Amazon: How a Bookstore Became One of the Most Powerful Companies in the World

There is a strange thing about Amazon.

Most people know it as the place where a package arrives at the front door. Type a few words into a search bar, click “Buy Now,” and somewhere in the background a remarkable machine starts moving. A warehouse finds the item. Software decides where it should come from. A worker or robot retrieves it. A delivery network gets it onto a route. Maybe an independent seller supplied it. Maybe Amazon advertised it. Maybe the software running part of that process is hosted on Amazon Web Services.

And then, a day or two later, there is a cardboard box on your doorstep.

That ordinary experience makes Amazon easy to underestimate.

The company that began selling books over the internet in 1995 is no longer simply a retailer. Today Amazon operates one of the world's largest online stores, a massive third-party marketplace, a global logistics network, a cloud-computing business, a digital advertising operation, subscription services, entertainment businesses, consumer devices, healthcare services and a growing collection of artificial-intelligence technologies.

In 2025, Amazon reported $716.9 billion in net sales, up from $638.0 billion in 2024. Operating income reached $80.0 billion and net income reached $77.7 billion. Amazon also employed approximately 1.576 million full- and part-time people at the end of the year.

Those numbers are almost too large to picture.

But the more interesting question is not how Amazon became so big.

It is why it kept becoming bigger.

The answer has surprisingly little to do with books.

It started with a bet on the internet

In 1994, Jeff Bezos was working in New York at an investment firm when he became interested in the extraordinary growth potential of the early internet. He decided to leave a comfortable career and build an internet business.

Bezos has said the decision was partly about avoiding a particular kind of regret: looking back decades later and wondering what might have happened if he had never tried.

The original idea was an online bookstore.

That sounds almost laughably modest compared with modern Amazon. But Bezos was not really interested in building an ordinary bookstore. He was interested in using the internet to create a store that could exist in a form a physical building never could.

At the time, a physical bookstore had a basic limitation: shelf space.

An enormous bookstore could stock thousands or perhaps hundreds of thousands of titles, but there was a physical limit to what could fit in a building and what a local population could support.

The internet removed that constraint.

When Amazon opened on the World Wide Web in July 1995, the company described itself as offering more than one million book titles. It pointed out that such a selection would have been impossible for a conventional store to maintain.

That was the first important Amazon idea:

The internet could make selection itself a competitive advantage.

And Bezos understood something else. People did not necessarily want a bookstore.

They wanted books.

That distinction would become enormously important.

Amazon could give customers something traditional retailers struggled to provide: an almost absurd amount of choice, available at any hour, without requiring customers to travel to a store and hope the particular book they wanted was on a shelf.

The company did not need to persuade customers that shopping online was morally superior or technologically exciting. It simply needed to make online shopping useful.

There was, however, a considerable problem.

Very few people were sure the whole thing would work.

Bezos has recalled telling his parents that he believed there was roughly a 70% chance they would lose their investment when they helped fund the young company. He has also said it took more than 50 meetings to raise $1 million from investors, during a period when many potential investors were still asking a basic question: “What’s the internet?”

Amazon was not born as an obvious future giant.

It was born as a risky experiment at exactly the moment when the commercial internet was still being figured out.

The first secret was not technology. It was customer obsession.

Amazon's most famous cultural idea is “customer obsession.”

The phrase can sound like standard corporate language. Almost every large company claims to care about customers.

But Amazon has treated the concept more like an operating system.

Its first Leadership Principle is Customer Obsession: leaders start with the customer and work backwards, focusing intensely on earning and keeping customer trust. The company currently describes 16 Leadership Principles that guide decision-making, hiring and day-to-day work.

The important part is the second half of the idea: work backwards.

Instead of starting with, “What can we build?” the question becomes, “What does the customer wish were easier?”

That way of thinking helps explain many of Amazon's strangest decisions.

Why offer millions of products rather than only the most profitable ones?

Why build massive fulfillment infrastructure?

Why introduce customer reviews that could sometimes discourage a purchase?

Why spend years building cloud infrastructure that most shoppers would never even see?

Why launch Prime and essentially train customers to expect fast delivery?

Why make it possible for independent businesses to sell alongside Amazon?

Viewed independently, these decisions can look unrelated.

Viewed through customer obsession, they start to look like pieces of the same machine.

Amazon spent decades trying to reduce friction.

Finding something should be easier.

Comparing products should be easier.

Buying should be easier.

Getting the item should be faster.

