(This is the first part of a double issue of TheFamily Papers dedicated to better understanding Amazon and its business model. Read part 2 here.)
Amazon is one of our favorite examples when it comes to explaining TheFamily’s investment thesis. It is one of the oldest digital companies around. And contrary to most of its peers, from the beginning it operated a business that included tangible assets (operating warehouses, delivering stuff) and lots of employees. Amazon, which is located in Seattle, was reportedly despised in Silicon Valley for that reason: why would an Entrepreneur even bother founding a low-margin, difficult-to-scale retail business when they could make tons of money in the ad-clicking industry?
Today, Amazon remains one of the most fascinating tech companies out there. It also sets an inspiring example for traditional brick-and-mortar companies that are looking to become more digital. If Amazon can operate a business model that is both digital and tangible, why can’t the US Postal Service or Air France-KLM?
The problem we encounter while trying to communicate our Amazon passion is that it inspires mistrust and hostility. “That is all very interesting, but why do they treat their employees so badly?” is a frequent reaction. “When will they post a profit, anyway?” is another. From union workers to corporate CFOs, everyone has good reasons to hate Amazon or, at the very least, to refuse to draw lessons from its success. So for all of you Amazon skeptics, below are 11 notes that could perhaps lead you to reconsider.
A Retailer Turned Technology Company
1/ When the tech bubble burst in 2000, it didn’t wipe out the whole digital economy. On the contrary, as explained by Carlota Perez and William Janeway, it marked the beginning of the deployment phase that has now seen digital companies enter industries such as healthcare, car manufacturing, energy and education. But the bubble was a tough time for companies that were operating at the time, especially if they had low margins. Amazon survived by laying off workers and cutting costs even more, but it was a wake-up call. Something had to be done to hedge against those low margins and to finally become a technology company.
According to Brad Stone’s The Everything Store, Tim O’Reilly provided Bezos with the solution. “Why not,” he essentially said, “open your infrastructure to harness user participation and enroll outsiders in the process of making the company more scalable?” That marked the day when Amazon entered the digital economy: not only using digital technology to take orders and process payment, but also leveraging it to make the whole business model create more value. The company’s marketplace for third-party sellers had been launched in November 2000. Amazon Web Services was envisioned in 2002. After turning its first quarterly profit in Q4 2001, Amazon finally broke even in 2003.
Meanwhile, the company’s engineering staff had apparently been killing itself to transform Amazon’s large IT infrastructure into a collection of Web services. It was not an easy task. In his now famous “Platform Rant”, former Amazon employee Steve Yegge detailed what it took for Amazon to deploy its current IT architecture. Everybody had to work incredibly hard to deploy the new Web services while ensuring continuity of operations. Multiple unexpected pitfalls were spotted along the way. The legendary and frightening Rick Dalzell, Amazon’s CIO at the time, had to whip everybody for years to make sure that the radical IT upheaval was achieved on time at a mastered cost.
Amazon succeeded where many others would have failed and it gained a lot in the process. Forcing each team to design its own Web service imposed a virtuous pressure that incentivized everyone to improve reliability and performance. The agility derived from using Web Services instead of integrated architecture made it easier to innovate and to A/B test many new features. Finally, the internal Web services used by Amazon itself were opened to outside customers in 2006, becoming Amazon Web Services, the company’s main driver of revenue. As suggested in Jeff Bezos’s 2010 letter to Amazon’s shareholders (a most rousing text about digital technology), Amazon has fully become a technology company.
Lower Prices v. Exceptional Experience
2/ A company is a contract between many parties. The value it creates is then distributed among its various stakeholders: its shareholders of course, but also its employees, executives, customers, creditors and suppliers. The more competitive the market is, the more value has to be distributed to the customers in the form of lower prices. This is what Sam Walton understood would help him turn Wal-Mart into one of the biggest corporations in the world. For a long time, Wal-Mart’s tagline read “Always Low Prices. Always.” Sam Walton meant it: each year, suppliers had to lower their price, or else they were out. And as related by Charles Fishman’s The Wal-Mart Effect, prices went down in Wal-Mart stores even in the absence of local competition.
Jeff Bezos understood something else that didn’t exist when Sam Walton operated his first stores in Arkansas: competition is even higher on digital markets. And his conclusion was not that you should drop the prices even lower than Wal-Mart ever did: if Amazon had done that it would have disappeared just as many — among them Thomas Friedman— predicted it would. Bezos concluded that the only way to retain customers without bringing down the prices was to provide an exceptional customer experience.
