, he could have easily added.

Gauthier isn’t alone on his bullish outlook on voice assistants. “There’s no denying that voice is the future,” consultancy PwC said in a 2018 report on voice assistants. Gauthier likened voice’s current status to the early days of apps on smartphones. At first people were skeptical about buying products through them, but today it’s common. We even do our banking through apps.

Voice commerce, not by coincidence, fits neatly into Amazon’s strategy of reducing barriers to purchase wherever possible. “The top two drivers of voice commerce are ease-of-use and speed,” Amazon’s survey stated. “We speak two times faster than we type. Through this speed, voice facilitates convenience and satiates consumers’ desire to get more by doing less. This ‘doing less’ aspect is preferred by consumers because it frees them to pursue other activities.”

An early model for how Amazon thought of voice commerce was to imagine how you would interact with a computer while doing the dishes, according to Gauthier.

One challenge it raises is how customers select the product they want to order when they’re not seeing their options laid out on a screen. It’s no problem if you know the exact brand you want, but if you tell Alexa to order, say, dish soap, it gets trickier.

In that instance, Alexa will see if you’ve previously ordered the item, and if you have, put that item in your cart. But if it’s something you haven’t previously ordered, Amazon says Alexa may recommend an “Amazon’s Choice” product. The company introduced the designation in 2015 as a way to let people shop by voice over its Echo speakers, effectively allowing Amazon to pick the product for you when there’s no order history to consult. What “Amazon’s Choice” means, however, has never been entirely clear. Amazon has refused to fully explain it, except to say it considers a number of factors, from customer ratings to whether the item is ready to ship.

There’s no indication Amazon gives preference to its own products, but how it selects items has the potential to become another topic of scrutiny, particularly as Alexa spreads across the consumer landscape. Amazon said last year it had already sold more than 100 million Alexa-integrated devices, and it continues to bring Alexa to new objects, from smart ovens to smart eyeglass frames. In its competition with other makers of voice assistants, notably Google, it has followed its usual playbook and opted for market share over profits, selling its devices at an apparent loss at times in exchange for getting them into more homes. This holiday season it was practically giving away its mini Echo Dot speakers.

At the same time, other companies have been cranking out their own Alexa-enabled products. At this year’s annual Consumer Electronics Show, or CES, in Las Vegas this January, Alexa was inescapable. “It may have been harder to find a device that didn’t have Amazon’s voice assistant built-in at the tech industry’s annual showcase,” Quartz technology reporter Amrita Khalid wrote. “From Instant Pots to motorcycle helmets to electric toothbrushes, Alexa seemed to be everywhere.”

In fact, Amazon expects these products to ultimately outnumber its own. “I think someday, third-party Alexa-enabled endpoints may be vastly larger than what we produce ourselves,” Dave Limp, Amazon’s senior vice president of devices and services, told The Verge last year. Device sales aren’t really Amazon’s goal with Alexa. Limp likened its potential to AWS, which as The Verge put it has become “the default web services platform that a million software startups use.” Beyond making Echo product, Limp’s job is mainly to build tools and APIs other developers who want to integrate Alexa can use.

Amazon’s ambition appears to be for every home to some day be filled with devices that incorporate Alexa, allowing Alexa to become a genuine ambient computing platform where a user isn’t tethered to a device at all. Anywhere you are—in your home, in your car, at your office—you could simply speak your request for Alexa to fulfill. The goal is bigger than just shopping. You could get answers to questions, and control appliances or your environment, such as the light or heat.

But shopping is certainly a key opportunity. In Amazon’s survey, it asked respondents to pick the top five reasons they would or currently do use voice commerce. The reasons most people gave were its ease and its speed, but 28% of respondents said it allowed for in-the-moment purchasing, a finding Gauthier highlighted. “I submit that, once we have really mastered this opportunity, we’ll find that in-the-moment is what is the game changer, just like location [services] is the game changer for mobile and round-the-clock [connectivity] is the game changer for the web,” he said.

Voice commerce, as Amazon sees it, isn’t meant to replace shopping from a phone or computer, but to complement them. It’s the next step in our evolution to a permanently connected life where anything we want is available in a few clicks, swipes, or now, a voice command. Consumers are shopping across a variety of channels: in stores, online, in apps. Any separation among them is diminishing. “What unites the use of all these purchase channels?” Amazon asks in the introduction to its survey. “The consumer’s desire for an effortless and frictionless experience across all channels.”

That remains Amazon’s goal: less friction; fewer steps; fewer days, minutes, and seconds between you and any product you might want, to make sure you don’t go anywhere else, and for any need you have, you turn to Amazon to meet it.