Returning it should be simpler.

Running a business should be easier.

Developing software should be easier.

Watching something should be easier.

Even discovering what you want should become easier.

The company's enormous expansion can therefore be understood as a series of attempts to remove friction from different parts of modern life.

That is much more powerful than simply selling more products.

The “Day 1” philosophy

There is another Amazon phrase that reveals a lot about the company's personality: Day 1.

Bezos used the expression in his famous 1997 shareholder letter, writing that Amazon was still at “Day 1” of the internet. The idea eventually became a broader philosophy: a huge company should continue behaving as though the market is still young, uncertain and full of opportunities.

In 1997, Amazon had served more than 1.5 million customers and generated $147.8 million in revenue. That represented 838% revenue growth over the previous year. Amazon had also expanded its employee count from 158 to 614 and dramatically increased its distribution capacity.

Those figures were remarkable at the time.

But Bezos was thinking about something much larger.

He argued that the internet could transform commerce and that Amazon should focus on creating long-term value rather than optimizing for short-term profits.

That philosophy is one reason Amazon spent years in situations where it looked less like a traditional retailer and more like a company intentionally sacrificing today's earnings to build tomorrow's infrastructure.

The gamble was that once the infrastructure existed, it would be extremely difficult for competitors to catch up.

That turned out to be one of the defining patterns of Amazon's history.

Amazon stopped being just a bookstore

Once Amazon had established itself in books, the obvious next question was:

Why stop there?

Books were an excellent starting point because they were relatively standardized. A customer could search for a book by title or author, compare editions, read reviews and order it without needing to physically inspect it.

But the underlying system was not really about books.

It was about online retail.

Amazon gradually expanded into music, electronics, toys, household goods, clothing and countless other categories. The company's original advantage—selection, convenience, search and personalization—could travel into virtually any retail category.

Then Amazon made another decision that was arguably even more important.

It invited other businesses to sell on the platform.

The decision that changed Amazon's economics

In November 2000, Amazon Marketplace began allowing independent sellers to offer products alongside Amazon's own inventory.

That transformed the company's model.

Instead of Amazon needing to purchase, store and sell every possible item itself, outside merchants could supply additional selection while Amazon provided the marketplace, customer interface and increasingly the infrastructure around the transaction.

This was more than a feature.

It was a structural change.

Imagine a traditional retailer deciding it wants to carry ten million products. It needs warehouses, purchasing systems, inventory, employees and money tied up in stock.

A marketplace can potentially obtain much of that selection from other businesses.

Amazon became not only a retailer, but a place where retailers could operate.

That distinction matters enormously today.

Amazon's 2025 filing reported $172.2 billion in revenue from third-party seller services, compared with $269.3 billion from online stores. Third-party seller services include commissions, fulfillment and shipping-related fees and other services for sellers.

Amazon itself has said independent sellers now account for more than 60% of sales in its store, and that over the first 25 years of the seller program they generated more than $2.5 trillion in sales through Amazon's store.

This created a powerful feedback loop.

More sellers meant more products.

More products made Amazon more useful.

A more useful Amazon attracted more customers.

More customers made Amazon more attractive to sellers.

More sales generated more data and justified more investment in infrastructure.

And that infrastructure made the platform more valuable.

The flywheel started spinning faster.

Then came Prime

One of the biggest changes in modern retail did not look particularly revolutionary when it launched.

In February 2005, Amazon introduced Amazon Prime for $79 a year, offering unlimited two-day shipping on eligible products.

The idea was deceptively simple.

Customers would pay in advance for a membership.

In exchange, shipping would become faster and easier.

But Prime changed customer psychology.

Once people had paid for Prime, ordering became less of a decision.

A customer did not have to think, “Do I really want to pay $8 for shipping on this?”

The shipping cost had already been psychologically absorbed into the membership.

The result was a change in shopping behavior.

Small purchases became easier to justify.

Customers had another reason to stay inside Amazon's ecosystem.

And Amazon had another reason to continue improving delivery.

Prime eventually expanded far beyond shipping. It became a subscription bundle containing entertainment and other benefits, tying together retail, video, music, reading and additional services.

Amazon says Prime now has more than 200 million members across 25 countries, although the company does not regularly disclose a precise current global membership number alongside its quarterly financial statements.

Prime's deeper achievement was not simply convincing people to pay an annual fee.

It helped redefine what shoppers considered normal.

Fast delivery stopped feeling like a luxury.

It became an expectation.

