Amazon is undoubtedly one of the most consistent companies when it comes to operational excellence. Over the last 20 years, the development of its stock price outperformed the S&P 500 by nearly 200 times.
Such a development and market position, as Amazon enjoys today, does not come out of nowhere. It requires clear competitive advantages and barriers that keep competitors away.
But competitors can build fulfillment centers and launch marketplaces.
So how does Amazon manage to protect its market position?
That is what I want to get to the bottom of in this article.
1. The Moat at a Glance
Rating: Wide Moat
Trend: Stable Development
Identified Powers: Scale Economies, Network Economies, Switching Costs, Process Power
Strongest Power: Scale Economies
Biggest Threat: Regulatory Developments
Core Thesis: Amazon combines several Powers across its individual segments, which merge into a wide moat that is difficult to overcome in multiple ways.
New here? Make sure to subscribe to Antares Research and get articles like this delivered straight to your inbox.
2. The Sources of Power
Most people know Amazon primarily for its Commerce business. But Amazon has long been much more than that. What is probably the most promising segment today, AWS, is not even on most people's radar. It would therefore be wrong to look at Amazon only as one whole company in the following moat analysis. Each segment has different and partly unique moats. Mixing them together would be misleading.
Power #1: Scale Economies
This Power applies to several segments at once.
Commerce
However, this Power is probably most pronounced in the Consumer business. Scale Economies emerge as a company grows. Through higher sales and production volumes, fixed costs can be spread across a larger number of units.
The Benefit for Amazon becomes clear: As order volume increases, utilization of fulfillment, sorting and delivery infrastructure increases. The result is a broader distribution of fixed costs across a larger number of units, which reduces costs. Amazon also benefits from higher margins when costs decline while sales volumes increase. The resulting higher profits and cash flows can then be used to finance new investment projects.
Amazon mastered this. The marketplace has generated high order and revenue volumes for many years. Even a small improvement in the cost structure can lead to significantly higher profits. The company even describes these scale effects itself in its reports. Amazon aims to continuously reduce shipping costs through higher sales volumes, optimization of the fulfillment network and better terms with suppliers. All of these aspects are a result of Scale Economies.
But by now, these Scale Economies go far beyond simply spreading fixed costs. Amazon operates more than one million robots in its fulfillment centers. According to Jassy, this creates an additional advantage through the feedback loop.
My friend Gianni wrote a separate article about this. You can read more about the topic here:
Scale Economies not only lead to cost reductions, but also to better service quality. In the U.S., there are more than 85 specialized Same-Day Fulfillment Centers, which had delivered more than 500 million Same-Day units as of the beginning of 2026.
The relationship can be illustrated as follows: More volume → more local fulfillment centers → more products can be stored closer to customers → faster delivery with better utilization.
Competitors have long noticed the Benefits of the Commerce business as well. The prospect of an attractive business naturally means that other companies also try to enter the market in order to benefit from attractive returns.
For Amazon, it is therefore particularly important to protect this moat.
The Barrier that emerges here and keeps competitors away is the sheer volume. Of course, a competitor can build warehouses, hire drivers or install robots. But at Amazon’s size, correspondingly high investments are required to achieve similar efficiency and offer customers a similar level of value.
For most competitors, especially small to medium-sized ones, this is a battle they cannot win. They would try to undercut Amazon on price, at least temporarily. But Amazon’s pricing power allows it to temporarily lower prices as well without threatening the existence of the business. For competitors, it is simply economically unattractive to make such investments. Instead of creating value, they would inevitably destroy it through this process.
Only a few very large providers with nearly similar Scale Economies have managed to take market share from Amazon. Walmart is one example. Chinese providers such as TikTok Shop or Temu/Shein are also increasingly entering the market.
Amazon’s fulfillment network remains unique for now and continues to enjoy significant competitive advantages. Third-Party Sellers also use the same infrastructure, making the network increasingly stronger.
AWS
At AWS, Scale Economies look somewhat different. The Cloud business is highly capital-intensive and therefore requires enormous investments before a data center can even begin operating.
Typically, the lead time for investments in land, buildings, power, chips and networking is around 6–24 months. Only after this period can the infrastructure be monetized.
Amazon’s Custom Silicon business, meaning its own portfolio of chips for Cloud and AI, is a major driver of AWS’s Scale Economies.
Management expects that, once Trainium chips used for AI training reach sufficient scale, they could save double-digit billions in CapEx annually and create an operating margin advantage of several hundred basis points compared with relying exclusively on third-party chips. The Benefit for Amazon becomes very clear here.