Amazon has accordingly set a new standard. It has extended the experience it provides to the customer far beyond the transaction, building one of the most powerful closed ecosystems of the digital economy. You can now buy almost everything on Amazon, even from third-party sellers; you can find inspiration; you can watch movies; you can read e-books on a device that is provided by Amazon; now you can even book home services. And the more you rely on Amazon for your day-to-day life, the more difficult it becomes to even consider offers by Amazon’s competitors. Of course, you can compare prices whenever you have something to buy. But saving a few cents here and there may not be enough to compensate for the effort that it takes to go browse prices all over the Internet. In fact, Amazon proves that you can design an experience that invites customers to simply trust the company over the long term instead of constantly evaluating the competition.
This is one key to understanding Amazon. There are many detractors willing to say bad things about Amazon. But any criticism about Amazon lowering prices doesn’t stand, since Amazon is not about lower prices. Pricing the Wal-Mart way was dominated by economies of scale: the higher the volume, the lower the cost, and the price. In the digital economy, it’s different. In a business driven by increasing returns, there will always be new entrants willing to lose money on certain goods if only to monetize later or elsewhere in their business model. Replicating Wal-Mart’s strategy in the digital economy would be suicidal.
Above all, prices can evolve in real-time, and new pricing strategies can now be devised that are radically different from the ones we were used to in the Fordist economy. Operating a digital business enables one to monitor what happens in the entire business and use machine-learning algorithms to predict how money can be made later even though prices are currently low. As pointed out by David Streifeld a few years ago, prices on Amazon are a mystery and they change fast. In some cases, digital companies can even raise the prices as the volume goes up, because other incentives will inspire customers to buy anyway: those could be peer pressure and virality on social networking platforms, or… an exceptional user experience.
The Architecture of Participation
3/ It takes several steps for an industry to go digital. The first step is when startups begin to multiply and try to test new ways of doing business. The last step is when one of these startups, having formed an alliance with hundreds of millions of users, has taken over the entire industry, evicting or commoditizing every established company in the process. At every step, what the incumbent companies tend to express is denial.
When traditional chain store retailers consider Amazon, the expression of denial usually consists in declaring that Amazon and traditional retailers essentially do the same business, with only tiny differences of no importance. To be frank, we all have the same bias: when someone vaguely resembles you, you tend to concentrate your attention on similarities instead of focusing on the differences. But in the case of Amazon and traditional retailers, it really is the differences that matter.
Because of that familiarity bias, when they look at Amazon, traditional retailers only see what I call the Northern Side: operating a massive logistics infrastructure, negotiating with suppliers, delivering products to customers as mail-order businesses have been doing for a very long time. All these various businesses, combined in a traditional retailer’s business model, imply diminishing returns: the more you grow, the more you have to pay to convert and serve additional customers. As a result, the retail sector is usually dominated by oligopolies — such as the one formed in France by Carrefour, Casino, Auchan, E. Leclerc, and others. When a company reaches a larger scale of operations in such a business, its logistics infrastructure is so huge and the employees are so numerous that, as author Matt Dickinson wrote about the North Face of the Everest, “a simple slip would mean death.”
This is the reason why traditional retailers prefer to stop growing at some point and why they’re convinced that Amazon will ultimately crash when it passes that point. But what those old retail professionals are missing is Amazon’s other side, the Southern Side, on which the company has deployed Tim O’Reilly’s famous “Architecture of Participation”, thus making the most of digital technologies by harnessing the power of its users to create even more value. As O’Reilly wrote in 2004 in his masterful article “Open Source Paradigm Shift”,
Products identical to those Amazon sells are available from other vendors. Yet Amazon seems to enjoy an order-of-magnitude advantage over those other vendors. Why? Perhaps it is merely better execution, better pricing, better service, better branding. But one clear differentiator is the superior way that Amazon has leveraged its user community…
Amazon doesn’t have a natural network-effect advantage like eBay, but they’ve built one by architecting their site for user participation. Everything from user reviews, alternative product recommendations, ListMania, and the Associates program, which allows users to earn commissions for recommending books, encourages users to collaborate in enhancing the site. Amazon Web Services, introduced in 2001, take the story even further, allowing users to build alternate interfaces and specialized shopping experiences (as well as other unexpected applications) using Amazon’s data and commerce engine as a back end.
