We have all been there; you hit Send, the little wheel spins, and… nothing. More often than not, the fix is a quick restart or a Wi-Fi hiccup. Thursday, November 12, 2026, was different. Late that morning (Eastern time), Outlook stopped cooperating for offices across the Americas — and then for much of the planet. When the East Coast logged off that evening, it was not fixed. When those same people logged back on the morning of Friday, November 13, it STILL was not fixed.
Make no mistake; this was not a blip. Microsoft did not mark the incident resolved until late on November 13 — roughly 36 hours after it began — and it landed squarely on the one tool that small businesses, schools, and government offices lean on the hardest; email. So, with the dust settled and the first round of answers in, let’s break down what actually happened, what we have learned since, and the question every business owner should be asking: how do we keep the next one from taking us down with it?
(Spoiler: nobody outside of Redmond can stop Microsoft from having a bad day. What we CAN do is stop Microsoft’s bad day from becoming yours.)
The Timeline: From Late-Morning Hiccup to 36-Hour Ordeal
Microsoft’s service health dashboard logged the start of the Exchange Online incident (EX1550543) at 11:17 a.m. ET on November 12. The first public alert followed at 12:54 p.m. — initially framed as trouble reaching mailboxes in the Americas — and just 13 minutes later, the warning widened to customers anywhere in the world. At 1:26 p.m., Microsoft opened a second, broader incident (MO1550821) for the long list of Microsoft 365 services that lean on Exchange behind the scenes.
From there, the afternoon turned into a game of whack-a-mole. By 2:06 p.m., Microsoft had paused all deployments and configuration changes, redistributed traffic around the world, and begun restarting servers. Around 6 p.m., it reported that some servers were back and holding steady; by 7 p.m., those same servers were falling over again. Even the diagnosis moved around — at 2:57 p.m., Microsoft said nothing it had changed lined up with the start of the trouble; by 8:06 p.m., it was reverting a network change after all. Shortly before 2 a.m. on November 13 (more than 14 hours in), the company’s own update described only “marginal success.”
The turn FINALLY came at 2:24 a.m. ET on November 13, when Microsoft reported positive results from a configuration patch to an upstream authentication and routing layer. With that holding, it worked region by region — draining traffic, restarting servers, and bringing them back in stages. Shortly after 4 a.m., Microsoft extended its email timeout window to 48 hours to help protect delayed messages. Parts of Asia Pacific were recovered by 6:01 a.m., Europe and North America improved through the morning, and by 9:01 a.m., the queues of waiting email were significantly reduced.
Microsoft’s final incident records list the impact as ending at 11:10 a.m. ET on November 13 — just shy of 24 hours after the dashboard’s start time — BUT the cleanup dragged on. The company spent the afternoon draining queued messages, marked both incidents resolved at 11:15 and 11:17 p.m. ET that night, and confirmed on X in the early hours of November 14 that service was back for the last remaining customers and regions. Start to finish; roughly 36 hours. (One oddity worth watching: those closing entries also list an earlier start time of 5:05 a.m. ET on November 12, a discrepancy the post-incident report will hopefully clear up.)
Let’s Talk Numbers; Outage Sized Numbers
Downdetector logged roughly 130,000 Outlook problem reports at the peak (one tally put the high-water mark at 138,058), with reports coming in from a whopping 74 countries and territories within the first hour. And remember; Downdetector only counts the people who take the time to file a report. Microsoft says more than 430 million people use Microsoft 365 apps, and while nobody knows exactly how many of them were affected, the real number is undoubtedly far higher than any outage tracker shows.
Then there is the price tag. Anderson Economic Group, a Michigan research firm, estimated that lost productivity could approach an astonishing $7 billion, with as many as 26 million users on November 12 and 51 million on November 13 disrupted badly enough to cause an economic loss, according to CNN. (To be fair, that is an early model rather than an audited total; it still gives a sense of the scale.)
The real-world list is even more telling. In the U.K., the National Health Service was affected, police could not take online crime reports, and a judge in Manchester halted a trial and sent the jurors home. Australia’s spy agency and federal police were caught up in it, and the Australian government activated a formal agreement with Microsoft to get priority in the restoration line. Ministries in Japan, Singapore, Greece, and Norway reported email disruptions, and some banks in Hong Kong were disrupted as well. Closer to home, the U.S. House of Representatives said its users were affected, New York City’s Emergency Management Department pointed residents to the phone instead of email, and Rep. Josh Gottheimer of New Jersey told CNBC he had asked for a briefing. Even OpenAI and Anthropic reported trouble connecting their AI tools to Microsoft 365.
