Amazon reports second-quarter 2026 results after the market closes on Thursday, July 30, and I can already predict the ritual.
The revenue number will arrive. The earnings-per-share figure will follow. Television anchors will begin speaking faster. A stock chart will twitch violently in one direction, reverse course six minutes later, and then move again when an executive uses an adjective Wall Street was not expecting.
Some investors will celebrate. Others will announce that civilization has ended. Social media will produce 10,000 confident interpretations before most people have opened the earnings release.
I will be watching the margins.
Revenue tells me how much economic activity passed through Amazon. Margins tell me how much value Amazon kept after paying the staggering cost of making that activity possible.
That distinction matters because Amazon is no longer simply an online retailer. It is a collection of businesses with radically different economics forced to share one income statement. The company sells groceries, rents computing power, delivers packages, streams television, manufactures chips, operates a marketplace, sells advertising, builds artificial-intelligence infrastructure, launches satellites, and occasionally reminds me that the household item I ordered at 11:47 p.m. can arrive before I have emotionally accepted buying it.
Treating all Amazon revenue as equal is like evaluating a restaurant by counting how many objects entered the kitchen.
A dollar of first-party retail revenue is not economically equivalent to a dollar of advertising revenue. One may require inventory, warehouse space, transportation, packaging, returns, customer service, and a driver attempting to locate an apartment that urban planners apparently designed during a fever dream. The other can be generated by placing a sponsored result in front of a shopper who was already standing inside Amazon’s digital store.
Both dollars count as revenue.
They do not carry the same weight.
That is why this earnings report is ultimately a test of Amazon’s economic architecture. I want to know whether high-margin businesses are becoming more important, whether the retail network is gaining efficiency, and whether management’s enormous artificial-intelligence investments are building future earning power or merely constructing the world’s most expensive collection of optimistic assumptions.
The Setup: Amazon Enters the Quarter From a Position of Strength
Amazon’s first quarter created a demanding comparison.
Revenue rose 17% year over year to $181.5 billion. North American sales increased 12% to $104.1 billion, International sales climbed 19% to $39.8 billion, and Amazon Web Services revenue advanced 28% to $37.6 billion.
More important to me, operating income rose from $18.4 billion to $23.9 billion. That produced a companywide operating margin of approximately 13.1%, the highest Amazon had reported.
The segment figures were equally revealing. North America generated $8.3 billion in operating income on $104.1 billion of sales, or a margin near 7.9%. International produced $1.4 billion on $39.8 billion, or about 3.6%. AWS generated $14.2 billion on $37.6 billion, giving the cloud division an operating margin of roughly 37.7%.
Those numbers come directly from Amazon’s first-quarter earnings release, and they describe a company whose profitability has changed dramatically.
Amazon also guided for second-quarter revenue between $194 billion and $199 billion, representing growth of 16% to 19%. Management projected operating income between $20 billion and $24 billion, compared with $19.2 billion in the prior-year quarter.
Wall Street’s expectations sit near the middle of the revenue range. Visible Alpha estimates cited by Investopedia call for roughly $196.75 billion in revenue, earnings of about $1.99 per share, and AWS revenue near $40.49 billion.
Those estimates matter because stocks trade on the distance between expectations and reality, not on whether a number looks impressive in isolation.
But I do not intend to spend the entire evening comparing $196.75 billion with whatever number Amazon prints. At this scale, a billion-dollar revenue difference can result from currency movements, calendar timing, product mix, or consumers briefly deciding they need another kitchen appliance that will eventually live beside the bread maker.
The margin tells me more.
If sales grow strongly while margins deteriorate, I need to know why. If margins expand while growth remains healthy, I want to understand whether the improvement is structural or temporary. If both growth and margins accelerate, I become interested. If management also converts that profit into cash, I begin sitting up straighter.
AWS Is the Profit Engine, but the Engine Is Being Rebuilt While Running
AWS will receive more attention than any other Amazon segment, and rightly so.