The Barrier comes from the capital intensity mentioned above. To build a similar business, a competitor would have to pre-finance billions in infrastructure, provide global capacity and then utilize that infrastructure sufficiently to compete with the cost base of a hyperscaler.
That is why there are only a few worthy competitors. The Cloud business is largely dominated by three companies: Amazon or AWS, Microsoft and Google. Smaller providers have almost no chance of breaking into the market and catching up with these Scale Economies.
This also makes something else clear: Only the massive Scale Economies of Amazon’s Commerce business made it possible for Amazon to launch AWS in the first place.
I therefore consider Scale Economies Amazon’s strongest individual Power.
Power #2: Network Economies
Here, the focus is once again on the Commerce business. And particularly on the marketplace that brings buyers and sellers together. Third-Party Sellers play a central role here. Their share of all units sold through the platform is above 60%.
This mechanism of bringing buyers and sellers together creates Network Economies.
But what exactly are network effects?
They describe how the value of a service or offering for customers increases with every new customer joining the network. Network Economies are similar to Scale Economies in the sense that the provider with the most customers wins. It is therefore about scaling as quickly as possible.
For Amazon specifically, this means: The more buyers use the platform, the more attractive the platform becomes for merchants because they can reach more potential demand there.
More merchants on the platform, in turn, mean: more product selection, more variants, stronger price competition and better product availability. Amazon becomes increasingly valuable for buyers because they directly benefit from this.
This creates a real flywheel:
More buyers → Amazon becomes more attractive to sellers → more sellers → greater product selection, lower prices, better product availability → better shopping experience → more buyers → Restart
The Benefit for Amazon lies in the operating leverage. Third-party sellers partly store their products themselves. Amazon therefore does not have to grow its own warehouse capacity proportionally with the higher revenue generated by a larger network. This leads to higher profit margins while the platform becomes increasingly valuable as selection grows.
Another advantage is that Amazon can increasingly distribute its own products such as Alexa, Kindle, Fire TV, Fire Tablet, Ring or Amazon Essentials and serve more and more product categories over time. Additional services such as Prime Video or Music are also monetized, which helps keep users within the ecosystem.
The Barrier emerges from the classic marketplace problem: A new marketplace launched by a competitor needs buyers in order to become attractive to sellers. But at the same time, it needs sellers to create enough product selection to attract buyers.
Amazon already possesses both sides at a very large scale. Since buyers do not ask which platform has the nicest seller software, but where sellers can reach the most customers, Amazon has an advantage over its competition.
In addition to a broad product selection, the ecosystem provides offerings that help keep customers inside the network. These include technology products such as the Fire TV Stick or Amazon Alexa. Even more important, however, are the Prime membership program and access to Prime Video and Prime Music. Prime provides free and fast shipping, which is a decisive criterion for consumers when shopping online. Its introduction helped Amazon establish itself as the market leader early on. The additional streaming offerings create enormous value for customers at a comparatively low price. The Benefit for customers quickly becomes obvious from the shipping savings alone. For this reason, a large share of customers remains tied to Amazon’s ecosystem.
For competitors, it is extremely difficult to offer similar value. To convince customers to switch away from this ecosystem, some would probably demand some form of compensation. At the very least, competitors would have to spend significant financial resources to deliver similar value to Amazon. As with Scale Economies, this creates an unattractive cost/benefit of gaining share for competitors, which can become so high that trying to win users for their ecosystem simply does not make economic sense.
Amazon’s Network Power does have one weakness: More than 95% of independent third-party sellers also sell their products through other retail channels. Amazon therefore does not have a closed network effect.
However, this does not destroy the network effect, because merchants do not have to leave Amazon in order to offer their products elsewhere, and Amazon does not need to own these sellers exclusively.
As long as the additional demand Amazon provides to a merchant is large enough, it remains economically attractive to be present on the platform.
Amazon’s Network Economies are therefore very strong, even though this is not a “winner-take-all” situation.
Power #3: Switching Costs
Although Amazon can keep customers within its ecosystem through offerings such as Prime, customers can generally order from another retailer such as Walmart without facing meaningful Switching Costs.
At AWS, however, the situation looks completely different.
A company that builds its entire IT infrastructure on AWS develops a high level of dependency on the service through stored data, databases, cloud-native services, APIs, networking, internal processes and so on.
A company will not simply switch to Microsoft Azure without hesitation just because a slightly better offer appears. Many companies commit to a Cloud provider only once.
A switch would involve high costs and a high degree of uncertainty. It must be ensured that all stored data, applications and so on are completely transferred to the new provider. Such a process can take weeks and interrupt the company’s operations, naturally resulting in revenue and profit losses. Very few companies would accept this uncertainty and would instead rather pay a higher price.
