Microsoft 365 Outage (Nov 12-13, 2026): What Happened and What to Do Now
What caused the November 12-13, 2026 Microsoft 365 outage, how Microsoft fixed it, and the steps small businesses should take before the next one.
Last updated November 16, 2026, based on Microsoft's preliminary post-incident report. Microsoft says a final report is due November 20, and we'll update this post when it's out.
Late Thursday morning on the East Coast, November 12, the phones started ringing. Outlook wouldn't connect, emails sat in outboxes, and people were restarting laptops and reinstalling Office to fix a problem that wasn't on their end. If your team lived through the Microsoft 365 outage of November 12 and 13, 2026, you know how quickly a normal workday stops when email goes away.
Microsoft has now published a preliminary post-incident report explaining what broke, why it spread worldwide, and how they fixed it. I read it closely, and in this post I'll translate it into plain English.
More importantly, I'll cover what a small business or manufacturer should take from it. You can't stop Microsoft from having a bad day. You can decide how much that bad day costs you.
In this guide
- What happened during the Microsoft 365 outage
- The root cause, in plain English
- How Microsoft fixed it
- What Microsoft is changing to prevent a repeat
- What your business should do now
- How to build a simple outage plan for next time
What happened during the Microsoft 365 outage
According to Microsoft's preliminary report (incident IDs EX1550543 and MO1550821), the impact window ran from 4:20 PM UTC on Thursday, November 12, to 4:10 PM UTC on Friday, November 13, 2026. For those of us on Eastern time, that's about 11:20 AM Thursday to 11:10 AM Friday. Microsoft kept monitoring and sent final customer updates early on November 14.
Who was affected
The biggest hit was Exchange Online. Microsoft says most users couldn't reach their mailboxes through Outlook or other mail clients, and many saw delayed email.
Other services that depend on Exchange Online were also affected. The report lists Copilot, Teams, SharePoint, OneDrive, Defender, Purview, Fabric, Intune, Planner, Power Apps, To Do, Power BI and Universal Print. Most of those stayed online, but features tied to email and calendars didn't work reliably.
The impact varied by region, connection method and time. That's why one person in your office might have had webmail while the person next to them couldn't open Outlook at all.
Was it a cyberattack?
No. Microsoft said during the incident that it had no indication of malicious activity, and described it as a technical issue not related to cyberattacks or AI agents. The preliminary report points to a software defect combined with traffic and capacity conditions, which I'll explain next.
The root cause, in plain English
Microsoft's report describes three things that lined up at the same time. None of them alone would have caused a global outage. Together, they did.
1. A hidden bug that had been sitting there since July
Microsoft 365 sits behind a networking gateway, a large fleet of servers that handles incoming connections before they reach services like Exchange. In July, Microsoft deployed a fix to those gateway servers. That fix contained a defect that caused lock contention in an internal Windows process.
Lock contention is when many tasks are waiting on the same shared resource and can't move forward. Under the right conditions, it ate up processor and memory until the servers stopped responding. The change had run worldwide for months without problems, so nobody knew it was there.
Microsoft is clear that this defect, and the fix for it, applied only to its Microsoft 365 gateway servers and not to broader Windows deployments. You don't need to patch your own Windows servers because of this.
2. A shift in Teams traffic
About 15 hours before the outage, Teams traffic failed over from an alternate gateway to the primary gateway, the same one that also supports Exchange Online. That added load and changed the traffic pattern the gateway servers were handling.
3. Less capacity than normal
On top of that, roughly 14% of gateway server capacity was already out of service because of an unrelated issue the day before.
How it became a worldwide outage
When the first servers became unresponsive, automation did what it was designed to do: it moved traffic to healthy servers. That extra traffic pushed those servers into the same failure. At the worst point, more than 80% of gateway servers were unresponsive.
Microsoft also has "partition bulkheads," boundaries meant to keep a failure in one area from spreading. The report says a separate issue kept those from working as intended. Teams, OneDrive, SharePoint and Office apps could route around the bad gateway, but Exchange Online couldn't, because of its scale.
This is a pattern I've seen on smaller networks too. A failover system that sends all the load to the "healthy" side can take down the healthy side as well, if the underlying problem goes with the traffic.
How Microsoft fixed it
The fix took most of a day, and the report is candid that early attempts didn't work. Here's the sequence, in UTC:
- November 12, around 4:03 PM: Alerts started firing, and automation began redirecting traffic.
- November 12, evening: Engineers restarted servers and tried failovers. Recovered servers kept going unresponsive again once traffic returned.
- November 13: Engineers identified the connection-handling process as the main cause.
- November 13, 7:20 AM: Microsoft applied a mitigation that reverted the defective change, starting with part of the capacity, then expanded it region by region.
- November 13, around 11:00 to 11:35 AM: Asia Pacific and then Europe recovered.
- November 13, 2:17 PM: Webmail access returned to expected levels globally.
- November 13, 4:10 PM: Microsoft confirmed the incident was resolved.
Microsoft also extended the email delivery window so delayed messages wouldn't bounce back as undeliverable, then worked through the queued mail. If someone tells you they never got an email from Thursday, check again. Most delayed mail should have been delivered once queues cleared.
What Microsoft is changing to prevent a repeat
In the preliminary report, Microsoft lists these items as completed:
- Reverted the connection-handling code to a known good state.
- Extended the email delivery window and processed all queued mail.
- Added proactive capacity to the gateway.
- Added monitoring for traffic-redirection patterns and connection-handling health.
- Added guardrails on traffic distribution to avoid cascading failures.