In the first quarter, AWS represented roughly 21% of Amazon’s total revenue but nearly 59% of its segment operating income. That is an extraordinary concentration of profitability inside a company still commonly described as an online retailer.
AWS is not simply another division. It is the economic engine that gives Amazon the capacity to invest aggressively across the rest of the empire.
When AWS grows, Amazon’s profit mix improves. When AWS margins expand, the effect on consolidated operating income can be enormous. When AWS disappoints, nobody is comforted by an unusually strong quarter for paper towels.
The second-quarter consensus of about $40.49 billion implies AWS growth above 30%. If Amazon delivers something close to that figure, it would suggest the cloud business is accelerating on an annualized revenue base approaching $160 billion.
Growth at that scale is difficult to overstate.
A small software company can double because it signed several large customers. AWS must add billions of dollars in quarterly revenue merely to move its growth rate a few percentage points. It is attempting to sprint while carrying a small country on its back.
The artificial-intelligence cycle is a major source of that growth. Businesses need computing capacity to train models, run inference, build agents, manage data, and integrate new systems into existing operations. Amazon is investing in data centers, networking equipment, power, custom Trainium chips, Graviton processors, and Nvidia graphics processors to meet that demand.
In the first quarter, management said Amazon’s custom-chip business had exceeded a $20 billion annual revenue run rate and was growing at a triple-digit percentage. Amazon also reported that Bedrock processed more tokens during the quarter than it had in all previous years combined.
Those figures sound thrilling, and perhaps they should. They also require context.
Demand alone does not determine the value of a business. The cost of satisfying that demand matters.
If AWS revenue grows 30% but Amazon must invest an increasingly large amount of capital to produce each additional dollar, the economic return may be less spectacular than the growth rate suggests. A cloud provider does not summon data centers from the atmosphere. It must acquire land, secure power, purchase expensive chips, install cooling systems, construct networking infrastructure, and hope that today’s advanced hardware does not become tomorrow’s unusually costly furniture.
The margin will give us an early reading on this tension.
AWS posted a 37.7% operating margin in the first quarter. I do not assume that figure will rise forever. Capacity investments, depreciation, energy costs, compensation, product mix, and competitive pricing can all pressure it.
If AWS growth accelerates while its margin remains in the mid-to-high 30% range, I would view that as powerful evidence that Amazon is scaling AI demand economically. It would mean the company is absorbing an enormous infrastructure expansion without surrendering the profitability of its most important segment.
If the margin declines modestly because Amazon is installing capacity against contracted or clearly visible demand, I may accept that tradeoff.
If the margin falls sharply while management speaks vaguely about long-term opportunity, I become less charitable.
I have invested long enough to know that “long-term investment” is one of the most useful phrases in corporate language. Sometimes it describes a rational decision to sacrifice current earnings for exceptional future returns. Sometimes it describes spending that management hopes nobody examines too closely until the people responsible have vested their stock awards.
My job is to tell the difference.
Retail Margins Reveal Whether Amazon’s Logistics Network Has Become a Moat
The North American retail business is less profitable than AWS, but its margin progression may reveal more about management quality.
For years, Amazon trained investors to tolerate thin retail margins because the company was building scale. That explanation was reasonable. Warehouses had to be constructed. Delivery stations had to be opened. transportation routes had to be established. Prime had to become indispensable. An enormous physical network could not appear fully optimized on the first day.
Eventually, however, scale must produce economic benefits.
Otherwise, it is merely size wearing an expensive suit.
Amazon’s North American operating margin reached approximately 7.9% in the first quarter, up from roughly 6.3% a year earlier. That is meaningful improvement in a business generating more than $100 billion in quarterly sales.
A single percentage point of margin on $100 billion of revenue is $1 billion of operating income. This is why I refuse to dismiss small movements in Amazon’s retail profitability. At Amazon’s scale, decimal points become office buildings.
The company has spent years redesigning its fulfillment network around regionalization. By placing inventory closer to customers and reducing the distance packages travel, Amazon can improve delivery speed while lowering transportation costs.