One detail worth noting: this was an Exchange problem at its core. By the end, Microsoft’s broader incident listed 20 different services — from Copilot and Planner to Intune and Power BI — but it was the features that depend on Exchange that broke. Teams messages and calls kept working, for example, while Teams calendars and meeting scheduling did not. SharePoint files stayed reachable. And in a twist that should make every security-minded reader sit up, Defender’s Safe Links (the feature that checks a link at the moment someone clicks it) could not always verify URLs opened in Outlook and Teams.
The Cause: A Hidden Flaw, a Routine Change, and a Domino Effect
So what broke? Microsoft’s explanation evolved as the hours went by — from elevated load, to servers dropping offline, to failed health checks — before the company landed on an answer on November 13. The incident traces back to a problem introduced during an earlier network capacity upgrade; it sat quietly until a later network change brought it to the surface. Microsoft was also emphatic that this was not a cyberattack and (in a very 2026 clarification) that AI agents had nothing to do with it.
The mechanics are worth understanding, because they explain why the fix took so long. Large cloud services constantly run automated “health checks” against their own servers; fail the check, and a server gets pulled out of rotation so that users are not sent to a broken machine. On November 12, a core process that sets up secure connections became overwhelmed on the affected servers, those servers stopped answering, and the health checks began pulling them from service — a lot of them. Microsoft was clear that traffic had not spiked; capacity had shrunk. In its simplest form, fewer healthy servers = more load on the survivors = even fewer healthy servers.
That is also why “just restart it” was not enough. A restored server would rejoin the pool, take on traffic, and fall right back over; at one point, the repair work itself set off a surge of calls from Microsoft’s routing layer to the servers behind it. It took that upstream patch — plus a slow, staged return to service — to break the cycle.
“We are deeply sorry,” the company said in its November 13 apology. As of November 14, 2026, the explanation above is still officially preliminary; Microsoft has committed to a preliminary post-incident report within two business days and a final one within five, which should put both in the week of November 16, 2026.
The Pattern: This Was Not a One-Off
Here is the part that should get every business owner’s attention; November 12 was not a freak event. In January 2026, a Microsoft 365 outage tied to North American infrastructure dragged on for nearly 10 hours. In July 2026, a bug in Microsoft’s automated network maintenance system pulled routes from more devices than intended, disrupting Microsoft 365 and Azure for close to five hours. (The first 24 hours of the November outage ran longer than both of those combined.) And in April 2026, Tony Redmond of Office 365 for IT Pros reported that Microsoft 365’s uptime for the first quarter of 2026 came in at 99.526% — the lowest figure since the site began tracking in 2013. That may sound close to perfect; by our math, it works out to the equivalent of roughly ten hours of downtime in a single quarter.
Notice the common thread? None of these was pinned on hackers. In every case, the trouble started inside Microsoft’s own infrastructure — and both the November outage and the July one trace back to changes Microsoft made to its own network. That is not a knock on Microsoft so much as a fact of life at this scale. Sanchit Vir Gogia of Greyhound Research made the point to CIO during the outage: too many organizations treat Microsoft 365 as just another license on the invoice, when it is really a single point of failure for the whole business. The smart money plans for the next outage rather than hoping there won’t be one.
The MSP Reality: We Can’t Prevent the Outage — We Can Prevent the Downtime
Let’s be honest about what a managed services provider (MSP) can and cannot do. No MSP on Earth can reach into a Microsoft datacenter and fix Exchange Online. What we can do is design a client’s environment so that when a single vendor stumbles, the business keeps moving. Put simply: an outage only becomes downtime when there is no Plan B.
