Ringing a colleague sounds like the sensible fallback when email and video calls vanish on a Friday morning, right up until you go to find their number and discover it’s saved in the system that’s currently down. It’s a small and slightly daft moment, but it’s exactly the kind of problem people can run into when Microsoft’s services go offline and disrupt email, calendars and meetings.
A few people probably enjoyed the break from a meeting they didn’t want to attend, but for some businesses, the working day effectively stopped. What companies should do about it is less clear-cut, with suggestions ranging from keeping a second provider in reserve to rethinking how their systems are connected or writing a plan that fits on a single sheet of paper.
A Working Day Built On Someone Else’s Servers
The people who felt this outage most weren’t the ones pleased about a cancelled call. Matt Saunders, Field CTO at Adaptavist, says thedisruption had a disabling effect on a huge number of organisations. He also warns that attempts to restore services can face another problem when large numbers of users try to reconnect at once. He describes this as a “‘thundering herd’ problem”, where the growing queue of requests can actively damage recovery efforts. He compares it to the pressure on a ticketing website when a major music event goes on sale.
Working around the outage presents its own challenge: “People can use the old-fashioned phone for meetings, but that’s challenging when everyone’s numbers are in a system that is also offline.”
Paul Adams, Group Technical Director at Utilize, points to habit rather than the outage itself: “The biggest risk for businesses isn’t necessarily that a provider like Microsoft experiences an outage; it’s that so many organisations have built their entire working day around the assumption that these services will always be available.”
When email, calendars, documents and chat all sit within the same ecosystem, a disruption can affect far more than one application. It can also leave a business unable to tell customers what’s happening, a problem Mel Owen, Founder of TimeToPost, highlights: “If your mail, calendar and documents all sit with one provider, an outage also takes out your ability to tell customers what’s going on.”
Matt Falconer, CEO and Co-Founder of Ares Intelligence, believes the samedependence is building around AI tools. He sees the Microsoft outage as a reminder of what losing access to essential technology could look like: “The current Microsoft outage is a technical failure, not a product being withdrawn, but it’s a useful rehearsal for what a withdrawal would look like.” Businesses are handing more work to AI models they don’t own or control, he argues, meaning access could disappear through a technical fault, a commercial decision or a geopolitical one.
A Spare Provider Won’t Help If You Can’t Log In to It
Keeping a second service in reserve sounds sensible, but different productivity suites aren’t necessarily interchangeable. Saunders explains:
“It’s not just a problem of synchronising data so you can use the other when one is down; any company of any size will not be able to transparently switch from one to the other, because the offerings are different and no one designs systems to interoperate with both.”
Vladimir Beskorovainyi, Enterprise AI Architect and CTO at Besk Tech, points to another weak spot: the sign-in process. When several systems rely on the same identity provider, losing access to it can lock employees out of those systems at the same time. It’s like having a spare office across town but keeping its only key inside the first one. As he puts it: “Companies discover they have one dependency wearing twelve different logos.”
It’s also why he’s unimpressed by one of the usual solutions: “Multi-cloud is usually theatre.”
Rob Mueller, Founder, Director and Chief Technical Officer at Fastmail, approaches the problem from the supplier side. Productivity suites are often bundled together, which he says allowsfailures to spread across services. His alternative is to use independent products linked through open standards, allowing different providers to communicate. Mail, calendars and contacts could sit with one company, be accessed through another company’s app and connect to a chat tool or file store from a third. In his view: “A distributed ecosystem gives you a smaller blast radius on any failures, resilience through diversity of suppliers and your choice of tools rather than what one provider dictates.”
Jens Hagel, Founder and Owner of Kjello, isn’t convinced that breaking things up is the safer route. His Hamburg firm has 32 employees running IT for more than 150 mid-sized companies, and in 2013 he moved clients to Microsoft’s cloud instead of keeping Exchange servers in the basement. From his experience, a server run by a small firm fails more often than Microsoft’s services do. His conclusion being: “I would make that call again. Leaving Microsoft is not the answer for most companies.”
More from News
- How Far Can COP31’s Antalya AI Pledge Go Without Binding Rules?
- Jensen Huang Is Out Amid Spy Allegations. How Bad Could The Fallout Get For Nvidia, AI And Global Chip Supply?
- Brits Want AI to Help Not Take Over New Research Finds
- Australia Is Drafting An AI Safety Regulator After An OpenAI Model Hacked A Government Site – What Will It Be Able To Enforce?
- Waymo’s Software Fix Was Supposed To Make Robotaxis Safer – Why Are Regulators Still Concerned?
- Uh Oh, Elon, Texas Just Turned Blue. Will Tesla Hit The Road Again, Or Will Big Tech Have To Adapt?
- Huntress Offers Breached Businesses 60 Days Of Free Protection To Stop Hackers Coming Back
- Finite State Raises $27 Million To Help Device Makers Fight ‘Vulnpocalypse’ With AI-Powered Product Security
A One-Page Plan For When Teams And Email Disappear
Where Hagel does see a gap is in what businesses do while a service is down: “What worries me more is that few firms have written down what they do while a service is down.”
In three recent infrastructure reviews, he found that not one company had a written emergency plan. He says a single page is enough to get started, covering who makes decisions, how staff reach each other when Teams and email are gone and which tasks need to carry on by phone or on paper.