While returns are clearly diminishing on the Northern Side, the opposite trend — increasing returns — is at work on the Southern Side. For Amazon, every new warehouse costs more than the previous one, especially because it has to be located closer to the city so as to shorten delivery time (= diminishing returns). But the new customers that this warehouse will enable Amazon to serve will drive more than revenue: as they join the experience made possible by the architecture of participation on the Southern Side, they create value for Amazon through many channels: revenue, higher volumes, network effects, machine learning, and content-driven virality (= increasing returns).
Life is very different on the Southern Side as compared to the Northern Side. On the North Face of the Everest, the higher you climb, the more painful it gets. On the Southern Side, however, increasing returns radically change the rules of the game: the higher you climb, the easier it gets (and thus the faster you run up to the peak). Additionally, unlike on the Northern Side, you can slip on the Southern Side without that simple slip meaning death. Here, even failures generate valuable data that can be put to work to improve the customer experience: as Justin Fox reminds us in Harvard Business Review, “an apparent failure like the Amazon Fire phone can be treated as a learning experience rather than a crisis.”
Thus another key to understanding Amazon’s business model is that precarious balance between the Northern Side and the Southern Side: as long as returns increase more on the Southern Side than they diminish on the Northern side, Amazon can open many new, costlier warehouses and create even more value for its growing user community. I have no insight into this, but somehow I’m convinced that every night and every morning, Jeff Bezos glances at some dashboard just to check that the Southern Side weighs more than the Northern Side — which means that the company, even if it’s already so large, can continue to sustain its growth with (slightly) increasing returns. That would mean Bezos actually manages Amazon’s increasing returns, as envisioned by Michael Schrage in this Harvard Business Review article: “Effectively managing the network effects portfolio will become one of the most important challenges tomorrow’s management will confront.”
Finally, participation does not only serve the goal of improving the experience at the margins and making customers more loyal; it also empowers the users within the supply chain to take control and help make Amazon more scalable. It may even explain Amazon’s work culture. You’d think that user participation is the exclusive mark of nice companies. You’re wrong: imagine thousands of impatient, whimsical bullies storming in the factory and telling the workers to serve them better and to hurry up. This pressure from the customers, amplified by Bezos’s obsession with serving them better, explains why it apparently is so hard to work at Amazon, both on the Northern Side (= the pressured workers in the ever-more-efficient warehouses) and on the Southern Side (= the software and marketing folks constantly evaluated on their capacity to enhance user participation).
Prime At The Core
4/ Amazon seems very far away from the sharing economy. It doesn’t mean that user interactions are of a lesser importance. Quite the contrary: there are so many opportunities for Amazon users to interact with one another, from customer reviews to recommendations, not to mention the affiliate program and, of course, the marketplace. In truth, even though Amazon doesn’t stress that point in its communication, it really is as user-driven as other Web 2.0 businesses such as YouTube, Airbnb, or Wikipedia. Indeed, Amazon became a platform long before the term made its way into mainstream media.
As you can learn in politics (and by reading authors like Saul Alinsky), the key to sustaining a vibrant and growing user community on a digital platform is to cater to the power users — those who belong to the vanguard and use the product the most. Power users may not be the ones who spend the most on the platform, but their individual spending multiplies many times over because they reveal emerging trends and they talk a lot about what they do on the platform. As a result, they create value on every side of the business model: they inspire other users to spend more, and they help the company and its suppliers reach better decisions when it comes to allocating resources and setting prices.
Prime is designed for Amazon’s power users. As such, it serves many purposes. Like coupons, it enables Amazon to detect those among its customers who are really attached to having their goods delivered the next day for no additional cost. Once they’re revealed by their Prime subscription, Amazon is able to redeploy its resources and concentrate its next-day (and more and more same-day) delivery efforts on Prime subscribers, if only to keep them satisfied. Also, as the unit cost of delivery has been converted into a yearly subscription, a major source of friction has been removed and Prime subscribers can begin ordering goods on impulse, and not think twice. Hence it helps reveal what people want, almost in real time, which generates useful data flows from which Amazon can extrapolate larger market trends. Thanks to its Prime subscribers, Amazon can reach higher volumes, achieve much-needed economies of scale and increase its market power over suppliers.
