For years, I treated Amazon’s retail business as the world’s most elaborate customer-acquisition program for AWS.
That was the accepted story: Amazon retail created scale, loyalty, Prime memberships, and an ocean of consumer data, while Amazon Web Services produced the margins Wall Street actually loved. Investors tolerated the thin economics of shipping millions of low-priced objects to millions of impatient people because AWS made the consolidated income statement look civilized.
Then something inconvenient happened to that tidy narrative.
Amazon’s retail operation started making serious money.
Not “nice little improvement” money. Not “the holiday quarter went well” money. In the second quarter of 2026, Amazon’s North America segment generated $9.1 billion in operating income, up from $7.5 billion a year earlier. The International segment added another $1.7 billion, compared with $1.5 billion in the prior-year period. Together, those two commerce-heavy segments produced $10.8 billion in quarterly operating profit.
For the full year 2025, North America delivered $29.6 billion in operating income and International produced $4.7 billion. That is $34.3 billion combined from the parts of Amazon that many investors once regarded as strategically essential but economically underwhelming.
AWS remains the profit champion. It produced $45.6 billion in 2025 operating income and another $16.6 billion in Q2 2026. I am not confusing the crown prince with the king. But Amazon retail is no longer the family member everyone politely thanks for bringing traffic to the reunion. It has built a business model that can extract profit from nearly every stage of a transaction—and often from companies that are competing for the privilege of appearing on the same screen.
Wall Street expected Amazon to dominate retail. What it did not fully expect was for domination itself to become this profitable.
The Old Amazon Retail Model: Sell Everything, Keep Almost Nothing
Amazon trained an entire generation of shoppers to expect endless selection, competitive prices, fast delivery, simple returns, and free shipping that was never truly free but felt free after the Prime fee disappeared into the fog of annual subscriptions.
That customer experience was expensive to build. Warehouses had to be constructed. Robotics had to be installed. Planes, trucks, vans, sorting centers, software, and delivery stations had to be coordinated. Inventory had to sit close enough to customers for “two-day shipping” to become “tomorrow,” then “later today,” and eventually “we may arrive before you finish ordering.”
Retail margins suffered because Amazon was not merely running a website. It was building a physical network large enough to challenge the concept of distance.
Wall Street accepted this because Amazon kept growing. Revenue expanded, Prime became embedded in household routines, third-party sellers flooded the marketplace, and AWS subsidized patience. The retail business did not need to resemble a traditional profit engine. Its strategic value was obvious even when its margins were not.
Andy Jassy’s response was not to abandon the network. It was to make the network earn its keep.
Amazon cut costs, reduced layers of management, canceled or delayed some facilities, improved inventory placement, and redesigned its U.S. fulfillment system around regional networks. The goal was simple to describe and difficult to execute: store products closer to the people likely to buy them, move packages through fewer handling steps, and deliver faster at a lower cost per unit.
The result is becoming visible. Retail sales are rising, but operating income is rising faster. That is the sound investors hear when scale finally stops demanding an allowance and starts paying rent.
The Regional Fulfillment Network Changed the Math
I consider Amazon’s regionalization effort one of the most important changes in the company’s recent history, even though it lacks the glamour of generative AI and the conversational appeal of a robotaxi.
Amazon used to move many domestic orders through a broadly national network. That approach maximized selection, but an item might travel through several facilities and across long distances before arriving at the customer’s home. Every extra mile and every additional touch cost money.
The regional model divides the country into smaller fulfillment areas and uses forecasting software to place popular inventory closer to expected demand. When Amazon accurately predicts that households in a region will soon need air filters, phone chargers, dog food, and a surprising number of inflatable lawn decorations, those products can already be nearby.
Shorter travel distances reduce transportation costs. Fewer handoffs reduce handling. Better inventory placement improves delivery speed. Faster delivery encourages customers to order more frequently, which increases package density along delivery routes. Greater density then lowers the average cost per package.
That loop matters:
Better placement leads to faster delivery. Faster delivery increases demand. More demand creates denser routes. Denser routes lower unit costs. Lower unit costs support better margins.
