Amazon is taking over the world

By Pascal-Emmanuel Gobry on August 12th, 2013

One upon a time, a Texas teen  created a summer camp. It was called “The DREAM Institute.” It charged $600 a pop. It was all about technology and had a reading list that included Dune and The Lord of the Rings. The teenager, who had grown up attending Montessori school, was named Jeffrey P Bezos.

Such it was that a kid with a knack for standing out, starting things and thinking big became the founder of Amazon, easily the most fascinating company on the planet. Jeff Bezos recently bought the venerable Washington Post and has a space company, and he’s probably the only person on Earth of whom it could be said these aren’t even the most interesting things he’s working on.

Amazon is fascinating because it lives by its founder’s tenets: relentlessness,  complete customer focus, what he calls “an explorer’s spirit”, and, most importantly, a clear orientation towards the long term. It took Amazon eight years to eke out its first profit – and that was on purpose.

Bezos realised from the start that e-commerce would start as a land grab, and would be an economies of scale business. To be sustainably profitable, Amazon had to spend a lot of time being unprofitable. This he literally bet the company on. He won. Today, Amazon is the only company from the first generation of the web to still be the leader in its sector. Amazon is as old as the commercial web, much older than Google or Facebook. And there’s no indication that Amazon’s pace of innovation is slowing down.

It’s thanks to Jeff Bezos’ unmatched determination to think about the long term – and to actually bring his mad schemes to fruition – that it’s worth painting a portrait of Amazon in 2030. No one knows what Apple or Google or Facebook will be like in 2030. But there’s no doubt that Jeff Bezos has an idea what Amazon will be like in 2030, and that it guides his decisions today.


It’s easy to forget with all the talk about Kindles and TV shows that Amazon is first and foremost a giant online retail empire—the Walmart of the twenty-first century.

So what will it look like in 2030? We can already see inklings of it today. First and foremost, in 2030 Amazon will have warehouses and distribution centers (and lockers and self-driving trucks) everywhere in the rich world. It is already starting to build out the infrastructure to enable same-day delivery in select towns in the US. It is spending on warehouses at an alarming clip.

The implications will be vast. There is a sort of addictive quality to shopping, particularly one-click shopping. It will become even more attractive once products get to your doorstep or desk the same day. Amazon will train consumers to expect same-day delivery, which will yield tremendous competitive advantage, since no one else will have the infrastructure to deliver the same service. Once you have the infrastructure for same-day delivery, you can deliver things like groceries and other daily essentials. Which means more and more people will shift their every significant regular purchase to Amazon.

It’s also worth pondering the impact on offline retail. When you think about it, 17 years later, online commerce hasn’t actually made that much of a dent in retail. For all the talk of Borders and Radio Shack going bankrupt, malls are still around, and people still spend time in stores. E-commerce is a single-digit percentage share of total commerce. But of course it is growing quickly. And the more convenient and cheaper e-commerce gets, the more people will do it. And at some stage there will be a tipping point.

Retailers operate on thin margins, and it doesn’t take much to put them under, with their very high fixed costs for running all those stores. Mall and commercial real estate operators are typically highly indebted after buying all that real estate. E-commerce is only eating a small share of total commerce now, but it’s not just our shopping habits that will change in the future: it will also be the character of our cities and how we live. Shops won’t go away completely, just like books won’t—we’ll still need cafés and showrooms. But life will still feel very different.

Another thing worth pondering  is that once Amazon has distribution centres everywhere, and potentially even its own self-driving cars and lockers to move things around, it will probably be the biggest and most sophisticated logistics company in the world. If I were FedEx, I’d be very worried. Amazon built its tough website architecture for itself, and then began renting it out to other companies, becoming the early leader in cloud computing.

It wouldn’t be shocking if Amazon went after FedEx and became a leader in logistics. Indeed, Amazon already offers fulfillment for its third-party merchants, and its Fulfillment By Amazon service even has an API. It’s only a small step from there to having you, me and General Electric having Amazon ship things around the world for what will most certainly be a low fee.