This is precisely why Prime is a key component of Amazon’s business model: it was launched to detect and attract Amazon’s power users, and to nurture them. The improved features and exclusive offers could then be leveraged to make the rest of the business grow further. In a thoughtful article published last year on Wired, Marcus Wohlsen dubbed Prime “one of the most bizarre good business ideas ever,” discussed Prime economics in great detail and rightfully concluded that “Prime’s seemingly random hodgepodge is, it turns out, a finely tuned engine that drives the consumption of physical and digital goods in a seemingly unstoppable cycle optimized for the 21st-century economy.”
Those Low Margins
5/ Ah, the curse of consumer markets. Consumers are merciless, ready to deprive a company of its profit, even to force it into bankruptcy, if it’s the only way for them to obtain lower prices. This was not obvious before the advent of the Internet, because then-dominant companies defended themselves with high barriers to entry to prevent competition from bringing prices down. But in the digital economy, barriers to entry are much more difficult to erect. As a result, no company wants to be on its own facing consumers: if it does, chances are high that it will soon see its margins squeezed because of the necessity to keep on lowering prices.
Why, then, are consumer companies so dominant in the digital economy? The reason is that it is so difficult to scale up operations selling to business customers. As written in a previous issue, “enterprise markets may be a trap: FedEx sells to companies because it’s expensive (no ordinary individual can pay for it on a regular basis); and it’s expensive because its scalability is limited by the absence of user participation on an enterprise market.” This is why there are very few giant, fast-growing tech companies that are 100% old-style enterprise when it comes to selling: Palantir Technologies comes to mind as a rare exception.
To conquer enterprise markets, some enterprise companies behave as if they’re on a consumer market, even though they sell to businesses: Slack is a good example. Others, such as Google or Facebook, occupy the sweetest spots in the digital world. They have formed a strong alliance with billions of individuals, harnessing their power to sustain increasing returns, and yet they sell to business customers which are more prone to leaving a producer surplus on the table, for two reasons: businesses can cover the spending with superior earnings, and businesses are oftentimes suckers when it comes to buying—as observed by Paul Graham: “A large part of what big companies pay extra for is the cost of selling expensive things to them.”
Those enterprise exceptions are so rare, because of the difficulty to scale up on enterprise markets, that many Empire builders, among them Jeff Bezos, have to deal with the curse of consumer markets. The way to survive that curse is either to remain confined on a high-end segment (= Apple) or… to live with low margins (= Amazon). In other words, the fact that Amazon never made a profit is not because, as Matthew Yglesias once put it, “it is a charitable organization being run by elements of the investment community for the benefit of consumers.” Rather, it is because the only ways to keep on growing on a consumer market are to create an exceptional customer experience, to use superior growth as a barrier to entry, and to live on low margins.
Eugene Wei, one of the most inspiring writers on Amazon, explains it in a more figurative way:
A profitless business model is one in which it costs you $2 to make a glass of lemonade but you have to sell it for $1 a glass at your lemonade stand. But if you sell a glass of lemonade for $2 and it only costs you $1 to make it, and you decide business is so great you’re going to build a lemonade stand on every street corner in the world so you can eventually afford to move humanity into outer space or buy a newspaper in your spare time, and that requires you to invest all your profits in buying up some lemon fields and timber to set up lemonade franchises on every street corner, that sounds like a many things to me, but it doesn’t sound like a charitable organization.
Finally, margins also derive from a company’s culture, not necessarily from the industry in which it operates or the market it serves. Apple prefers high margins, and they’re willing to sacrifice market share to get them. Amazon prefers low margins. It is easy to go from high to low margins: that happens if a company gets dragged into a competition for lower prices. But it is equally easy to go from low to high margins—and it is also very dangerous, because not only does it mean that you have to sacrifice growth, it doesn’t fit in the company’s culture: thus it makes people complacent and there is no way back.
Refusing to renounce low margins even though there have been many opportunities to do so is probably one of Jeff Bezos’s most impressive achievements to date. As Eugene Wei suggested in another post, Amazon favors low margins so as to suck all the air out of the room and thus make its competitors suffocate. Hence those low margins are not a liability. Rather they are Amazon’s most potent comparative advantage.
(This is an issue of TheFamily Papers series, which is published in English on a regular basis. It covers various areas such as entrepreneurship, strategy, finance, and policy, and is authored by TheFamily’s partners as well as occasional guest writers. Thanks to Kyle Hall and Laetitia Vitaud for reviewing drafts.)