That is the rare operational change capable of making customers happier and shareholders wealthier at the same time.
Faster delivery encourages customers to use Amazon more frequently. Shorter shipping distances reduce cost. Greater order density improves route economics. Higher purchasing frequency attracts more third-party sellers. More sellers increase selection. Greater selection attracts more customers.
This is the flywheel investors have heard about for decades, except now it arrives in a van before dinner.
During the second-quarter report, I will examine whether North American sales growth continues to produce operating leverage. If revenue rises faster than fulfillment and transportation expenses, the network is becoming more efficient. If unit growth remains strong while the margin holds or improves, Amazon may be proving that convenience and profitability are no longer opposing forces.
Prime Day complicates the analysis.
Amazon assumed Prime Day would occur in the second quarter, and the event can increase sales volume while temporarily affecting product mix, promotional expense, and fulfillment costs. A margin decline caused by a surge in high-quality customer engagement may not concern me. A decline caused by worsening shipping economics would.
The earnings release will provide numbers. The conference call must provide the explanation.
I will listen carefully to how management discusses delivery speeds, inventory placement, shipping cost per unit, and the balance between first-party and third-party sales.
Third-party marketplace activity is especially important because Amazon collects fees without owning every item sold. That generally makes marketplace revenue more attractive than traditional first-party retail revenue. The same is true of subscription services and advertising.
Retail revenue growth is fine.
A more profitable retail mix is better.
Advertising Is the Margin Multiplier Hiding in Plain Sight
Amazon’s advertising business is one of the finest assets inside the company, yet it still receives less attention than AWS.
In the first quarter, advertising revenue reached approximately $17.2 billion, growing 24% year over year. Management said trailing-12-month advertising revenue had exceeded $70 billion.
That is no longer a side business. It is a global advertising platform built directly into the moment of purchase.
Traditional advertising often tries to infer intent. Amazon does not need to infer very much when a customer types “running shoes,” “coffee grinder,” or “emergency anniversary gift” into the search bar. The customer has already disclosed what they may want to buy. Sellers pay to move closer to that demand.
The economics are attractive because advertising revenue rides on infrastructure Amazon already built for commerce. The company does not separately disclose advertising operating income, but the business is widely understood to carry much higher incremental margins than first-party retail.
This creates a powerful effect on consolidated margins.
When advertising grows faster than total company revenue, Amazon’s revenue mix shifts toward a more profitable activity. The same customer visit can produce a retail transaction, a seller fee, a Prime subscription benefit, and advertising revenue.
This is monetization stacked on top of monetization.
At some point, of course, Amazon must avoid ruining the shopping experience. If every useful search result is buried beneath sponsored products with names that resemble Wi-Fi passwords, customers may begin questioning whether convenience has been replaced by a digital flea market.
There is a balance between monetizing intent and exhausting it.
For this quarter, I want advertising growth comfortably above the rate of total company growth. I also want evidence that Amazon’s broader advertising ambitions—in streaming video, live sports, connected television, and third-party properties—are expanding without compromising the value of the core shopping platform.
If advertising growth remains in the 20% range or better, it should continue supporting companywide margin expansion even if the retail business improves only gradually.
That is one reason Amazon’s consolidated margin can move higher over time without every warehouse suddenly becoming twice as efficient. The mix of the business is changing.
Investors who focus only on packages miss the invisible toll booths Amazon has built around the transaction.
International Margins Tell Me Whether Amazon Can Export Its Economics
Amazon’s International segment has historically tested investor patience.
International expansion requires local warehouses, transportation networks, employees, regulatory compliance, content, technology, and marketing. Many markets lack the purchasing density and logistics infrastructure that made Amazon’s North American model so powerful.
The result has often been admirable revenue growth accompanied by profits that seem to have missed the connecting flight.
That pattern has improved.
The International segment produced $1.4 billion of operating income in the first quarter, compared with $1.0 billion a year earlier. Its operating margin rose to approximately 3.6%.