The beauty of the system is that customer convenience and operating efficiency reinforce each other. Amazon does not have to choose between faster delivery and lower cost in every case. Once the network reaches sufficient density, speed can become part of the cost advantage.
For years, investors saw fulfillment as a necessary expense. I now see it as infrastructure with widening economic leverage. The network was brutally expensive to build, but each efficiency improvement can spread across billions of units. Saving a small amount per package becomes meaningful when the package count resembles a national census.
Amazon Does Not Make Money From Retail in Only One Way
Calling this a retail business almost undersells what Amazon has created. A traditional retailer buys a product, marks it up, sells it, and hopes the difference covers wages, rent, shrinkage, and the customer who returned a blender after using it to process landscaping gravel.
Amazon has more ways to earn.
It sells merchandise directly. It collects commissions and service fees from third-party sellers. It charges for fulfillment and storage. It collects Prime subscription revenue. It sells advertising to brands and merchants seeking visibility. It offers payment, logistics, and other services around the transaction. One customer order can sit at the center of multiple revenue streams.
This is the crucial reason retail profitability can expand even if Amazon continues offering low product prices. The company does not need the entire profit to come from the merchandise markup. It can monetize the traffic surrounding the purchase.
Third-party sellers are especially important. When an independent merchant sells through Amazon’s marketplace, Amazon may avoid owning the inventory while still collecting marketplace and fulfillment fees. The seller takes product risk. Amazon supplies access to customers, software, trust, payment processing, and logistics. This model can be more attractive than first-party retail because Amazon participates in the transaction without funding every box of merchandise sitting on a shelf.
Then there is advertising, the profit stream hiding in plain sight at the top of the search results.
When I search Amazon for headphones, several brands are willing to pay for the chance to interrupt my decision. That ad has unusual value because I am not watching a cooking video or scrolling vacation photos. I am standing inside a digital store with purchase intent. Amazon knows what I searched, what I viewed, what I bought, and whether I am likely to return next week for replacement ear pads.
This makes Amazon’s ad inventory extremely valuable. Advertisers are not paying merely for attention; they are paying for proximity to a transaction. The retail operation creates the audience, and the advertising operation monetizes competition for that audience.
I think of it as a department store that charges brands rent for the shelves, commissions on the sale, fees to pack the item, fees to ship it, and an advertising toll to make sure customers notice it. The store may also charge the customer a yearly subscription for better access. Somewhere, an old-school merchant is staring at this arrangement and quietly reconsidering the limits of capitalism.
Prime Is Less a Membership Than a Habit-Making Machine
Once I pay the annual fee, Amazon becomes the default answer to thousands of small purchasing decisions. I do not compare delivery charges because shipping appears free. I do not wait until I have a large basket because a single low-priced item can arrive quickly. I do not always visit competing websites because convenience has already shortened the decision process.
That frequency is economically powerful. Prime increases loyalty, order volume, and the data Amazon can use to forecast demand. It supports video, music, pharmacy, grocery, reading, gaming, and other services that make the subscription harder to cancel. Even if I use only a fraction of the benefits, the bundle makes any single price comparison feel incomplete.
The subscription revenue itself is attractive because it is recurring and paid in advance. More important, Prime changes the economics of the customer relationship. Amazon receives money to make customers more likely to shop at Amazon.
Most retailers spend heavily on loyalty programs. Amazon persuaded customers to fund theirs.
Retail profitability improves when that friction disappears. A customer who visits frequently is easier to monetize through products, seller services, advertising, and subscriptions. Amazon’s advantage is not simply that it has many shoppers. It has many shoppers who begin with Amazon rather than merely end up there.
The International Business Finally Stopped Setting Money on Fire
Amazon’s International segment used to be an excellent demonstration of how quickly global expansion can consume profit.
Every new country required local warehouses, transportation relationships, technology, content, customer support, regulatory compliance, and enough selection to make the service useful. Amazon had to fund the infrastructure before demand reached efficient scale. Losses were treated as the price of building future markets, which is comforting unless one has a sentimental attachment to present-day money.
The segment has now become consistently profitable. International operating income was $3.8 billion in 2024, rose to $4.7 billion in 2025, and reached $1.7 billion in Q2 2026 alone. That progression matters because it suggests Amazon has moved beyond isolated quarterly improvements.