It also helps to be clear-eyed about what Microsoft actually promises. The standard service level agreement (SLA) for Microsoft 365 targets 99.9% monthly uptime — which still leaves room for roughly 43 minutes of downtime every single month. Fall short, and the remedy is a service credit; for Exchange Online, that is 25% of the month’s fees below 99.9%, 50% below 99%, and 100% below 95%. In a 30-day month, the 50% tier takes a little more than seven hours of downtime per user and the 100% tier takes roughly 36 hours, so most organizations hit by the November outage are looking at 25% or 50%, depending on how long their own people were actually down. BUT, the credit is not automatic (somebody has to file a claim), it only applies to the fees for the affected service, and personal and family plans are not covered at all. Dollar for dollar, a few dollars back per mailbox does not come close to covering a lost day of business. An SLA is a refund policy, not a continuity plan.
The Playbook: Keeping the Lights On When the Cloud Goes Out
So what does a real plan look like? Here are the moves that matter most; nearly all of them cost less than a single day of idle payroll.
Put a continuity layer in front of email. Many email security platforms can sit between the internet and Microsoft 365 at the mail-routing (MX) level. When Exchange Online goes dark, that layer holds incoming mail safely and gives users an emergency inbox in a web browser, where they can keep reading, replying, and sending. When Microsoft recovers, the held mail flows back into the real mailboxes. For a business that lives in email, this is the single biggest difference between “we lost a day” and “we barely noticed.” (One caution: API-only email filters that plug directly into Microsoft 365 generally cannot do this; they depend on the very service that is down.)
Keep working copies local. Classic desktop Outlook keeps a synchronized copy of the mailbox on the PC; meaning, users who had it could still pull up mail that had already been downloaded while the servers were struggling. The same idea applies to OneDrive and SharePoint files that are set to always stay on the device. It is a small configuration choice that pays off enormously — but only if somebody verified it BEFORE the outage.
Have a second way to talk. Teams messages and calls stayed up during the November outage, but Teams calendars did not — and Teams had its own, separate hour-long incident in the Americas on the morning of November 12. In July 2026, it was on the casualty list outright. Every business needs one communication path that does not share a fate with its primary platform — a group text list, a conference bridge, a secondary chat tool — plus a contact sheet for staff and key clients that lives somewhere other than the mailbox. New York City’s “call us instead” advisory was the right instinct; the trick is deciding on that fallback before it is needed.
Keep an independent backup. Microsoft’s own services agreement recommends that customers regularly back up the content they store in its services. A third-party backup that lives outside of Microsoft’s cloud is first and foremost insurance against data loss; it also means critical files and mailboxes exist somewhere reachable when the front door is locked.
Know how your security tools fail — and expect the scams. When Safe Links cannot check a link, does the click go through, or does it get blocked? Decide (and document) whether each security control fails open or closed before an outage, not during one. And warn people; researchers at Sophos cautioned that scammers could use the confusion to pose as Microsoft, or as an affected business, offering a fix. An outage is a terrible time to click on an unexpected link.
Watch the status so your people don’t have to. Within minutes of the first alert on November 12, the right message to staff was a simple one: it is Microsoft, not you; here is what still works; and please do NOT start reinstalling Outlook or resetting passwords (the problem was on Microsoft’s servers, not on anyone’s laptop). A good MSP tracks the incident numbers, keeps an eye on Microsoft’s public status page (status.cloud.microsoft) when the admin center itself is unreachable, translates the updates into plain English, and tells clients when to expect the next one. That alone saves hours of collective frustration.
Give the help desk a back door. If the only way to reach IT support is by email, an email outage takes the support line down with it. (The ripple reached third-party tools, too; Zendesk reported mail delivery issues with its Microsoft Exchange connector, and Twilio reported delays delivering email to Microsoft-hosted inboxes.) A phone number, a ticket portal, and a status page hosted somewhere else should all be on the menu — and yes, that goes for the MSP’s own tools as well.
Write the runbook, then rehearse it. A one-page “email is down” plan — who makes the call, who notifies clients, which fallback gets switched on — turns a scramble into a checklist. Walk through it once or twice a year; a plan nobody has practiced is just a document.
File the claim. Service credits are requested through a support ticket in the Microsoft 365 admin center, citing the incident IDs (EX1550543 and MO1550821). The deadline is one month after the close of the billing month that contained the incident; for most customers hit by the November 2026 outage, that means the end of December 2026. If the licenses were bought through a partner, the request goes through that partner (and some set earlier cutoffs) — exactly the kind of paperwork an MSP should be handling for its clients.
There is a lesson here for those of us on the provider side of the table, too. The November outage started with a change; so did the one in July 2026. MSPs make changes to client environments every week, and the same discipline applies to us — staged rollouts, real maintenance windows, and a tested way to roll back.
