23 September 2026

How to Keep Your Business Running When Microsoft 365 Goes Down

Photo by geralt on Pixabay

Most small businesses run almost everything through a handful of online services now. Email, files, accounting, bookings, payments, it's all sat in the cloud rather than on a machine in the office. That works brilliantly right up until one of those services stops responding and nobody in the building can do anything about it.

You don't need a hacker for that to happen. On 19 July 2024, a faulty update from the security vendor CrowdStrike caused around 8.5 million Windows devices worldwide to crash within hours, according to Microsoft's own figures. Flights were grounded, banks and hospitals lost systems, and none of it involved an attacker at all, just one bad software update rippling out through machines that all trusted the same vendor.

Microsoft 365 has had its own outages too, days where Outlook or Teams stop responding for hours at a time with no warning. When that happens, there's genuinely nothing on your end to fix. The provider is working on it, and everyone downstream just waits.

Being with a big name doesn't make you immune

It's tempting to assume that a service as large as Microsoft 365, or your bank, or your accounting platform, is too well-resourced to ever go down. Scale doesn't work that way. A big provider has more customers depending on it, not fewer failure points, and when something does go wrong, the outage tends to be more widely felt precisely because so many businesses lean on the same handful of platforms.

The practical takeaway isn't to distrust the big providers. It's to accept that any online service you depend on, however reputable, will occasionally be unavailable, and to have already decided what your business does during that gap.

What actually breaks when a core service goes down

The disruption is rarely subtle. Staff can't reach email or shared files, so anything that needed those grinds to a halt. If your booking or payment system is the one affected, you can't take money or appointments until it's back. Customers who are trying to reach you can't, and you've got no easy way to reach them either. People end up sitting idle rather than working around the problem, mostly because nobody agreed in advance what "around the problem" looks like.

None of this needs a security incident, a breach, or anything dramatic. It just needs one ordinary Tuesday where the wrong service has a bad day.

The one-page plan

A continuity plan for outages doesn't need to be a formal document nobody reads. It needs to answer five questions, and it needs to be reachable even when the system it's protecting against is the one that's down.

• Which tools would actually stop the business? List the small number of services that halt real work if they're offline, such as email, your payment system, or your booking tool. Leave off anything you could realistically live without for a day.
• How does the team stay in touch without it? Agree a backup channel that doesn't depend on the service that's down, whether that's phone numbers, a group chat on a separate app, or plain text messages.
• What do people need to be able to reach offline? A copy of your customer contact list, key phone numbers and anything you'd need mid-crisis, kept somewhere that doesn't require the internet. A printout or a phone's own contacts app is enough.
• Who's actually in charge during an outage? Decide in advance who makes the call and who keeps customers updated. Agreeing this ahead of time means people act, rather than waiting around for someone to tell them what to do.
• Where do you check, and who do you call? Know the provider's status page for spotting a wider outage, and know who to call for help. For most businesses we work with, that call goes to their IT provider rather than to Microsoft directly.

What to do the moment it happens

1. Call your IT provider, if you have one. They can usually tell you within minutes whether it's a known, wider outage or something specific to your setup, and they're the ones best placed to help you get moving again.
2. Check the status page yourself if you're waiting, or don't have anyone to call. If it's a confirmed outage, there's nothing to fix locally, so stop trying to reboot the router.
3. Tell the team what's down and what to use instead. This saves an hour of people independently trying to troubleshoot a laptop that was never the problem.
4. Move to your backup way of staying in touch, so the team can keep coordinating while the primary system is out.
5. Tell customers if it affects them, and let them know roughly when to try again.
6. Note when it started and what was affected. A couple of lines is enough, and it makes it much easier to follow up afterwards or spot anything that needs attention once things are back.

Making the next outage hurt less

A handful of things, mostly set up once, make a real difference when this happens:

• Keep recent files available offline. With Known Folder Move turned on, OneDrive can sync a device's Desktop and Documents folders automatically, so those files stay reachable even while the service itself is degraded. We've covered how to decide what actually belongs in OneDrive versus SharePoint in more detail if your team has never had that split explained.
• Have a backup way onto the internet. A mobile hotspot from a phone can get a few key people working again if your main office connection is what's actually down. If that's a recurring worry rather than a one-off, a dedicated data SIM for a spare device is a steadier version of the same idea, since it isn't competing with anyone's personal data allowance.
• Bookmark your status pages. The Microsoft 365 status page, and pages for whatever other core tools you rely on, are the fastest way to confirm whether it's them or you.
• Keep essential contacts off the cloud. Important numbers stored somewhere that doesn't need the internet, a printout or a phone's own contacts list, save a lot of stress mid-outage.
• Ask your IT provider about proactive alerts. Many can flag a wider outage before your customers start telling you about it.

An outage is a different problem from an attack, and worth telling apart: if what you're actually dealing with is a breach or ransomware rather than a provider outage, our step-by-step guide to what to do in a cyberattack covers that scenario instead.

A plan nobody has tested is a guess

Redundancy only counts once it's actually been proven to work. One of our clients, Mindvision, had paid for dual-internet failover for years before we started looking after their IT, without anyone ever finding out it had been configured wrong. The switch was there, but it simply didn't kick in when their main connection dropped. Nobody knew until it mattered. Whatever you've got in place for an outage, whether that's failover internet, synced files, or a one-page plan sitting in a drawer, treat "we set it up" and "we know it works" as two different things, and test the second one on purpose rather than finding out by accident.

Keep it to one page, and revisit it

Write the plan down on a single page you can reach without your main systems, whether that's a printout in a drawer or a note in a separate app. Cover your critical tools, your backup channel, where the essentials live, who's in charge, and who to call. Review it once or twice a year so the names and numbers are still right when you actually need them.

If your business has never had this conversation, it's a short one to have. Our ongoing IT support includes exactly this kind of planning, and we're happy to talk it through with you if you'd rather have someone else map it out.