Microsoft 365 outage explained with server rows and email iconWas your business one of the many cut off from Outlook last Thursday? On November 12, Microsoft 365 suffered one of its biggest outages in years. Email went down worldwide, and Teams, SharePoint, Copilot and other tools that depend on email were disrupted along with it.

Microsoft has now released a preliminary report explaining what went wrong. Here's the plain-English version.

What happened

According to Microsoft, the outage ran from about 11:20 a.m. ET on Thursday, November 12 to 11:10 a.m. ET on Friday, November 13. Exchange Online, the system behind Outlook and Microsoft 365 email, took the biggest hit:

  • Most users couldn't open their mailbox in Outlook or other email apps.
  • Many messages were delayed.
  • Other services were affected where they rely on Exchange Online, including parts of Teams, Copilot, SharePoint, OneDrive, Defender and Purview. Most of those services largely stayed online.

What caused it: three problems at once

Microsoft says no single thing caused the outage. Three separate problems lined up at the same time.

1. A hidden bug from July. In July, Microsoft rolled out a fix for the gateway servers that connect users to many Microsoft 365 services. That fix contained a bug that could make the servers freeze up under certain heavy loads. It ran worldwide for more than three months with no problems. Microsoft's testing missed it because its labs mostly used older, less powerful hardware, and the bug shows up more on newer machines with more processor cores. Microsoft says the bug only affects these specific gateway servers, not Windows computers in general.

2. Extra traffic. About 15 hours before the outage began, Microsoft's systems automatically moved Teams traffic onto the same gateway that handles Exchange Online email. Microsoft says this kind of switch is a normal part of how it keeps services running.

3. Less room to spare. At the same time, about 14% of that gateway's servers were already offline because of an unrelated issue the day before. Microsoft says it had enough spare capacity that this wouldn't have caused trouble on its own.

Together, the extra traffic and the reduced capacity set off the July bug, during peak business hours in the United States.

Why it spread so fast

When a server stops responding, Microsoft's systems automatically send its users to other servers. That's normally what keeps the service reliable. This time it made things worse: each server that failed pushed its traffic onto the rest, and they started failing too. In under two hours, the problem had spread across Microsoft's network, and at the worst point, more than 80% of the servers weren't responding.

Microsoft's network is divided into regions, partly to stop a problem in one area from spreading to others. Microsoft says a separate issue kept those barriers from working, which is how the outage went global.

Why Teams came back sooner than email

Services like Teams, OneDrive, SharePoint and the Office apps were able to move off the faulty gateway, so they recovered earlier. Teams chats and meetings worked for most people through most of the outage. Exchange Online can't move off that gateway because of its size, so email took the longest to recover.

Why it took so long to fix

Microsoft gives three reasons:

  • Repaired servers were overwhelmed right away. As servers came back online, the backlog of waiting traffic swamped them and they froze again.
  • Automatic recovery wasn't built for this. Microsoft's normal automated recovery isn't designed for a situation where almost everything has stopped responding at once.
  • The bug was months old. Microsoft's usual tools for quickly undoing a bad change couldn't simply roll back something that had been in place since July.

Once engineers confirmed the bug, Microsoft built an update that put the affected code back the way it was before. It then took groups of servers offline, applied the update, restarted them, and brought them back gradually. Microsoft says customers stopped seeing active problems by 11:10 a.m. ET on Friday, November 13. It then monitored the service and cleared the email backlog before posting its final status updates late Friday night.

Was it a cyberattack?

No. Microsoft describes this as a technical failure inside its own systems. Nothing in its report points to a hack.

What Microsoft has done so far

Microsoft says it has already:

  • Fixed the faulty code.
  • Extended how long it holds delayed email, so messages queued during the outage were delivered instead of bounced back.
  • Added server capacity.
  • Added monitoring for traffic shifts and connection problems.
  • Put limits in place on how traffic is shifted, to help keep one failure from cascading into others.

Microsoft calls this report preliminary. It plans to publish a final report with a full list of improvements by November 20, 2026. We'll share the highlights when it's out.

What this means for your business

An outage like this is a reminder that even the biggest cloud providers can go down, and that a lot of everyday work runs through email. A few questions worth asking:

  • If Outlook went down at 11 a.m. tomorrow, how would your customers reach you?
    • Do they have a phone number that still works, and does your team know to use it?
  • Who on your team would know what's going on?
    • During this outage, Microsoft posted updates in its admin center and on its status page. Does someone at your business know where to look?
  • Which of your daily tasks quietly depend on email?
    • Think about calendar invites, password resets, invoices, and alerts from other software. This outage showed how many tools stop working when email does.
  • Do you have a backup way to reach your own team?
    • A group text, a phone tree, or a separate chat tool can keep everyone on the same page.
  • Would you know if important emails were lost?
    • Microsoft protected delayed messages this time. Do you have a way to confirm that time-sensitive messages arrived?

If you'd like help planning for the next outage, Net Tel One can help. Get a free consultation and learn how we can make your technology worry-free.

For IT admins: Microsoft's full preliminary report is in the Microsoft 365 admin center under Service health, incident EX1550543 (also MO1550821).