Switching Costs become even greater when a company does not only use computing power but also cloud-native applications such as monitoring or observability software. Employees are often specifically trained to use such software, and their work is partly structured around it. This creates another dependency that increases Switching Costs.
The Benefit for Amazon is that it can charge higher prices for its services without customers immediately considering a switch. Amazon also benefits from recurring and highly predictable revenue. High customer retention allows AWS to grow alongside its customers and gives it a structural advantage when selling additional services and workloads.
A particularly strong piece of independent evidence for these Switching Costs comes from a Cloud investigation by the British Competition and Markets Authority, which considers the market power of AWS and Microsoft in the UK Cloud market to be substantial.
Developments around AI strengthen this effect even further, because according to Amazon, companies increasingly want to run their AI inference close to where their existing workloads are located, partly because of latency. Since some customer workloads are already hosted on AWS, existing infrastructure can pull additional AI demand into the same environment. Cloud providers with multiple locations across different regions have a strategic advantage here. Amazon continuously expands its capacity across various regions, both nationally and internationally, to meet this demand.
The Barrier again comes from the unattractive cost/benefit of gaining share for potential competitors.
To convince a customer to switch, a competitor must compensate them for the switching costs. These include migration costs, engineering effort, retraining employees and potential downtime.
The more complex and AWS-specific the existing infrastructure is, the greater this hurdle becomes.
When a customer exits the AWS Cloud, there are not only indirect costs such as employee retraining or the cost of company downtime. AWS also charges so-called Egress Fees when customers leave. These have increasingly become a concern for the British CMA, and there are efforts to reduce these fees.
Much more interesting, however, are not the direct costs of switching, but the indirect costs and the technical and organizational integration of a Cloud provider into a company’s stack.
I therefore consider this Power to be very strong and important as well. Its exact strength, however, depends on the depth of workload integration. The more services a company uses from AWS, the higher the switching costs tend to become.
Power #4: Process Power
Prime’s promise is easy to understand: Free package delivery with short delivery times. Sometimes, you receive your order the very next day.
The fulfillment system behind it, which makes such deliveries possible in the first place, is probably not quite as easy to understand. It is the result of years of experience and improvement. Such a system includes numerous steps from order to delivery and can only function through the interaction of many individual components, programs and algorithms. This also includes robotics systems used to accelerate processes or AI applications that create forecasts about where certain products are likely to experience the highest demand. These applications serve as the foundation for Inventory Placement, which in turn ensures that products are stored in the appropriate region.
This complex system creates a decisive competitive advantage.
And the Benefit is relatively simple: Amazon can provide greater selection and faster delivery while simultaneously reducing its own costs. According to Amazon, its newest fulfillment centers reduce Fulfillment Processing Time as well as Cost-to-Serve by up to 25%.
But how exactly can Amazon protect itself from competitors here? Competitors can buy robots, use AI and even build entire fulfillment centers. But they do not start from where Amazon already stands today. The Barrier here is time.
Amazon launched Prime in 2005, and the Kiva acquisition in 2012 significantly accelerated its robotics development. Amazon’s shipping system is the result of years of experience. This process includes continuous investments that do not always lead to the desired outcome. In the 2024 Shareholder Letter, Jassy explicitly describes Prime and Regionalization as examples of developments that did not emerge in a straight line, but through numerous detours.
And there is no shortcut for these developments. Every potential competitor has to go through this process and gradually adapt the system to its own company. For this reason, it is extremely difficult and time-consuming to replicate this Power. Amazon’s fulfillment system becomes increasingly complex over time, further increasing its lead. This growing complexity also means that the exact processes are not visible from the outside and therefore cannot be replicated one-to-one.
I consider Process Power probably the most overlooked Power at Amazon, because only the result of the process is visible, while the mechanisms behind it remain hidden. A time advantage, experience, high complexity and continuous development make this moat nearly impossible to catch up with.
This is the first edition of my new Moat Analysis series. If you want to follow future breakdowns of the moats behind businesses like Amazon, subscribe to Antares Research.
3. Power That Shows Up in the Numbers
For me, one of the strongest pieces of evidence for an intact moat is the development of margins.
A moat is supposed to protect a company from competitors taking market share and putting pressure on prices and therefore profits and margins.
Amazon’s segments are partly very capital-intensive, as in AWS, or cost-intensive, as in Commerce. The absolute level of margins is therefore not comparable to that of an asset-light software company. What I find much more important, however, is the development of margins.
