AWS, Amazon Web Services, holds approximately 30 to 37 percent of the global cloud infrastructure market. When it goes down, it does not just affect one company. It takes entire categories of the internet with it. On October 20, 2025, a single DNS failure in one AWS data centre in northern Virginia disrupted services across 3,500 companies in more than 60 countries. Over four million outage reports were filed within two hours. Apps like Zoom, Snapchat, Slack, Roblox, Signal, and Coinbase all went offline. The reason this keeps happening comes down to one thing: too much of the internet is running on too few providers.
The Morning Everything Stopped
On the morning of October 20, 2025, a restaurant owner in Houston named Dia Giordano woke up to her phone going off at 2 a.m. It was DoorDash, warning her that the online ordering system was down. By the time she got to her mental health clinics later that morning, her staff could not verify patient insurance information. The software was offline. Across the country in Indiana, a couple named Debi Dougherty and her husband arrived at a doctor's office for radiation therapy. It took forty minutes to book twenty-five days of follow-up appointments because the scheduling software was barely functioning. At the local Kohl's store, the credit card reader was down.
None of these people worked in technology. None of them had any connection to Amazon Web Services. But all of them were affected by a single technical failure in a data centre in northern Virginia.
That is what AWS outages actually look like in practice. Not server error pages on developer dashboards. Real disruption to real people doing real things, on an ordinary Monday morning.
At Ascella Group, we think about this from a marketing and brand perspective. When your business cannot be found online, when your website is dark, when your booking system is down and your social media scheduling tools are unresponsive, it is not just a technical failure. It is a brand failure in the eyes of your customer. They do not know about AWS. They just know your business did not work when they needed it.
That is the story we want to tell in this piece.
References: CNN Business, October 21, 2025; CNN Live Coverage, October 20, 2025
What AWS Actually Is and Why It Matters
To understand why an AWS outage ripples across the global internet, you need to understand what AWS actually is.
Amazon Web Services is the cloud computing division of Amazon. It provides on-demand computing power, data storage, networking, databases, and developer tools to businesses of every size, from early-stage startups to some of the largest organisations in the world.
The scale of what AWS operates is extraordinary. It generated $107.6 billion in revenue in 2024. It operates across 38 geographic regions globally. It runs on more than 6 million kilometres of fibre optic cabling. Its clients include Disney, the US Army, Capital One, United Airlines, and the NFL. According to Gartner, it held approximately 37% of the global cloud market share as of 2024.
For context, the "Big Three" cloud providers, AWS, Microsoft Azure, and Google Cloud, together controlled 63% of the global enterprise cloud infrastructure market in Q2 2025. AWS alone commands by far the largest individual share. When Lance Ulanoff, editor at TechRadar, was asked by CNN to describe AWS in plain terms, he put it simply: "It's like: why build the house if you're just going to live in it?" That is the value proposition. Instead of building and maintaining your own servers, storage systems, and networking infrastructure, you rent it from AWS.
The problem is not that this model is wrong. It is actually more efficient, more secure, and more cost-effective for most businesses than the alternative. The problem is what happens when the building catches fire.
References: Newsweek, October 2025; Gartner via CNN, 2024; DemandSage Internet Outage Statistics 2026; The Conversation, 2026
The October 2025 Outage: What Actually Happened
On October 20, 2025, AWS experienced a major outage beginning at approximately 3 a.m. Eastern Time. The root cause was traced to a DNS resolution failure in the US-EAST-1 region, specifically linked to an empty DNS record and incorrect automation updates.
US-EAST-1 is the oldest and most heavily trafficked AWS region, located in northern Virginia. Because it is the default region for so many AWS services and has been the backbone of the platform since its early days, its outages have a disproportionate effect on everything else. A failure there does not stay local. It cascades.
The outage lasted approximately 15 hours. AWS reported all services had returned to normal by 18:01 Eastern Time. By that point, the damage was already done.
The numbers tell the story clearly. More than 4 million outage reports were filed within the first two hours. Over the course of the day, Downdetector recorded a 970% increase above its average daily baseline for reports. Snapchat alone received approximately 3 million outage reports, making it the single most-affected platform during the peak period. The outage disrupted services for more than 3,500 companies across more than 60 countries.
The apps and platforms affected during the outage included Zoom, Slack, Roblox, Fortnite, Duolingo, Canva, Wordle, Signal, Robinhood, Coinbase, and Perplexity. The New York Times gaming platform, home to Wordle, saw login failures affecting players protecting their daily streaks. Government agencies reported disruptions. Point-of-sale systems at restaurants and retail locations stopped processing cards. Ring doorbells stopped working. CCTV cameras went offline.