But profitability changes the strategic conversation. International is no longer merely a promise that requires endless investment. It can contribute earnings while preserving long-term expansion options.
For me, this is one of the clearest signs that Amazon’s operating discipline has matured. Growth is still important, but growth no longer receives diplomatic immunity from questions about returns.
The Margin Expansion Is Real—but It Needs to Be Read Correctly
Amazon reported North America sales of $426.3 billion in 2025 and operating income of $29.6 billion, implying a segment operating margin of roughly 6.9 percent. International sales were $161.9 billion with $4.7 billion in operating income, or about 2.9 percent.
In Q2 2026, North America generated $9.1 billion of operating income on $116.2 billion of sales, a margin near 7.8 percent. International produced $1.7 billion on $42.2 billion of sales, approximately 4 percent.
Those are not AWS margins. They do not need to be. Retail operates on enormous revenue. A one-percentage-point improvement across hundreds of billions of dollars can create several billion dollars of additional operating profit.
There is an important accounting nuance: Amazon’s North America and International segments are not pure retail disclosures. They include businesses associated with the stores ecosystem, and Amazon does not give investors a neat standalone income statement showing merchandise profit separate from advertising, Prime, marketplace services, and every logistics component.
Therefore, I would not claim that Amazon suddenly discovered gigantic margins in selling laundry detergent. The more accurate conclusion is that the entire retail ecosystem has become highly profitable. The low-margin transaction attracts customers. Marketplace fees, advertising, subscriptions, and logistics monetize the ecosystem. Fulfillment efficiency improves the cost base. The pieces work together.
That distinction makes the story stronger, not weaker. A retailer dependent only on product markup is vulnerable to price competition. Amazon has built layers of profit around commerce that competitors cannot easily copy without comparable traffic, seller density, logistics, data, and membership habits.
Why Wall Street Underestimated the Retail Operation
Wall Street did not fail to notice Amazon’s scale. It underestimated how much hidden operating leverage was buried inside the scale.
During Amazon’s investment phase, costs arrived before efficiencies. A new fulfillment center initially adds depreciation, staff, utilities, and complexity. As volume grows, fixed costs spread across more units. Automation improves. Routes become denser. Forecasts become more accurate. Inventory turns improve. The same network that once depressed margins can begin expanding them.
The retail business is the foundation beneath several businesses that look better on a spreadsheet.
That is why separating “retail” from “ads” can become intellectually misleading even when analysts need categories. Amazon built a commercial flywheel, not a row of unrelated revenue booths. More selection attracts more customers. More customers attract more sellers. More sellers improve selection and compete for advertising. More volume makes logistics denser. Better logistics improves speed and customer loyalty. Loyalty attracts more transactions.
The flywheel was famous long before retail margins improved. What Wall Street may not have fully appreciated was how profitable the mature version could become once Amazon stopped expanding capacity at any cost and began extracting efficiency from what it had built.
The Risks Have Not Vanished Just Because the Margins Arrived
I am impressed by Amazon’s retail transformation, but I am not willing to treat recent margin gains as a law of nature.
Competition remains intense. Walmart has a vast store network that doubles as local fulfillment infrastructure. Temu, Shein, and other platforms have conditioned shoppers to consider extremely low prices, though policy and tariff changes can alter their economics. Shopify supports merchants seeking more control over customer relationships. Grocers remain difficult because food margins are thin, delivery is costly, and customers are unforgiving about bruised produce.
Amazon also faces regulatory risk. Authorities in the United States and Europe have examined marketplace rules, Prime practices, seller treatment, advertising disclosures, and the power of large platforms. Remedies could affect how Amazon ranks products, bundles services, uses seller data, or structures fees.
Then there is capital spending. Amazon expects roughly $200 billion in capital expenditures during 2026, largely to pursue opportunities in AI, chips, robotics, and related infrastructure. Operating cash flow reached $161.4 billion for the 12 months ending June 2026, yet free cash flow turned negative by $7.6 billion because property and equipment spending surged.
That does not negate retail profitability, but it matters to shareholders. A company can produce excellent operating income while consuming enormous cash to build its next growth engine. Investors must decide whether those investments will earn attractive returns or become the world’s most expensive demonstration that enthusiasm and capital discipline are not synonyms.