A 3.6% margin will not cause AWS to look over its shoulder. But profitability matters because it suggests the international network is maturing.
I am not expecting every global market to reproduce American economics. Consumer behavior differs. Labor expenses differ. Population density, infrastructure, regulation, competition, and purchasing power differ.
What I want is evidence that Amazon can grow internationally without repeatedly resetting the investment clock to zero.
A durable improvement in the International margin would support the argument that Amazon’s global logistics network has reached a more productive stage. It would also diversify the company’s profit base beyond AWS and North America.
If International revenue grows at a double-digit rate on a currency-adjusted basis while remaining profitable, I will consider that a constructive result.
If margins deteriorate because Amazon is entering promising markets with disciplined investments, I will examine the opportunity.
If losses return without a clear explanation, I will ask whether Amazon is pursuing growth because the economics are attractive or because expansion has become part of its institutional identity.
Great companies can become addicted to opportunity.
There is always another geography, product category, device, service, or moonshot available. Capital allocation requires the discipline to recognize that an opportunity can be real without being worth the price.
The Companywide Margin Is Where the Businesses Meet
Amazon’s first-quarter operating margin of 13.1% was a milestone. But one quarter does not establish a permanent earning level.
The second-quarter guidance range implies a wide variety of possible outcomes. At the midpoint—$196.5 billion in sales and $22 billion in operating income—Amazon would produce an operating margin of about 11.2%.
At the low end, $20 billion of operating income on $199 billion of revenue would produce a margin near 10.1%. At the high end, $24 billion on $194 billion would produce approximately 12.4%.
That range is intentionally broad because Amazon manages several volatile businesses and continues investing heavily. It also gives management enough room to drive a delivery van through expectations.
I am less interested in whether Amazon reproduces the first quarter’s record margin than in why the margin changes.
First-quarter seasonality differs from the second quarter. Prime Day affects mix and expenses. Infrastructure investments can arrive unevenly. Stock-based compensation, depreciation, content costs, and other items can shift.
A slightly lower consolidated margin would not automatically indicate deterioration.
I would become concerned if three things happened together: AWS margin compressed materially, North American retail margin weakened without a convincing operational reason, and management raised capital-spending expectations without demonstrating stronger demand visibility.
That combination would suggest Amazon is funding an investment cycle from a profit base that is becoming less stable.
Conversely, I would view the quarter favorably if AWS growth exceeded 30% with a resilient margin, advertising remained above 20% growth, North American retail profitability held near recent levels, and management provided evidence that AI capacity was being absorbed quickly by customers.
Margins tell me the quality of current growth.
Management’s explanation tells me the likely quality of future growth.
Net Income May Be the Least Useful Headline
Amazon reported first-quarter net income of $30.3 billion, or $2.78 per diluted share.
That figure looked magnificent. It also included a $16.8 billion pre-tax gain related to Amazon’s investment in Anthropic.
This is why I do not worship earnings per share.
The gain was economically real in the accounting sense, but it did not tell me that Amazon suddenly became dramatically better at delivering packages, selling advertisements, or operating data centers during the quarter.
Investors must separate operating performance from investment revaluations, tax effects, and other non-operating items.
If Amazon reports a large EPS beat this quarter, my first question will be: What produced it?
If the answer is stronger operating income, I care.
If the answer is another investment gain, I note it but do not automatically capitalize it as recurring earning power.
Reported net income can be useful. Operating income is often more revealing for Amazon. Cash flow eventually matters most.
A sophisticated investor does not reject accounting. I simply refuse to let one accounting number perform every intellectual task.
Cash Flow Is Where the AI Debate Becomes Uncomfortable
Amazon’s first-quarter release contained one figure capable of spoiling the celebration.
Trailing-12-month operating cash flow increased 30% to $148.5 billion. Yet trailing free cash flow fell from $25.9 billion to only $1.2 billion.
The reason was a $59.3 billion year-over-year increase in purchases of property and equipment, driven primarily by artificial-intelligence investment.