And Amazon had to build an entirely new company to meet it.

The warehouse is part of the product

People often think of Amazon as a website.

That is increasingly misleading.

A better mental model is an enormous network connecting software, inventory, warehouses, sellers, transportation, payment systems, data centers and customers.

The warehouse is therefore not merely a cost center.

It is part of the customer experience.

Amazon has invested enormous sums into making products physically closer to the people who want them. Its current fulfillment network includes several types of facilities, with technology and robotics used alongside employees.

The logic is straightforward.

If the product is 100 miles away, delivering it quickly is difficult.

If the product is 10 miles away, the problem becomes easier.

If the product is already a few miles away, same-day delivery becomes plausible.

Amazon's recent logistics strategy makes this even clearer.

In 2025, more than 13 billion items were delivered to Prime members worldwide the same or next day, according to Amazon. In the United States alone, Prime members received more than 8 billion such items, an increase of more than 30% from the previous year.

Amazon has also been moving closer to the idea of ultra-fast retail. Its delivery network now includes same-day operations, one-hour and three-hour options in selected areas, and drone deliveries in parts of the United States. In 2026, Amazon introduced Amazon Now, promising delivery of thousands of groceries and household essentials in 30 minutes or less in dozens of U.S. cities.

The company that began by making books easier to find has spent decades making physical objects easier to move.

That is a remarkable evolution.

And then Amazon built a cloud company

Perhaps the single most surprising chapter of Amazon's history began when the company started selling something that cannot be put into a cardboard box.

Amazon Web Services, or AWS, launched its first major infrastructure service, Amazon S3, in 2006.

The basic idea was to let developers rent computing infrastructure rather than build their own. Amazon could take technology developed to operate its own growing internet businesses and make it available to other companies.

This was a brilliant example of Amazon turning an internal problem into an external business.

Running a gigantic internet retailer requires enormous computing capacity.

Amazon had to solve problems involving storage, databases, networking, reliability and scaling.

Once those capabilities existed, there was no fundamental reason other companies could not use them too.

AWS eventually became one of the world's most important cloud-computing businesses.

And today, it is one of the biggest economic engines inside Amazon.

In 2025, AWS generated $128.7 billion in sales and $45.6 billion in operating income. By comparison, Amazon's North American segment generated $426.3 billion in sales and $29.6 billion in operating income.

That comparison reveals something important about Amazon's business.

The shopping operation generates enormous revenue.

But some of Amazon's less visible businesses can generate disproportionate profits.

AWS is one of them.

The company that began selling books became a major supplier of the digital infrastructure on which other companies build their businesses.

Amazon is also an advertising company

There is another Amazon business that can be easy to miss because shoppers encounter it without necessarily thinking of it as advertising.

Search for a product on Amazon and you will often see sponsored products, display placements and other paid promotions.

Amazon has one of the most valuable advertising environments in the world because it sits close to the moment when someone is actually shopping.

A person searching Google for “running shoes” may simply be researching.

A person searching Amazon for “running shoes” is much more likely to be thinking about buying running shoes.

That makes the information surrounding the search commercially valuable.

Amazon reported $68.6 billion in advertising-services revenue in 2025, up from $56.2 billion in 2024. The company says this category includes sponsored ads, display and video advertising sold to sellers, vendors, publishers, authors and others.

This is another example of the Amazon flywheel.

Customers come because there is selection.

Sellers come because customers are there.

Sellers pay for visibility because customers are there and the platform can measure shopping behavior.

That advertising revenue can then support further investment.

The product is no longer just the product.

Increasingly, the infrastructure surrounding commerce is the business.

Amazon learned to sell subscriptions, entertainment and devices

The company also expanded into areas that look even further removed from its bookstore origins.

Kindle is perhaps the clearest example.

When the original Kindle launched in November 2007, Amazon was not simply selling an electronic gadget. It was trying to solve a particular customer problem: make digital reading feel more like reading a book.

The first Kindle came with access to around 90,000 books and sold out within hours of its launch. Amazon later expanded the Kindle ecosystem across devices and apps while also developing Kindle Direct Publishing, giving authors a way to publish directly to readers.

This was a recurring Amazon strategy.

Do not merely sell someone a product.

Build the surrounding experience.

Prime Video extended Amazon's subscription ecosystem into entertainment.

Amazon Music and Audible expanded it into audio.

Devices such as Kindle, Fire TV and Echo created physical entry points into Amazon's services.