Logistics is perhaps the most impressive of the pieces on Amazon’s chessboard, but it’s not the only one. Keep in mind Amazon’s relentless customer focus. And one of the key aspects of customer focus in the Amazon way is lower prices. Bezos says Amazon does lower prices because it tries to focus on things that will still be true decades from now: decades from now, customers will still want lower prices, more selection and more convenience.

A time-honored business practice for lowering prices is vertical integration. The way it works in retail is through private labels—retailers who make their own off-brand versions of branded products and sell them for slightly less than the branded version, whether they be diapers or detergent or something else. Given the high margins on these consumer staples (there’s a reason why Procter & Gamble is  a darling stock), retailers can sell them for less and still make a healthy profit – more than by selling another company’s merchandise.

Amazon already does this, actually: it has its own private label, Pinzon. What it doesn’t do is market it or build heavily on it. It’s still at the level of experiment. But, by 2030, I would expect Amazon’s private label brands of everything from razors to detergent to diapers to rival Unilever and Procter & Gamble for fame and market share.

Once Amazon has not only the world’s biggest retail storefront but the biggest logistics network and great manufacturing capacity, the next step will be obvious. Imagine you have a design for a new product, whether it’s a new pair of headphones or a smart watch or something else. Typically, the hardest part is finding suppliers in China who can build it for you. They can do it, but it’s a hard world to penetrate, and they might not give you a good deal, and the world of Chinese business is notoriously opaque and fraught with complexity and politics.

Then you have to figure out a way to market it. And then, almost as bad as finding a supplier is finding distribution. Big retailers won’t give shelf space to upstarts. Sure, you can sell on your own website, but that means you have to market your website in a very crowded world, and handle payment, fulfilment and so on. And being a manufacturing powerhouse is easier if you’re a leader in robotics, which Amazon is since its purchase of the warehouse robots startup Kiva Systems.

What if instead there was some sort of one-stop shop. Some sort of website, where you could not only get your designs manufactured easily, at a great price and reliably, but also shipped around the world, and with great shelf space. It would be like having the deal of your life with Foxconn, DHL and Walmart all at once. And, with a bit of clever software, the thing could be made to scale and be very user-friendly.

Expect Amazon to do this. Just as they turned computers into a service with cloud computing, they can turn manufacturing, logistics and distribution into a service too. And they will.


But Amazon is not just a retail empire. It is also a set of platforms, as Jeff Bezos highlighted in his 2011 Shareholder Letter. Amazon allows third parties to use its website to sell things, which is much more profitable for it than selling them itself, and allows it to lower prices elsewhere. It’s a better experience for customers—they have more selection, and they can buy things new or used.

But it’s also important to note what the implications for this are. We are reaching the beginning of the end of the industrial economy. Whether or not you work in a factory, your life is probably shaped by the industrial economy. If you work at an office at a set hours five days a week, that’s a pattern of work that was been set by the industrial economy in the late nineteenth century. There’s no fundamental reason for us to work or live like this.

We are going to have to invent new patterns of work and of labour organisation. It won’t happen overnight. But the internet is playing a key role in this, and Amazon within it. You might want to call it the eBay economy, but with its offer of a retail marketplace and its own fulfilment network, there’s reason to believe Amazon will win a big chunk of it.

Third parties make money from Amazon by selling on it, and also by earning affiliate fees. Amazon has already bred a sizeable economy of people who thrive on its platform. That can only grow.

In fact, Amazon not only has a marketplace for goods, it also has a marketplace for labour. Amazon Mechanical Turk is one of its most interesting services: a system of “artificial artificial intelligence”. Here’s how it works: if you have a menial task, the kind of thing that you think a computer should be able to do but actually can’t, you outsource it to Mechanical Turk. Tasks like data entry, finding listings, and so on.

On the other side of Mechanical Turk are people at computers in countries like India, who will do the tasks for you. It’s like artificial intelligence. For now Mechanical Turk only works on menial tasks, but there’s no reason it can’t move up the value chain to become a more global marketplace for labour like oDesk and eLance.

