Outlook Went Down, the World Stopped: Too Dependent on Microsoft?

Thursday evening, around 7 p.m. Turkey time. You open Outlook and the inbox spins, and spins, and spins… First you think, "is my internet down?" Then Teams won't open either. Then Copilot. Then you realize the problem isn't you; millions of people in 74 countries are staring at the same spinner.

As of the morning of November 13, when I'm writing this, the outage still isn't fully over, and Microsoft describes its progress as, in its own words, "marginal success." Let me walk through what happened, what Microsoft has said, and the question I think we really need to talk about. Because the question this incident asks us is as interesting as its technical side.

What Happened?

According to Al Jazeera, problems started around 16:00 GMT on Thursday, November 12, and within an hour Downdetector had reports from 74 countries and territories. According to ABC News, Outlook reports alone topped 130,000. Even the next morning, 15 hours in, there were more than 43,000 reports from 72 countries.

It wasn't just Outlook:

• Exchange Online mailboxes (web, desktop and mobile)
• Teams, SharePoint Online and Microsoft 365 Copilot
• Integrations that companies such as OpenAI and Anthropic rely on

And it went beyond corporate email. Government systems in Singapore and Japan, the Philippine Supreme Court's communications, and Australia's spy agency and federal police were affected. A server problem slowed down public communication in several countries at once.

What Does Microsoft Say?

The root cause hasn't been published yet; Microsoft says it will share the full report once the incident is mitigated. As SecurityWeek reports, here's what it has said so far:

• "Traffic didn't increase": The extra load came not from more users but from a large number of servers becoming unexpectedly unavailable. The remaining servers had to carry everything.
• The secure connections process: A process that establishes secure connections became overwhelmed, and servers stopped handling requests. They also started failing automated health checks.
• A network change: Microsoft began rolling back a recent network configuration change that coincided with the start of the disruption. It has not confirmed that change was the cause.
• Not an attack: Microsoft's Charles Lamanna said there is no indication this was the result of malicious or cyber activity.

The most telling detail: restored servers can become unresponsive again once traffic hits them. So "turn it off and on again" isn't enough; the moment a server comes back, everyone waiting piles onto it at once.

Through an Engineer's Eyes: A Domino Story

This looks a lot like a classic distributed-systems scenario. The exact root cause won't be known until Microsoft publishes it, but this chain is familiar:

• Some servers go down.
• The rest take on extra load and start failing health checks too.
• Servers failing health checks are pulled from the pool, so the load lands on even fewer machines.
• A server recovers, millions of clients try to reconnect at the same moment, and it falls over too.

The literature calls this a cascading failure and a thundering herd. Microsoft testing a workaround that bypasses the component processing health checks fits the picture: sometimes the mechanism you add to protect a system is what tears it down faster in a crisis.

The Real Question: Is This Much Dependence Normal?

Now the debatable part. This isn't Microsoft's first big outage this year. In January 2026, Microsoft 365 was down for about 10 hours across North America. In late August, thousands of Outlook users were locked out. Remember CrowdStrike in 2024 and the AWS outage in 2025, too. Even Australia's Home Affairs Minister Tony Burke compared this incident to CrowdStrike.

There used to be a saying: "Nobody ever got fired for buying IBM." Today's version is Microsoft 365. Nobody gets fired for saying "let's trust Microsoft with our email." But if the communications of governments, courts, intelligence services and companies can all stop at once because of one failure at one provider, is that really an engineering choice, or just habit?

The counterargument is strong, too. Would running your own mail server be more reliable than Microsoft's multibillion-dollar infrastructure? Is moving to Google Workspace the answer, or does every cloud go down eventually? I lean toward the second view: the problem isn't choosing Microsoft, it's having no plan B for when a single provider goes down.

One irony I can't skip: at a company that puts AI into every product, Copilot went quiet this week right along with email.

Who Is Affected / What Should You Do?

Every company using Microsoft 365, and every developer building on that ecosystem. There are a few lessons here you can apply today:

• Follow the status from the right place: If you're an admin, track incidents EX1550543 (Exchange Online) and MO1550821 (Microsoft 365) in the Microsoft 365 admin center; look at official updates instead of guessing.
• Set up a plan B for communication: Decide in advance where your team talks when email and Teams go down together. Don't let critical workflows depend on a single channel.
• Prepare your own systems for the thundering herd: Make your clients retry with exponential backoff and jitter, so that when a server recovers, not everyone hits it in the same second.
• Question your health checks: If your health check kicks healthy servers out of the pool under load, it may be the thing making the crisis worse.
• Review your Microsoft-dependent integrations: If your product depends on Exchange or the Graph API, test how it behaves when that service is gone. An error message, a silent hang, or a crash?

I'll update this post when Microsoft publishes its root cause report. Until then: is the problem Microsoft, or us for trusting everything to a single company? Let's argue it out in the comments.