Acquisitions expanded the company's capabilities further.

Amazon acquired Zappos in 2009, gaining a major online footwear and apparel company while allowing Zappos to retain its own brand and culture.

In 2017, Amazon agreed to acquire Whole Foods Market for approximately $13.7 billion, pushing further into physical retail and groceries.

In 2018, it completed its acquisition of Ring, expanding into connected home-security products.

And in 2022, MGM joined Amazon's entertainment operations, bringing a vast library of film and television content including franchises such as James Bond, Rocky and Creed.

Amazon also moved into healthcare, completing its acquisition of One Medical in 2023. The company has since combined healthcare services with digital tools and Prime-oriented experiences.

From books to groceries to video streaming to healthcare, the individual moves can look random.

They are less random when viewed through Amazon's central question:

What is frustrating customers, and can Amazon make that experience easier?

The numbers reveal what Amazon really is

Amazon's 2025 financial statements provide perhaps the clearest picture of the modern company.

Online stores generated $269.3 billion in sales.

Third-party seller services generated $172.2 billion.

AWS generated $128.7 billion.

Advertising services generated $68.6 billion.

Subscription services generated $49.6 billion.

Physical stores generated $22.6 billion.

The rest came from other activities.

These categories overlap with the way customers experience the company, but they show something important.

Amazon is no longer adequately described as an e-commerce company.

E-commerce is still enormous, but it is only one part of a much broader ecosystem.

The company operates at several layers simultaneously:

It sells products.

It hosts sellers.

It delivers products.

It sells advertising.

It sells subscriptions.

It provides computing infrastructure.

It produces entertainment.

It operates physical stores.

It develops devices.

It offers healthcare services.

And increasingly, it is building artificial-intelligence services.

The interesting part is not merely that Amazon participates in all these industries.

It is that many of them reinforce each other.

Why AWS matters so much

AWS deserves particular attention because it illustrates how Amazon creates businesses from capabilities that originally existed for internal reasons.

A traditional retailer might build a sophisticated technology department because it needs a sophisticated technology department.

Amazon asked a different question:

Could other companies pay us for this?

The result was a business that reached $128.7 billion in annual sales in 2025 and $45.6 billion in operating income.

And AWS is not merely an old success story.

It is increasingly intertwined with the next technological wave.

Amazon's 2025 annual report repeatedly identifies artificial intelligence and machine learning as major areas of competition and investment. The company says it expects to continue making additional investments in AI initiatives.

In other words, another Amazon chapter may be following the old AWS pattern:

Build infrastructure.

Use it internally.

Improve it.

Then sell it to everyone else.

Amazon's enormous spending problem

There is a price for this strategy.

Amazon has always been willing to spend heavily on infrastructure and future growth.

In 2025, the company spent $128.3 billion in cash capital expenditures, compared with $77.7 billion in 2024. Amazon said those investments primarily reflected technology infrastructure—most of which supports AWS—and additional fulfillment capacity, with those investments expected to increase in 2026.

That is an astonishing number.

And it illustrates why comparing Amazon with a conventional retailer can be misleading.

A retailer generally tries to minimize the infrastructure required to sell its products.

Amazon often builds the infrastructure first and then tries to find more ways to use it.

That approach can depress short-term free cash flow while creating enormous long-term capacity.

In 2025, Amazon's operating cash flow reached $139.5 billion, while free cash flow was $11.2 billion, affected heavily by the increase in capital spending.

This is the same philosophy Amazon has pursued for decades: build first, monetize the resulting capabilities over time.

The risk is obvious.

Infrastructure is expensive.

Technology changes.

Customers change.

Competitors improve.

And not every big Amazon experiment becomes a winner.

The company has made plenty of mistakes

It is easy to tell Amazon's story as though every decision was brilliant.

It wasn't.

Amazon has launched products and services that failed, been criticized for strategic mistakes, struggled with businesses it entered and retreated from areas that did not develop as expected.

The company's willingness to experiment inevitably produces failures.

That is not a side effect of the Amazon model.

It is part of the model.

Amazon's culture explicitly emphasizes invention, experimentation and learning. Its leadership principles include “Invent and Simplify,” “Learn and Be Curious,” “Think Big” and “Insist on the Highest Standards.”

The important distinction is that Amazon often attempts to make failure affordable.

A small experiment can fail.

A new service can fail.

A product can fail.

But occasionally something works.

And when something works, Amazon has an extraordinary ability to scale it.

That is the real power of the company's approach.