This is the central tension in Amazon today.
The underlying businesses are producing extraordinary operating cash. Management is reinvesting nearly all of it—and then some—into infrastructure intended to support future growth.
That can create enormous shareholder value if the investments earn high returns.
It can also destroy value with remarkable efficiency if demand disappoints, technology changes rapidly, pricing becomes competitive, or customers capture most of the economic benefit.
Capital expenditure does not become intelligent merely because the equipment is used for artificial intelligence.
I want management to discuss the relationship between spending and demand. Are new facilities supported by customer commitments? How quickly does capacity become revenue-producing? Is Amazon constrained by supply, power availability, or customer adoption? What returns does management expect from custom silicon? How much of current spending supports AWS, and how much belongs to projects with more speculative economics?
Amazon has announced major capacity arrangements involving AI companies and is building an ecosystem around Trainium, Bedrock, and its broader cloud platform. That provides reasons for confidence.
It does not eliminate execution risk.
The most dangerous period for capital allocation is when an opportunity is both genuine and fashionable. Genuine demand gives every competitor a rational reason to invest. Fashionable enthusiasm gives management teams permission to spend before returns are fully visible.
I am not asking Amazon to protect free cash flow at the expense of a historic computing transition. That would be shortsighted.
I am asking whether each incremental dollar of investment is likely to create more than one dollar of present value for shareholders.
That is the standard.
Anything less is expensive theater.
The Human Element Behind the Margins
It is easy to discuss margins as though they emerge naturally from spreadsheets.
They do not.
A retail margin is built from millions of human decisions. A worker places inventory in the correct location. A software engineer improves routing. A driver finds a hidden entrance. A manager reduces damage. A seller chooses Amazon’s marketplace. A customer decides the convenience is worth the subscription.
AWS margins also reflect human judgment. Engineers design chips. Sales teams negotiate contracts. Customers decide whether to move workloads. Executives allocate tens of billions of dollars before demand becomes perfectly visible.
Advertising margins depend on preserving enough consumer trust that shoppers continue using the platform.
Every percentage point is an accumulation of choices.
This is why I do not evaluate Amazon as a machine. I evaluate it as an institution capable of coordinating more activity than any individual mind can fully comprehend.
That institutional ability is Amazon’s deepest advantage.
The company has repeatedly built capabilities that appeared uneconomic during development and obvious in retrospect. Fulfillment, Prime, AWS, marketplace services, and advertising all required management to tolerate criticism while investing ahead of visible profit.
That history earns management credibility.
It does not earn management immunity.
Intelligent investing requires holding two thoughts simultaneously: Amazon has demonstrated an exceptional ability to invest for the future, and even exceptional management can overestimate returns during a competitive investment boom.
Admiration without scrutiny is fandom.
Scrutiny without historical perspective is cynicism.
I prefer judgment.
My Earnings Scorecard
When Amazon reports, I will organize the quarter around several questions.
First, is AWS revenue growth at or above 30%? A result near the current consensus would confirm that AI demand is accelerating the cloud business at scale.
Second, does the AWS operating margin remain resilient? I would prefer a figure in the mid-30% range or better, although the explanation matters as much as the precise number.
Third, does North American retail maintain a healthy high-single-digit margin? Stability would suggest the regional fulfillment model and shipping efficiencies remain intact despite Prime Day volume.
Fourth, is the International segment still profitable? Continued progress would show that Amazon’s economics are becoming more portable.
Fifth, does advertising growth remain comfortably above total company growth? That mix shift is essential to long-term margin expansion.
Sixth, how much more capital does management intend to spend? More importantly, what evidence does it provide that demand and future returns justify that spending?
Seventh, does third-quarter guidance imply that operating leverage is continuing? The market may forgive heavy investment. It becomes less forgiving when heavy investment arrives with slowing growth and weaker margins.
Finally, I will listen to management’s tone. Are executives specific about capacity, customer demand, and utilization? Or do they retreat into grand language about transformation while avoiding measurable commitments?