If Amazon had not been able to preserve its competitive advantages and market share over the years, it would have been impossible to expand margins. The central Benefits of the Powers mentioned above are usually cost reductions or the ability to charge higher prices. The margin development shown below suggests that these Powers are present.
The gross margin in particular could be an indicator of Amazon’s pricing power. Over the years, it has slowly but steadily increased from around 27% to almost 50% today, nearly doubling. The operating margin stands at 11% today, compared with 1% in 2013. At AWS alone, it is even 39%. This demonstrates clear operating leverage, which has helped Amazon make further investments and expand its moat.
Investments have always been part of Amazon’s core strategy and are a major factor behind its success. For this reason, the FCF margin has essentially moved sideways over the years, sometimes reaching 5%, sometimes 3%, and is currently even negative due to high AI investments.
In my view, however, this is a sign that high Free Cash Flow is not always a guarantee of years of dominance. If Amazon had focused on generating as much FCF as possible, it probably would not be where it is today.
Amazon’s years of dominance and wide moat are reflected in the development of its stock price. Since its IPO in 1998, the long-term direction has essentially only been upward. Over the last 20 years, Amazon stock has outperformed the S&P 500 by nearly 200 times. Not many companies achieve such significant outperformance.
4. How Durable Is the Moat?
Amazon currently faces several threats that could potentially weaken its moat.
4.1 Technology & Competition
Purchase decisions could increasingly be made through AI Agents in the future. Amazon could therefore lose importance as the starting point for product discovery. AI Agents can compare several retailers at the same time and select the cheapest offer. The Network Economies of the Commerce business could potentially be weakened as a result.
Amazon, however, is responding aggressively with its own offerings such as Alexa for Shopping and AI Agents. AI could just as easily strengthen the Power.
Amazon also continues to face strong competition. Large companies such as Walmart in Commerce or Microsoft and Google in the Cloud business possess sufficient capital and scaling capabilities themselves and could cause Amazon to lose market share.
The commoditization of Cloud infrastructure represents a threat to AWS. If Compute, Storage and increasingly AI infrastructure become more standardized across hyperscalers, workloads could become more portable. This could lead to a reduction in Switching Costs.
4.2 Regulation
Future regulation could weaken Network Economies and Switching Costs. Because of its strong market position, Amazon has long been in the focus of several regulators that are increasingly trying to reduce this market power. This is a significant threat because there is relatively little Amazon itself can do about it.
4.3 When Scale and Complexity Turn Against Amazon
Amazon’s complexity regarding its fulfillment system creates Process Power. However, this complexity can become a problem if inventory is placed in the wrong locations, excess capacity emerges or the coordination of operations becomes increasingly difficult.
I currently consider the regulatory threat to be the most significant. Amazon itself is relatively powerless here and cannot take many direct measures.
Regarding technological threats, I am not particularly worried because Amazon has historically stood out through adaptability and targeted investments. They stay on top of current developments and are taking several measures themselves to avoid falling behind.
5. The Moat Verdict
Amazon as a whole is characterized by four clear Powers that form a central moat. These Powers create significant Barriers that preserve Amazon’s competitive advantages and allow the company to deliver superior products and financial results. Against new or emerging competitors, continued strong execution creates a wide moat.
Against existing competitors, Amazon must continue to protect its market share through targeted investments and the continued development of its systems and offerings.
For future development, the following KPIs should be monitored:
Commerce: Delivery Speed, Cost-to-Serve, Third-Party Seller Activity
AWS: YoY Growth Rate, Margins and indications of customer switching
All in all, Amazon receives a Wide Moat rating with the outlook for stable development.
Thank you for reading,
Finn
Antares Research
If you enjoy my strategy and way of thinking, and don’t want to miss future deep dives, quarterly reviews, watchlist updates, and further insights into my analysis process, consider subscribing to Antares Research.
That way, you’ll stay up to date on future opportunities, strong businesses, and attractive setups within my framework.
If you found this article valuable, I’d also appreciate you sharing it.
This article is for informational and educational purposes only and does not constitute investment advice or a recommendation to buy or sell any security. All views expressed reflect my personal opinion at the time of publication. Any investment decision should be based on your own research and judgment.
what an awesome read - thanks for this but it's also so frustrating to realise how cheap this stock was for such a long time before it really took off. Hindsight ... ugh!
Looks and feels a lot like my Moat analysis a while back! Great minds:
https://rudajev.substack.com/p/episode-22-the-moat-checklist-and?r=1uglxj&utm_campaign=post-expanded-share&utm_medium=post%20viewer