Not perfect prediction.

Repeated experimentation combined with enormous scaling capability.

The darker side of Amazon's success

No serious story about Amazon can stop at innovation and convenience.

A company employing approximately 1.576 million people and operating such a huge physical and digital infrastructure naturally has enormous effects on workers, sellers, competitors and communities.

Amazon has faced persistent scrutiny over working conditions, labor practices, delivery arrangements, marketplace policies, competition and antitrust issues.

Its latest annual report acknowledges that Amazon is involved in litigation and investigations involving areas including labor and employment, competition and antitrust, privacy, consumer protection, data security and other matters. The company specifically notes litigation involving allegations of price fixing, monopolization and consumer protection.

These issues are not trivial footnotes.

They are a natural consequence of becoming infrastructure for commerce.

When a company becomes one of the primary places where consumers shop, where small businesses sell and where digital services run, decisions made inside that company can affect entire markets.

The same scale that produces convenience can also produce power.

And power attracts scrutiny.

That tension is likely to define Amazon's next decade as much as any new product.

The seller paradox

Amazon's relationship with independent sellers is particularly fascinating.

For a small business, Amazon can be extraordinarily useful.

Instead of building an entire e-commerce operation, attracting customers from scratch, negotiating payment systems and developing a complex fulfillment network, a seller can plug into an existing ecosystem.

Amazon can provide access to millions of customers.

But the same platform can also become indispensable.

A seller may depend on Amazon for traffic, ranking, fulfillment, advertising and sales.

That creates a complicated relationship.

Amazon is simultaneously a marketplace, a partner, a service provider, an advertising platform and, in some categories, a competing seller.

That tension is one of the central questions surrounding platform businesses everywhere.

The more useful a platform becomes, the more difficult it can be to avoid dependence on it.

Amazon's scale has made that question impossible to ignore.

The culture that ties everything together

Perhaps the most underestimated part of Amazon's success is that the company did not only build warehouses and software.

It built a distinct managerial culture.

Amazon expects employees to make decisions through a common set of principles. The Leadership Principles are used not just as inspirational slogans but in hiring, evaluation and decision-making.

One principle is “Ownership.”

Another is “Dive Deep.”

Another is “Bias for Action.”

Another is “Think Big.”

Another is “Have Backbone; Disagree and Commit.”

And sitting at the front is Customer Obsession.

Whether one admires Amazon's culture or finds parts of it demanding, its consistency is notable.

The company has attempted to create a system in which millions of employees can make decisions without every decision being made by senior executives.

That becomes critical at enormous scale.

A company with a few hundred employees can be coordinated through direct communication.

A company with more than 1.5 million employees cannot.

At some point, culture becomes infrastructure.

Amazon in 2026 is playing a different game

The Amazon of 2026 is very different from the Amazon of 1995.

The original company asked:

“How can people buy books online?”

Modern Amazon asks much larger questions.

How can products be delivered within hours?

How can a global cloud platform provide computing infrastructure to startups and governments?

How can artificial intelligence improve shopping, logistics and enterprise software?

How can Amazon make grocery shopping more convenient?

How can healthcare be accessed more easily?

How can advertisers reach customers at the moment of purchase?

How can independent merchants grow their businesses?

And perhaps most importantly:

What will customers expect five years from now that feels impossible today?

Amazon has already started testing answers.

Artificial intelligence is being integrated into logistics, shopping and healthcare. Amazon says AI helps determine where products should be stored to reduce delivery distances and improve speed. Its Health AI tools are being developed to answer questions, explain health information, help manage prescriptions and assist customers in accessing care.

Meanwhile, AWS is investing heavily in the infrastructure required for the broader AI economy.

This could become Amazon's next great transformation.

Or it could prove to be an enormously expensive race in which other technology companies win.

That uncertainty is important.

Amazon's history is not a story of inevitability.

It is a story of repeated bets.

So how did a bookstore become Amazon?

The simplest explanation is that Amazon never really thought of itself as a bookstore.

The bookstore was a wedge.

It was a product category perfectly suited to the early internet.

Books gave Bezos a place to start because the internet could solve a genuine customer problem: physical stores could never offer the same searchable selection at the same scale.

From there, Amazon kept applying the same underlying logic.

More selection.

Lower friction.

Faster delivery.

Better discovery.

More convenience.

More services.

More infrastructure.

Then more customers.

Then more sellers.

Then more data.

Then more infrastructure.