Retail margins could also fluctuate as Amazon chooses to reinvest. Management may lower prices, accelerate delivery, enter new categories, or expand in countries where the network lacks scale. The company has never treated current margins as sacred when it sees a large long-term opportunity.
Anyone extrapolating a smooth upward line should remember that Amazon’s favorite use for profit is funding something ambitious enough to make the current business look conservative.
What the New Retail Economics Mean for Amazon Stock
I do not view Amazon solely as an AWS company with a giant shopping website attached. That framework now misses too much value.
AWS remains the company’s largest operating-profit contributor and a central beneficiary of AI demand. But North America and International together are generating profits large enough to stand beside major corporations on their own. Their combined $34.3 billion of operating income in 2025 represented roughly 43 percent of Amazon’s consolidated operating income.
The mix matters. AWS gives Amazon exposure to cloud infrastructure and AI. Advertising provides a high-margin stream tied to purchase intent. Marketplace services produce fee revenue without requiring Amazon to own every item. Prime strengthens recurring revenue and customer frequency. The fulfillment network creates a physical advantage that pure software competitors cannot summon with a product announcement.
This diversification deserves a better valuation conversation than “AWS is valuable and retail is there.” Retail has become an earnings engine with several internal cylinders.
The bull case is straightforward: sales keep growing, delivery density improves, automation lowers unit costs, international margins mature, advertising expands, and small margin gains produce enormous incremental profit. The bear case is equally real: valuation anticipates much of that improvement, competition forces reinvestment, regulation restricts monetization, and colossal AI spending suppresses free cash flow or earns disappointing returns.
I would never buy the stock simply because one segment had a strong quarter. I would watch whether North American margins remain durable, whether International stays profitable across seasonal cycles, whether advertising growth remains strong, whether capital spending produces revenue and efficiency gains, and whether free cash flow recovers after the infrastructure surge.
But I would also stop using Amazon’s low-margin past as the default forecast for its retail future.
The Business Wall Street Thought It Understood Has Changed
Amazon spent decades teaching investors to value growth over retail profit. It used low prices, broad selection, Prime, and relentless infrastructure investment to make itself the default store for a large portion of the internet. The strategy worked so well that its very success concealed the next stage.
Once the network existed, Amazon could regionalize it. Once customer traffic reached enormous scale, advertising became more valuable. Once sellers depended on the marketplace, services and fulfillment produced recurring fees. Once Prime became habitual, shopping frequency deepened. Once international markets matured, losses turned into operating income.
The retail business did not suddenly become a traditional retailer with better markups. It became a platform, a logistics network, an advertising marketplace, a subscription bundle, and a transaction engine wearing a retailer’s name badge.
That is what Wall Street did not expect. The assumption was that Amazon would always sacrifice retail margins to preserve price leadership and growth. Instead, efficiency and ecosystem revenue are allowing it to maintain a powerful customer proposition while producing tens of billions of dollars in operating profit.
I still believe AWS deserves enormous attention. In Q2 2026, it grew 37 percent and generated $16.6 billion in operating income. It is not being replaced as the star of the story.
But the supporting actor has learned how to steal scenes.
Amazon’s North American and International segments generated $10.8 billion in operating income during the same quarter. A business once dismissed as a thin-margin scale machine is now producing quarterly profit at a level many companies never achieve in a year.
The most interesting part is that Amazon may still have room to improve. Robotics can reduce handling costs. Better forecasting can lower inventory waste. Denser same-day networks can improve speed and economics. Advertising can keep growing with transaction volume. International operations can mature. Seller services can expand.
None of those outcomes is guaranteed, and Amazon’s investment appetite ensures the path will never be tidy. Yet the evidence is already substantial enough to retire the old assumption.
Amazon retail is not merely useful, dominant, or strategically important. It is becoming highly profitable.
For years, Wall Street waited for AWS to explain why Amazon deserved its valuation. Now the boxes arriving at the front door are beginning to make a much stronger argument of their own.
Sources
This article reflects my interpretation of publicly available information and is not personalized investment advice.
Comments
Post a Comment