But Amazon doesn’t just have marketplaces for goods and services. It also has marketplaces for media. They are tremendously interesting and disruptive, and will be huge by 2030. And they’re not even the most interesting thing about the Kindle ecosystem.


It’s a little bit hard to wrap one’s head around what Kindle has become. It’s not just devices. It’s not just media. It’s not just commerce. It’s all of these things, and some of these things, and more.

The Kindle project is about building an ecosystem of devices, media and commerce to keep us hooked – and buying from Amazon. This is actually more ambitious than the designs of Apple and Google.

To understand what Kindle will be like in 2030, let’s start with the start: the humble ebook reader. The ebook reader is replacing the book. A technology that’s been around in its current form for centuries. It is happening. This is because – for all the protestations of people unable to see differently, ebook readers are simply a superior technology. Just like printed pages, they don’t tire your eyes, but you can can carry a practically infinite library with you in a tiny, light device.

You can expand the type (very useful for eyes that are not so young). And you can buy books instantly and start reading them instantly. And ebook readers will be around for a very long time, because they’re a different technology from smartphones and tablets and other devices. Electronic ink is a separate technology. And there are  countless voracious readers out there who care enough to buy a small cheap secondary device just for reading.

It’s quite possible that this will affect literacy in a positive way. Your writer must confess that since encountering the internet he actually read very few books – until, that is, he got a Kindle, and could buy them with one click and have them delivered instantly, and could take them with him anywhere. It has made him a much, much more enthusiastic reader.

It’s hard to believe that making book-reading more convenient and cheaper won’t lead to the same effect with many other people. This is an unfashionable view to hold, because an unspoken rule is that educated people must worship the object of the codex, but it is nonetheless happening.

When you think about it, this is a stunning achievement—successfully reinventing the book, one of the most important and momentous inventions of civilization. And that’s only the first remarkable thing about Kindle.

Because, as I mentioned, Kindle is now a family of devices. Tablets. Smartphones, soon. Wearable devices, next. By 2030 there will be an entire family of Kindle devices with associated services. Again, it’s hard to overstate what a competitive advantage this will be for Amazon, as people are using its devices thinking about what to buy and when and have it delivered almost instantly.

But Kindle is also trying to redefine media.

Kindle has redefined the book. Not just the form of the book via readers. But everything else. Yes, it is redefining its form, through Kindle Singles, novella-length works, and through subscriptions to stories with new instalments each week.

But it’s also not an exaggeration to say that Kindle will probably lead to a new renaissance of literature. On the demand side, the kindle platform really makes reading quite pleasant, attractive and convenient. And on the supply side, of course, to the great wailing and gnashing of teeth of the publishing industry, Kindle is finally allowing anyone to publish a book.

For sure, this will mean lots of dregs will wash up, as it happened with every previous opening up of the media. But, as like with every previous opening up of the media, it will lead to new creativity and new works. By 2030, this will be in full swing.

Another Amazon venture no one ever talks about is Amazon Studios. Netflix is loudly going into the original content business… but so is Amazon. But it is doing it in its own, Amazon, way. Amazon Studios lets anyone submit a script to Amazon, whether they’re in Hollywood or not. Amazon will then pick some scripts, test them and produce them.

With this, we see an inkling of Amazon doing to Hollywood what it is currently doing to the publishing business: disintermediating it and platformising it.

Amazon is doing this because it has Amazon Instant Video, its Netflix-like system. Amazon Instant Video costs many hundreds of millions of dollars per year in content rights, and yet is priced nothing—it’s extra as part of Amazon Prime, its two-day shipping service. This leaves many observers puzzled, but it’s actually brilliant.

First, witness how Amazon is gleefully disrupting itself. Half  its revenue comes from “media”: that is to say, books, CDs and DVDs. Things that soon won’t exist anymore. But instead of just simply becoming a bigger mall with broader selection to shrink its reliance on these – though it is also doing that – Amazon is also hard at work disrupting the media businesses on which it currently relies.

Second, it gets its most valuable customers – its Prime customers – hooked on Amazon’s ecosystem, by giving them more than just free shipping. Finally, it gives Amazon the option of becoming the next Hollywood.