Numbers tell me what happened.
Language often tells me how confidently management understands why it happened.
What Would Make Me More Bullish
I would become more bullish if Amazon demonstrated that its AI investment cycle is strengthening rather than diluting its economic model.
That would mean AWS growth above expectations, margins holding up despite infrastructure expansion, strong customer adoption of custom chips and Bedrock, and evidence that new capacity is monetized quickly.
I would also welcome continued retail-margin resilience. Amazon does not need its stores business to resemble a software company. It needs to prove that years of logistics investment have created durable cost advantages.
Advertising growth above 20% would further support the thesis that Amazon is steadily converting its enormous customer traffic into higher-margin revenue.
The ideal quarter is not simply a revenue beat.
It is balanced strength: cloud acceleration, retail discipline, advertising momentum, international profitability, and capital spending tied to visible demand.
That combination would suggest Amazon’s current margins are not a temporary peak. They would be evidence of a business mix capable of becoming structurally more profitable over time.
What Would Make Me Cautious
I would become cautious if AWS growth missed expectations while AWS margins contracted sharply.
That would suggest Amazon is spending heavily without receiving enough near-term revenue or is facing pricing pressure in an increasingly competitive AI market.
I would also watch for deterioration in North American margins. One weak quarter can have innocent explanations. A sustained reversal would challenge the idea that fulfillment regionalization has permanently improved retail economics.
A major increase in capital-spending plans without corresponding backlog, contract, utilization, or growth evidence would concern me.
So would excessive reliance on non-operating gains to produce an earnings beat.
The market may celebrate EPS for an evening. Eventually, investors remember that recurring value comes from recurring economics.
My Final View Before the Report
Amazon is one of the strongest businesses I have ever studied because it combines scale, infrastructure, customer trust, technological capability, and a willingness to reinvest that few companies can match.
It is also becoming more difficult to value.
The retail business is producing better margins. Advertising is changing the revenue mix. AWS is accelerating. AI is creating extraordinary demand while consuming extraordinary capital. Investment gains can distort net income. Free cash flow can vanish precisely when operating performance appears strongest.
Anyone reducing this story to whether Amazon beats quarterly EPS by seven cents is examining a cathedral through a keyhole.
I will be watching the margins because they reveal the economic character of the growth.
AWS margin tells me whether artificial-intelligence demand is profitable.
Retail margin tells me whether logistics scale has become an enduring advantage.
Advertising growth tells me whether Amazon can monetize customer intent without carrying more inventory.
International margin tells me whether the model travels.
The consolidated operating margin tells me how those pieces interact.
Free cash flow tells me what remains after ambition sends the invoice.
I do not need Amazon to maximize this quarter’s profit. A company with Amazon’s opportunities should invest aggressively when returns justify it. I would be disappointed if management sacrificed a decade of value to make one earnings call more comfortable.
But I also refuse to treat every dollar of spending as sacred simply because it arrives beneath the banner of AI.
Capital has a cost. Time has a cost. Technological obsolescence has a cost. Investors deserve evidence that tomorrow’s cash flows will compensate them for today’s sacrifice.
That is what this earnings report must begin to reveal.
The stock may rise or fall 6% after the announcement. Options pricing suggests traders are preparing for something close to that range. I cannot predict the immediate reaction, and neither can the people currently predicting it with impressive facial expressions.
What I can do is evaluate the business.
If margins remain strong, AWS accelerates, advertising continues compounding, retail efficiency holds, and management connects capital spending to visible customer demand, the long-term Amazon thesis becomes stronger regardless of what the stock does at 4:07 p.m.
If margins weaken across the board and cash consumption rises without adequate explanation, I will not allow the words “AI opportunity” to suspend arithmetic.
The market trades the headline.
I invest in the economics beneath it.
On Thursday evening, those economics will be written in the margins.
This article is an independent analysis for informational purposes and is not personalized investment advice.
Comments
Post a Comment