Barring any extenuating circumstances, a business with the list above in place rides out the next outage with a shrug instead of a lost day. Helping clients get there is exactly the kind of work Hardwired’s managed services exist to do.
What Comes Next?
The next milestone is Microsoft’s post-incident report, which should spell out how a flaw from an earlier upgrade went unnoticed, why the first rounds of fixes did not hold, and what is being done to keep it from happening again. The questions will not stop there. When a single email platform can stall a national health service, a jury trial, and a city’s agencies in the space of a day — and when a national government has to activate a formal agreement to secure its place in the repair line — resilience stops being an IT topic and becomes a public one. Within days of the outage, lawmakers outside the U.S. were arguing that it shows the risk of leaning so heavily on American tech.
Will businesses walk away from Microsoft 365 over this? Absolutely not — and they shouldn’t. It remains a remarkable platform, and every major cloud provider has had its own bad days. BUT, even so, the conversation is changing. “What is our plan when the cloud is down?” is quickly becoming as standard a question as “do we have a backup?” — and the tools to answer it (continuity services, independent backups, offline-capable apps) are well within reach of a 25-person office. The cloud is not going anywhere; our expectations of it are simply growing up. And how Microsoft — and the rest of the industry — responds will certainly be fascinating to watch.
Sources
- Abrams, Lawrence. “Microsoft Blames Massive Microsoft 365 Outage on Maintenance Bug.” BleepingComputer, 24 July 2026, https://www.bleepingcomputer.com/news/microsoft/microsoft-blames-massive-microsoft-365-outage-on-maintenance-bug/. Accessed 13 Nov. 2026.
- Aiello, Chloe. “Microsoft Outage Cost: Confusion, an Opening for Criminals, and $7 Billion in Lost Productivity.” Inc., 13 Nov. 2026, https://www.inc.com/chloe-aiello/microsoft-outage-cost-confusion-an-opening-for-criminals-and-7-billion-in-lost-productivity/91436447. Accessed 14 Nov. 2026.
- Alger, Jordyn. “Microsoft Outage Could Lead to $7B in Losses.” Security Magazine, 13 Nov. 2026, https://www.securitymagazine.com/articles/102736-microsoft-outage-could-lead-to-7b-in-losses. Accessed 14 Nov. 2026.
- Associated Press. “Microsoft Says Services Fully Restored after Hourslong Global Outage.” Yahoo News, 14 Nov. 2026, https://www.yahoo.com/news/world/articles/microsoft-says-services-fully-restored-111052468.html. Accessed 14 Nov. 2026.
- Bowden, Zac. “‘We Are Deeply Sorry’ Microsoft Apologizes for Major Outage That Knocked Outlook, Teams, and Microsoft 365 Offline for 24 Hours.” Windows Central, 13 Nov. 2026, https://www.windowscentral.com/microsoft/we-are-deeply-sorry-microsoft-apologizes-for-major-outage-that-knocked-outlook-teams-and-microsoft-365-offline-for-24-hours-insists-it-wasnt-ai-that-caused-it. Accessed 13 Nov. 2026.
- Chen, Christine. “Australia’s Spy Agency, Federal Police Hit by Microsoft Outage.” Reuters, 13 Nov. 2026, https://www.yahoo.com/news/world/articles/australias-spy-agency-federal-police-030851057.html. Accessed 14 Nov. 2026.
- Clayton, Laura. “Inside the Microsoft 365 Outage That Took Down Outlook.” UptimeRobot, 13 Nov. 2026, https://uptimerobot.com/reports/microsoft-365-outage-november-2026/. Accessed 14 Nov. 2026.
- Habib, Heba. “Microsoft Hit by Global Outages: What Happened and Why?” Al Jazeera, 13 Nov. 2026, https://www.aljazeera.com/news/2026/11/13/microsoft-hit-by-global-outages-what-happened-and-why. Accessed 13 Nov. 2026.
- Leib, Mason. “Outlook Outage Updates: Microsoft Apologizes for Email Disruption, Says There Was No Hack.” ABC News, 13 Nov. 2026, https://abcnews.com/US/outlook-down-microsoft-outage-reported-mail/story?id=138088203. Accessed 13 Nov. 2026.