Mehdi Daoudi, CEO of internet performance monitoring firm Catchpoint, told CNN: "The incident highlights the complexity and fragility of the internet, as well as how much every aspect of our work depends on the internet to work. The financial impact of this outage will easily reach into the hundreds of billions due to loss in productivity for millions of workers who cannot do their job."
References: DemandSage Internet Outage Statistics 2026; CNN Business, October 21, 2025; Newsweek, October 2025; Rest of World, January 2026; Al Jazeera, October 2025; CNBC, October 2025
Why One Data Centre Can Take Down the Global Internet
This is the question that keeps coming up after every major outage. If AWS has 38 global regions, hundreds of Availability Zones, and billions of dollars invested in redundancy, how does a failure in one data centre in northern Virginia bring down Snapchat for millions of users worldwide?
The answer has several parts.
The first part is the nature of cloud architecture. Many businesses configure their AWS workloads to run primarily in one region, often US-EAST-1, either because it was the default when they set up their infrastructure, because their team is based in the US, or simply because nobody revisited the configuration after the initial setup. When that region fails, those businesses fail entirely, regardless of how many other AWS regions are technically operational.
The second part is dependencies. Even businesses that are not directly using US-EAST-1 may rely on services that are. A company running on a different cloud entirely might use Slack for team communication. Slack may be running on AWS. If Slack goes down, that company's team communication goes down with it, even though the company never made a conscious decision to depend on AWS.
The third part is concentration. When 63% of global enterprise cloud infrastructure runs through three providers, and one of those providers holds 30 to 37% of the market on its own, the concentration creates what engineers call a single point of failure at civilisational scale. As Corinne Cath-Speth, head of digital issues at Article 19, put it after the October 2025 outage: "When a single provider goes dark, critical services go offline with it. Media outlets become inaccessible, secure communication apps like Signal stop functioning, and the infrastructure that serves our digital society crumbles."
The fourth part is that these events are not rare. Between August 2024 and August 2025, AWS, Azure, and Google Cloud together experienced more than 100 service outages. That is roughly two per week across the three major providers. AWS's average incident resolution time was 1.5 hours per incident across 38 incidents during that period. Google Cloud's incidents averaged 5.8 hours. Azure's averaged 14.6 hours.
Global network outages rose 33.38% from January to May 2025. This is not a trend moving in the right direction.
References: DemandSage Internet Outage Statistics 2026; The Conversation, 2026; Al Jazeera, October 2025; Rest of World, January 2026
The Financial Reality of Downtime
When we talk about AWS outages with business owners, the conversation often starts in abstract terms. Cloud infrastructure. Redundancy. DNS failures. Eyes glaze over.
So let us put it in the language that matters most to any business: money.
IT downtime costs averaged $14,056 per minute in 2024 according to DemandSage's research. For a business experiencing a four-hour outage, that is over $3.3 million in lost productivity and business impact, before accounting for lost sales, cancelled orders, or damaged customer relationships.
For the US economy as a whole, a single one-hour internet outage produces an estimated $458 million in losses.
The CrowdStrike incident of July 2024 remains a useful comparison. A faulty software update from the cybersecurity firm caused Microsoft Windows systems to fail globally. Delta Air Lines was forced to cancel over 5,000 flights. Fortune 500 companies collectively suffered $5.4 billion in direct losses. It forced discussions at board level across every major industry about technology dependency risk that had previously been treated as an operational concern rather than a strategic one.
The October 2025 AWS outage was a different kind of event but carried the same message. Sam Harris, Vice President at global consultancy Proxima, told Newsweek: "Dependency risk isn't limited to manufacturing or logistics. It exists deep within digital infrastructure too." Businesses with redundant workloads across multiple providers avoided the worst of the disruption. Businesses that had placed everything on AWS had no fallback.
The mean time to restore across the industry averaged 80 minutes globally in 2025. Eighty minutes of complete downtime for a business that processes transactions online, runs digital marketing, or serves customers through a cloud-based platform is a serious commercial event, not a minor inconvenience.
References: DemandSage Internet Outage Statistics 2026; Newsweek, October 2025; CNBC, October 2025
What This Has to Do with Your Marketing
Here is where we bring this directly into the world of marketing and brand, because this is where the conversation gets interesting for us as an agency.
When an AWS outage hits, the first thing most businesses think about is their website. Is it down? Is the store broken? Can customers check out? Those are legitimate, urgent concerns. But they are the surface layer of the problem.
The deeper layer is this: during an outage, every piece of your digital marketing infrastructure can go offline simultaneously. Your email marketing platform. Your CRM. Your social media scheduling tools. Your analytics dashboards. Your advertising campaign management system. Your lead capture forms. Your customer chat widget.
In the October 2025 outage, a tech worker writing for Rest of World described sitting helpless while their company's platforms, including HubSpot, Slack, and Salesforce, went offline at the same time. They could not communicate with clients. They could not access customer data. They could not provide service under their own SLA commitments. Their entire operational and marketing stack disappeared at once.