That Instant Video is part of Prime shows how impossible it is to dissociate commerce, services and content from the Amazon bundle. Prime, Amazon’s free two-day shipping service, puzzles analysts, even though it’s great. Prime shoppers, by definition, are the heaviest shoppers. And usually, where an e-tailer makes margin is more on shipping than property. So why give it away to those who consume most, and therefore become the most costly instead of the most profitable?

First of all, Prime is probably profitable. Its economics are very hard to decipher, but Amazon probably breaks even on the things it sells through Prime, and makes money on the subscription.

But, most importantly, critics forget that Prime is one of Amazon’s best economic moats. The reason? In e-commerce, profitability often comes from the heaviest buyers. There is a long tail of infrequent buyers who give scale, but it’s the frequent and returning buyers who create the volume and margin. And by now, the most frequent, addicted online buyers in the United States are on Prime, where they are also addicted to the free two-day shipping that they can’t get anywhere else.

They’re captive to Amazon, and they’re never going to go anywhere else. And they’re paying Amazon recurring, high gross margin revenue for the privilege of their slavery. And, of course, Kindle devices come with a free trial of that addictive Prime subscription. Which includes both free shipping and video.

Again, we see the interplay of devices, services, commerce and media in Amazon’s Kindle ecosystem. You have platforms that try to disintermediate media production and turn Amazon into the platforms to distribute that media. You have devices that become the distribution channel for that media and the hook for buying into the ecosystem. You have commerce and subscriptions that ties it all together.

By 2030, when all the pieces are together, will anyone ever want to leave the Amazon ecosystem… for anything?

The Cloud

All of this would suffice to make Amazon the most interesting company in the world, and yet we haven’t even talked about one of their most interesting businesses: cloud computing.

The hype around cloud computing is unbearable. Suffice it to say that for businesses, cloud computing is important because it turns a fixed cost into a variable cost. This is why cloud computing is so financially important for businesses.

What’s  worth noting is that Amazon has brought its Amazon Way to this sector where nobody expected it to be a player, particularly in the way it is going aggressively for scale through cost cutting. Amazon keeps building data centres and keeps investing into its cloud computing services even as it keeps cutting prices to howls of pain from its competitors. And Amazon isn’t simply interested in the hobbyist, self-serve crowd: it is currently slugging it out with IBM in courts over a $600 million contract for the US Government.

The key thing about Amazon’s cloud is that by 2030 we will all have access to a supercomputer in the sky. Any of us – any research project, but also any of our applications – will be able to tap into massive stores of computing power in the cloud to run all sorts of applications. It’s impossible to know what the implications will be.

The last and final thing, the thing that no one mentions about Amazon, and yet the thing that is almost certain? Amazon is working on artificial intelligence. How can your writer know this? Well, first of all, Amazon is run by Jeff Bezos. Of course it’s working on artificial intelligence. But, second of all, Amazon is already working on AI with Mechanical Turk.

The analogy here is how Google was able to build algorithms for translation or speech recognition. The way it did it was not by building better algorithms for understanding language. The way it did was by crowdsourcing as many translations and human voices as possible, and then using big data techniques and correlations to build a system that could come up with translations or speech recognition.

This is why, in the era before smartphones, Google heavily promoted its call-to-search service. Not because it thought people phoning Google to get search results would be a big business. But because it needed lots and lots of people to speak lots and lots of short phrases into phones so that it could train computers to understand them.

Meanwhile, Amazon has the biggest laboratory of artificial intelligence in the world with Mechanical Turk: actual humans behaving the way an AI would behave, solving the kinds of discrete tasks we wish computers could do. It’s simply impossible that Amazon doesn’t have a team trying to extract AI algorithms from all the data Mechanical Turk generates.

So, this is what Amazon will look like in 2030. The world’s biggest retail and logistics empire, and also a manufacturing powerhouse and the owner of leading consumer brands. An ecosystem of devices, software, services and commerce that follows us everywhere. A set of platforms disintermediating and redefining media. A giant supercomputer in the sky anyone can use. And perhaps even AI.

And people say Apple is innovative.