- Lynton, Jenny. “Microsoft Could Owe Your Company Up to 100% Back for the Outlook Outage, but There’s a Dec. 31 Deadline.” Men’s Journal, 13 Nov. 2026, https://www.mensjournal.com/news/microsoft-could-owe-your-company-up-to-100-back-for-the-outlook-outage-but-theres-a-dec-31-deadline. Accessed 14 Nov. 2026.
- Microsoft. “Explain Service Level Agreement (SLAs) Concepts.” Microsoft Learn, https://learn.microsoft.com/training/modules/describe-support-offerings-for-microsoft-365-services/3-explain-service-level-agreements. Accessed 13 Nov. 2026.
- “Microsoft 365 Outage Disrupts Outlook, Exchange Online and Teams.” SecurityWeek, 12 Nov. 2026, https://www.securityweek.com/microsoft-365-outage-disrupts-outlook-exchange-online-and-teams/. Accessed 13 Nov. 2026.
- “The Microsoft 365 Shared Responsibility Model Explained: Who Is Responsible for Your Data?” AvePoint, 22 May 2026, https://www.avepoint.com/blog/backup/microsoft-365-shared-responsibility-model. Accessed 13 Nov. 2026.
- “Microsoft Apologizes for Massive Outage That Took Down Outlook and Other 365 Services.” CNBC, 12 Nov. 2026, https://www.cnbc.com/2026/11/12/microsoft-outage-outlook-aws-google.html. Accessed 13 Nov. 2026.
- “Microsoft Outage Live Updates: Email Service Is ‘Approaching Normal Levels’ Following Widespread Global Outage.” CNN, 12 Nov. 2026, https://www.cnn.com/2026/11/12/business/live-news/microsoft-global-outage-intl-hnk. Accessed 14 Nov. 2026.
- “MO1550821 — Users May Be Unable to Connect to Multiple Microsoft 365 Services.” Cloud Admin Hub, 14 Nov. 2026, https://www.cloudadminhub.com/microsoft-365-service-health/mo1550821/. Accessed 14 Nov. 2026.
- Parmar, Mayank. “Microsoft 365 Outage: You May Get Money Off Your Bill If You Use It for Work, but You’ll Have to Ask Microsoft for It.” Windows Latest, 13 Nov. 2026, https://www.windowslatest.com/2026/11/13/microsoft-365-outage-you-may-get-money-off-your-bill-if-you-use-it-for-work-but-youll-have-to-ask-microsoft-for-it-as-a-server-failure-takes-down-outlook-teams-and-exchange/. Accessed 14 Nov. 2026.
- Redmond, Tony. “Microsoft 365 Quarterly Uptime Number Sinks to New Low.” Office 365 for IT Pros, 22 Apr. 2026, https://office365itpros.com/2026/04/22/microsoft-365-uptime/. Accessed 13 Nov. 2026.
- Singh, Punam. “Microsoft Outlook Outage Runs into Friday as Recovery Stalls and Root Cause Stays Unconfirmed.” CIO&Leader, 13 Nov. 2026, https://www.cioandleader.com/microsoft-outlook-outage-runs-into-friday-as-recovery-stalls-and-root-cause-stays-unconfirmed/. Accessed 13 Nov. 2026.
- Speed, Richard. “Microsoft 365 Outage Drags On for Nearly 10 Hours During Bad Night for North American Infra.” The Register, 23 Jan. 2026, https://www.theregister.com/2026/01/23/microsoft_365_outage/. Accessed 13 Nov. 2026.
- StatusStack Staff. “Microsoft 365 and Exchange Online Outage, November 12, 2026.” StatusStack, 14 Nov. 2026, https://statusstack.com/blog/microsoft-365-exchange-online-outage-november-12-2026. Accessed 14 Nov. 2026.
- Swain, Gyana. “Microsoft Struggles to Get M365 Services Back Online after Widespread Outage.” CIO, 13 Nov. 2026, https://www.cio.com/article/4248249/microsoft-struggles-to-get-m365-services-back-online-after-widespread-outage.html. Accessed 14 Nov. 2026.
Would your business keep moving through a 36-hour email outage?
We build continuity into the environments we manage, so a vendor’s bad day doesn’t become yours. A free assessment tells you where the gaps are.
Book a free assessment →