This is the reality of building a business entirely inside one ecosystem. It is convenient, it is efficient, and it works brilliantly right up until the moment it does not.
From a brand perspective, the damage extends beyond the hours of the outage itself. Customers who try to reach you during a disruption and cannot get through do not always come back. Research consistently shows that a poor digital experience damages brand trust, and an experience where your business simply cannot be reached at all is as poor as it gets.
Eighty-six percent of organisations have now adopted multi-cloud strategies for resilience according to 2025 data. Businesses that spread their critical infrastructure across more than one provider can route workloads away from a failing service and maintain continuity. The ones that have not are taking a concentrated risk that plays out very visibly during events like October 2025.
References: DemandSage Internet Outage Statistics 2026; Rest of World, January 2026
The Pattern of AWS Outages Over Time
October 2025 was not the first time AWS has caused widespread disruption, and it will not be the last.
In 2021, an AWS outage affected websites and services across the globe, including parts of Amazon's own delivery operations. In 2023, a disruption knocked many websites offline for several hours. In August 2024, an IAM authentication issue caused a regional login failure lasting around 50 minutes. In October 2025, the fifteen-hour DNS failure in US-EAST-1 disrupted 3,500 companies across 60 countries.
The pattern is consistent. The frequency of disruption is not decreasing. It is increasing. The scale of impact, because the internet's dependency on AWS has grown faster than the investment in alternatives, is also increasing.
A study found that 50% of data centres experienced at least one impactful outage over the past three years. Organisations achieving 99.9% availability, which sounds impressive, still experience a maximum of nine hours of downtime annually. For businesses with no contingency plan, nine hours of unexpected downtime spread across unpredictable moments throughout the year can mean missed campaign launches, offline sales periods during peak trading, and broken customer journeys that erode trust over time.
The issue is not that AWS is poorly built. By any technical standard, operating at the scale AWS operates with the reliability record it has is an engineering achievement. The issue is that the concentration of so much of the internet's infrastructure in a single organisation creates risks that no single organisation can fully mitigate.
As a computer scientist writing in The Conversation put it after the October 2025 outage: "The heavy reliance of the global internet on just a few major providers creates profound risks for both businesses and everyday users. A simple configuration error in one central system can trigger a domino effect that instantly paralyses vast segments of the internet."
References: The Conversation, 2026; CNBC, October 2025; DemandSage Internet Outage Statistics 2026
Building Resilience Into Your Business Presence
After the October 2025 outage, the founder of Honeycoin, a Kenyan cross-border payment company called David Nandwa, described what his team had done: "We rely on multiple clouds, including AWS, Google Cloud, and Heroku. During the recent AWS outage, we temporarily switched to Google Cloud to maintain service availability for most of our customers. If all cloud providers are down, we then fall back to using an on-premise server."
That is multi-cloud thinking in practice. Not all businesses need that level of sophistication. But the principle translates to every scale.
For businesses thinking about their marketing and digital presence specifically, resilience looks like:
Knowing which tools run on AWS. Most businesses do not know this. Your email tool, your CRM, your scheduling software, your analytics platform, and your advertising tools may all share the same underlying infrastructure. Understanding your dependency map is the starting point.
Having a backup communication channel. If your customer chat goes down, do customers have a phone number? An email address? If your email marketing platform is offline, can you still reach your audience another way? Resilience in marketing is about not being entirely dependent on any single channel or platform.
Testing your website independently. Many businesses assume their website is secure if they have it hosted somewhere. But if your website depends on third-party plugins, payment processors, or widgets that run on AWS, the site may appear but not function during an outage.
Having a crisis communication plan. During the October 2025 outage, the businesses that handled it best communicated proactively with customers. A short social media post acknowledging the issue, a brief email to your database explaining the disruption, and a clear statement on your website turn an infrastructure failure into an opportunity to demonstrate how you look after your customers under pressure.
At Ascella Group, we help businesses think about their brand not just in ideal conditions but in the moments when things go wrong. How you communicate during a crisis, how accessible and transparent you are, and whether your customers feel looked after regardless of what is happening technically, is a significant part of what builds lasting brand trust.
Frequently Asked Questions
What caused the October 2025 AWS outage and how long did it last? The October 20, 2025 AWS outage was caused by a DNS resolution failure in the US-EAST-1 region, linked to an empty DNS record and incorrect automation updates. The outage began at approximately 3 a.m. Eastern Time and lasted approximately 15 hours, with AWS reporting normal operations restored by 18:01 Eastern Time.
How many businesses were affected by the October 2025 AWS outage? The October 2025 outage disrupted services for more than 3,500 companies across more than 60 countries. Over 4 million outage reports were filed within the first two hours. Downdetector recorded a 970% increase above its average daily baseline report volume at the peak of the disruption.