The company's famous “flywheel” was never just a diagram in a strategy presentation.

It became an operating model.

And once the flywheel was moving, Amazon could use its scale in one business to create advantages in another.

The retail business helped create logistics expertise.

Logistics made Prime more valuable.

Prime strengthened customer loyalty.

Customers attracted sellers.

Sellers expanded selection.

Selection attracted more customers.

The massive technology operation required to run all of this eventually became AWS.

The customer information generated by the marketplace helped create an advertising business.

Subscriptions connected shopping with entertainment.

Entertainment made Prime more valuable.

Physical stores expanded Amazon's presence beyond the website.

Healthcare became another area in which Amazon could attempt to reduce friction.

The company continued building around the same central idea, even as the businesses themselves became radically different.

What Amazon really sells

There is another way to understand the whole story.

Amazon appears to sell products.

But at a deeper level, Amazon sells convenience.

It sells the ability to find something without driving to a store.

It sells the ability to compare dozens of options quickly.

It sells the ability to receive something tomorrow—or sometimes within hours.

It sells the ability to run a business without building a global fulfillment operation.

It sells developers the ability to rent computing power instead of building data centers.

It sells advertisers access to consumers at an unusually valuable moment.

It sells Prime members a bundle designed to make the Amazon ecosystem increasingly useful.

Even AWS, at its core, is a convenience business.

Instead of owning enormous computing infrastructure, a company can rent it.

Again and again, Amazon has looked at a complicated, expensive or inconvenient process and asked whether it could make that process feel simpler.

That is probably the most important continuity between the bookstore of 1995 and the technology giant of today.

From “Earth's Biggest Bookstore” to something much stranger

Amazon's early ambition was enormous for its time.

But even the phrase “the world's biggest bookstore” now feels almost quaint.

Amazon reported $716.9 billion in revenue in 2025.

Its business spans retail, cloud computing, logistics, advertising, subscriptions, entertainment, devices, physical stores and healthcare.

Its infrastructure reaches around the world.

Its workforce numbers in the millions.

Its cloud business powers organizations far beyond Amazon itself.

Its marketplace is intertwined with the businesses of independent merchants.

And the company is investing billions into technologies that may shape the next stage of computing.

Yet the weirdest part of Amazon's story is that its original logic has barely changed.

In 1997, Bezos wrote that Amazon's goal was to focus relentlessly on customers, create long-term value and make bold bets while the internet was still young. Amazon still describes itself in remarkably similar terms today. Its current leadership continues to emphasize long-term thinking, customer obsession, invention and experimentation.

The website changed.

The warehouses changed.

The products changed.

The industries changed.

The scale changed beyond recognition.

But the underlying question stayed surprisingly simple:

What do customers wish were easier?

That question took Amazon from books to everything else.

And it may be the reason the company is still, in its own peculiar way, treating the present as Day 1.

The final lesson of Amazon

There is a temptation to describe Amazon as a company that won because Jeff Bezos saw the future.

That makes for a good story, but it misses something important.

Amazon did not know exactly what the future would look like.

It made a series of bets.

Some worked brilliantly.

Some worked eventually.

Some failed.

What made the difference was the company's willingness to keep investing in capabilities that could support the next opportunity.

The bookstore created the customer base.

The customer base supported broader retail.

Retail created logistical challenges.

The logistical challenges created infrastructure.

Infrastructure supported Prime.

Technology built for Amazon became AWS.

Marketplace sellers expanded selection.

Customer traffic created advertising opportunities.

Subscriptions connected separate services.

Acquisitions expanded Amazon's reach.

And now artificial intelligence is becoming the newest layer on top of the whole system.

That is why Amazon is difficult to classify.

It is a retailer, but not only a retailer.

It is a technology company, but not only a technology company.

It is a logistics company, but not only a logistics company.

It is an advertising platform, a cloud provider, a media company, a marketplace and a subscription business.

Perhaps the best description is simpler:

Amazon is a company that keeps turning one customer problem into another business.

It started with the problem of finding a book.

Then came the problem of finding almost anything.

Then getting it quickly.

Then making online selling easier for other businesses.

Then providing the computing infrastructure behind the internet economy.

Then delivering entertainment.

Then connecting physical and digital retail.

Then making healthcare more convenient.

And now, increasingly, helping people navigate a world shaped by artificial intelligence.

That is how a bookstore became one of the most powerful companies in the world.

Not because it stopped being a bookstore.

Because it kept asking what came next.