The report says the final version, with a full list of longer-term repair items, is due November 20, 2026. I expect it to say more about why testing missed the bug and why the bulkheads didn't contain it. We'll update this post when it's published.
What your business should do now
Here's the good news: the failure was inside Microsoft's cloud, so there's nothing to fix in your tenant or on your PCs. You don't need to reinstall anything or reset anyone's password because of this outage. But there are a few things worth doing this week.
A quick post-outage checklist
- Confirm delayed email arrived. Ask key people, especially sales, purchasing and shipping, to check for anything important from Thursday and Friday that seems missing. Follow up with customers and suppliers directly if needed.
- Save the post-incident report. Download it from the Service Health area of the Microsoft 365 admin center. Service health history is only kept for a limited time, so keep your own copy.
- Write down your impact. Note which teams were affected, for how long, and what workarounds you used. This helps with any service credit request and with planning.
- Look into SLA credits. Microsoft 365 has a financially backed uptime commitment, and credits generally have to be requested by your admin or through your partner. Check your agreement or ask your Microsoft partner whether this event qualifies for you.
- Watch for outage-themed phishing. Attackers often use big news events as lures. Be suspicious of emails about "re-verifying your Microsoft account" or "recovering lost messages," and never enter credentials from a link in an email. Our 10 steps to protect your company covers the basics.
What not to do during the next outage
During this incident, plenty of people tried to fix the problem on their end. That usually wastes time and can create new issues. When Microsoft 365 is down:
- Don't reinstall Office or delete Outlook profiles.
- Don't reset passwords unless you have a real security reason.
- Don't make big configuration changes in your tenant. If something else breaks, you'll want to know it wasn't you.
How to build a simple outage plan for next time
Every cloud provider has outages. The businesses that handled this one best weren't the ones with the fanciest tools. They were the ones who already knew how to check status, how to reach their people, and what to do while they waited.
Set up early warning
- Turn on Service health email notifications in the Microsoft 365 admin center, and send at least one alert to an address that is not hosted in Microsoft 365. If your email is down, an alert sent to that same email won't help.
- Bookmark Microsoft's public status page at status.cloud.microsoft so you can check it even when you can't sign in to the admin center.
- Install the Microsoft 365 Admin mobile app on the phones of whoever manages IT.
Have a backup way to reach people
Decide now how you'll contact employees if email and Teams both go down. For most small businesses, that's a simple group text or a phone tree, plus a short list of key customer and supplier phone numbers kept somewhere outside Microsoft 365.
For manufacturers, think about production too. If work orders, shipping notices or supplier POs depend on email, decide what the manual fallback is and who owns it.
Know your critical processes
Ask one question for each department: "What stops if we lose email for a day?" Usually the answer is a short list, like quotes, purchase orders, shipping confirmations and customer support. Write down a workaround for each one.
Decide who calls the shots
Pick one person who confirms the outage, sends the message to staff, and gives the all-clear. If you work with a managed IT provider, they should be doing this for you and keeping you updated. That's one of the real advantages we covered in MSP vs. hiring a full-time IT person.
A 15-minute outage plan template
You can put your whole plan on one page:
- How we confirm it: admin center, status page, our IT provider.
- Who decides: name and backup name.
- How we reach staff: group text or phone tree.
- How we reach customers and suppliers: phone list stored outside Microsoft 365.
- Department workarounds: one line each.
- After it's over: check delayed mail, save the report, note the impact.
Frequently asked questions
What caused the Microsoft 365 outage on November 12, 2026?
Microsoft's preliminary report says a defect in a July update to its Microsoft 365 gateway servers caused lock contention in an internal Windows process. Combined with a shift in Teams traffic and about 14% of gateway capacity already offline, it triggered a cascading failure that left more than 80% of gateway servers unresponsive at its worst.
How long was Microsoft 365 down?
Microsoft's stated impact window was from 4:20 PM UTC on November 12 to 4:10 PM UTC on November 13, 2026, or just under 24 hours. Recovery happened region by region, so some users got access back earlier than others.
Was the Microsoft 365 outage a cyberattack?
No. Microsoft said it found no indication of malicious activity and that the cause was a technical issue. Even so, watch for phishing emails that use the outage as a lure.
Do I need to change anything in my Microsoft 365 tenant or on Windows PCs?
No. The failure was inside Microsoft's cloud, and Microsoft says the defect applied only to its Microsoft 365 gateway servers, not to broader Windows deployments. The useful steps now are confirming delayed email arrived and setting up better outage alerts.
Can my business get a credit for the Microsoft 365 outage?
Possibly. Microsoft 365 services carry a financially backed uptime commitment, and credits usually have to be requested by your admin or Microsoft partner within a set time window. Check the terms of your agreement, or ask your partner, since eligibility depends on your plan and how you buy Microsoft 365.
Not sure how ready you are for the next outage?
This outage wasn't anyone's fault at a small business. But it showed which companies had a plan and which ones were scrambling. If you'd like a second set of eyes on your setup, we offer a free cybersecurity assessment. We'll check your alerting, backup communication, account security and the gaps an outage like this exposes, and give you a short list of practical fixes. Request your free assessment here.
Related reading
- Managed IT Services in Roanoke, VA: What Local Businesses Should Expect
- MSP vs. Hiring a Full-Time IT Person: A Real Cost Breakdown
- Cybersecurity for Roanoke Small Businesses: 10 Steps to Protect Your Company
Harsha is a co-founder of SoftwarePac Technologies in Roanoke, VA, and a network engineer with 15 years of experience.