Why does one AWS region failing cause global internet disruptions? The US-EAST-1 region in northern Virginia is the oldest and most heavily used AWS region. Many businesses default to this region for their infrastructure. Additionally, widely used services and platforms like Slack, Zoom, and others run their own operations on AWS, meaning even companies not directly using AWS may be indirectly affected through the tools they rely on daily.
What is AWS's market share in cloud computing? AWS held approximately 30 to 37% of the global cloud infrastructure market as of 2024 and Q2 2025, according to Gartner and multiple market research sources. Microsoft Azure held approximately 20% and Google Cloud approximately 13%, giving the Big Three a combined share of around 63% of global enterprise cloud infrastructure.
How much does an AWS outage cost businesses per minute? IT downtime costs averaged $14,056 per minute in 2024 according to DemandSage's research. For the broader US economy, a single one-hour internet outage is estimated to produce $458 million in economic losses. The CEO of Catchpoint told CNN after the October 2025 outage that its total financial impact would "easily reach into the hundreds of billions."
How often do major cloud outages happen? Between August 2024 and August 2025, AWS, Azure, and Google Cloud together experienced more than 100 service outages, roughly two per week across the three providers. Global network outages rose 33.38% from January to May 2025. The mean time to restore globally averaged 80 minutes per incident.
What apps and platforms were affected during the October 2025 AWS outage? The outage affected a wide range of platforms including Zoom, Slack, Snapchat, Roblox, Fortnite, Duolingo, Canva, Wordle, Signal, Robinhood, Coinbase, Perplexity, Amazon itself, Delta Airlines, Venmo, and Google. Snapchat received approximately 3 million outage reports during the peak, making it the single most-reported platform.
What percentage of companies have adopted multi-cloud strategies to protect against outages? According to 2025 data compiled by DemandSage, 86% of organisations have adopted multi-cloud strategies for resilience. This approach spreads infrastructure across multiple cloud providers so that if one fails, workloads can be routed to an alternative, maintaining business continuity.
How does an AWS outage affect a brand's marketing operations specifically? During an AWS outage, marketing platforms that run on AWS infrastructure can go offline simultaneously. This can include email marketing platforms, CRM systems, social media scheduling tools, advertising management dashboards, analytics platforms, and live chat tools. Businesses can find themselves unable to communicate with customers, unable to track campaign performance, and unable to process leads or sales, all at the same time.
What happened during the CrowdStrike incident of July 2024 and how does it compare to AWS outages? In July 2024, a faulty software update from cybersecurity firm CrowdStrike caused Microsoft Windows systems to fail globally. The incident forced Delta to cancel over 5,000 flights and cost Fortune 500 companies an estimated $5.4 billion in direct losses. While the cause was different from an AWS outage, both events illustrate the same core vulnerability: too much of the global economy depends on too few technology providers.
What is the US-EAST-1 region and why do outages there have outsized global impact? US-EAST-1 is Amazon's oldest AWS region, located in northern Virginia, United States. Because it was AWS's first region and became the default for many services and configurations, a disproportionate share of global AWS workloads runs through it. When it fails, the impact extends far beyond the United States because so many globally used platforms and services have their infrastructure anchored there.
What practical steps can a business take to protect itself from cloud outage disruption? The most effective steps are: mapping your technology stack to understand which tools depend on which cloud providers, adopting a multi-cloud approach for critical services, maintaining backup communication channels for reaching customers during outages, testing your website's full functionality rather than just its availability, keeping a simple crisis communication template ready to deploy during disruptions, and reviewing which business functions would stop entirely if your primary cloud provider went offline for six hours.
Conclusion
An AWS outage is not a story about technology. It is a story about dependency.
When 37% of the internet's cloud infrastructure runs through one company, and that company experiences a failure, the consequences show up in a restaurant owner's phone notifications at 2 a.m., in a radiation therapy patient's appointment booking taking forty minutes instead of two, and in millions of people around the world unable to use apps they depend on every day.
For businesses, the lesson is not to stop using AWS. It remains one of the most capable and reliable platforms available. The lesson is to understand what you depend on, build in alternatives where the stakes are highest, and have a clear plan for how you communicate with your customers in the moments when your technology lets you down.
Because your customers are not tracking AWS status pages. They are tracking whether your business showed up for them.
At Ascella Group, we help businesses build marketing and brand strategies that account for this reality. A strong brand is not just what you do when everything works. It is how you behave when it does not. The businesses that communicate clearly, act quickly, and treat every disruption as a moment to demonstrate their values are the ones that come out with their customer relationships intact.
If you want to think about your digital presence with that kind of resilience built in, that is exactly the kind of work we do.