Owen’s advice is to identify the three services that would stop revenue coming in if they disappeared for a day, then write down a manual fallback for each. That includes another way to reach customers and somewhere to keep important files that can be opened offline. As she puts it: “You don’t need to leave the big providers. You need to stop assuming they’ll never go down.”
Adams adds that alternativecommunication channels, secure backups and a tested continuity plan don’t need to be complicated or expensive. The mistake, he says, is waiting until something goes wrong: “The real mistake is waiting until an outage happens to work out what to do.”
Testing The Plan Before Another Outage Does
A plan on paper only gets a business so far. Beskorovainyi advises businesses to know which systems can still authenticate users without their usual provider and keep an emergency access route that doesn’t depend on the systems they’re trying to survive losing. Then comes the step many plans skip: “And test it quarterly, because an untested failover is a plan, not a capability.”
Owen also recommends testing the fallback every quarter, rather than trying to work it out during an outage. Having an alternative written down doesn’t necessarily mean employees will know how to use it.
Saunders acknowledges that preparing for this is difficult: “An outage like this almost throws us back to the dark ages, so it’s hard to plan for these systems not being there.” His suggestion is to start with a small habit: each time you use an online system, think briefly about how you’d get that task done without it.
What Your Business Would Do Without Its Inbox
Walking away from Microsoft altogether isn’t the answer for most businesses, according to Hagel (and most would probably agree). The challenge is understanding where reliance on one provider creates risk and having a plan for working without the systems a business already uses.
That doesn’t necessarily call for a huge budget or an entirely new provider – it starts with identifying what can’t stop, writing down the alternatives and testing whether they actually work.
Our Experts:
- Matt Saunders: Field CTO at Adaptavist
- Matt Falconer: CEO and Co-Founder of Ares Intelligence
- Vladimir Beskorovainyi: Enterprise AI Architect & CTO and Founder of Besk Tech
- Mel Owen: Founder of TimeToPos
- Rob Mueller: Founder, Director and Chief Technical Officer at Fastmail
- Jens Hagel: Founder and Owner of Kjello
- Paul Adams: Group Technical Director at Utilize
Matt Saunders, Field CTO At Adaptavist
“Whilst many employees have reacted gleefully at not being able to log into a Friday morning online meeting, this outage affecting email, calendaring and videoconferencing is having a disabling effect on a huge number of organisations, which have ground to a halt without Microsoft’s services.”
“Classic IT disaster planning strategies suggest using a second service for resilience, but in practice, it’s almost impossible for a business to function without tools such as Microsoft. It’s not just a problem of synchronising data so you can use the other when one is down; any company of any size will not be able to transparently switch from one to the other, because the offerings are different and no one designs systems to interoperate with both.”
“People can use the old-fashioned phone for meetings, but that’s challenging when everyone’s numbers are in a system that is also offline.”
“These systems operate on a massive scale – and it’s clear that attempts to fix the issue have caused a ‘thundering herd’ problem – the increasing queue of requests from people trying to get back online actively damages the fix attempts, similar to the compounding problems we see when a big music event goes live on a ticket seller site.”
“An outage like this almost throws us back to the dark ages, so it’s hard to plan for these systems not being there. But each time you’re interacting with an online system, think a little bit about how you’d get that task done without it, so the world doesn’t stop turning when something goes down.”
Matt Falconer, CEO And Co-Founder Of Ares Intelligence
“The current Microsoft outage is a technical failure, not a product being withdrawn, but it’s a useful rehearsal for what a withdrawal would look like. Disruption has been reported in more than 70 countries, and the likes of OpenAI and Anthropic have even flagged knock-on issues.”
“Businesses are handing a growing share of their work to AI models they don’t own, can’t control and could lose access to overnight, whether through a technical fault, a commercial decision or a geopolitical one. No board would accept its workforce failing to turn up with no warning – yet that’s the risk many are carrying without a plan.”
“Every leader should be asking: if our main AI provider went offline tonight, what would we do on Monday morning? That means knowing which processes depend on which models, having a fallback ready, and treating sovereign or self-hosted options as a business-continuity question rather than a nice-to-have.”
Vladimir Beskorovainyi, Enterprise AI Architect & CTO And Founder Of Besk Tech
“The dependency most businesses miss is not the service they chose. It is identity.”
“When a provider like Microsoft has a bad day, the visible failure is email or Teams. The expensive failure is sign-in, because every unrelated system federated to that identity provider stops accepting logins at the same moment. Companies discover they have one dependency wearing twelve different logos.”
“Multi-cloud is usually theatre. The compute is duplicated, the identity layer is not, and the failover has never been run under load.”
“Practical contingency is narrower than people expect. Know which of your systems can authenticate without the provider. Keep one break-glass access path that depends on nothing you are trying to survive the loss of. And test it quarterly, because an untested failover is a plan, not a capability.”
Mel Owen, Founder of TimeToPost
“Every business now runs on someone else’s servers, and that’s fine until an outage shows you never decided what happens when they stop.”
My rule as a founder: list the three services that would halt revenue if they vanished for a day, and write down the manual fallback for each before you need it. Email is the classic blind spot. If your mail, calendar and documents all sit with one provider, an outage also takes out your ability to tell customers what’s going on.”
“Keep a second channel for customer updates, keep critical files somewhere you can open offline, and test the fallback once a quarter, not during the outage. You don’t need to leave the big providers. You need to stop assuming they’ll never go down.”
Rob Mueller, Founder, Director And Chief Technical Officer At Fastmail
“Modern productivity suites are more and more bundled as all-in-one. Tight integration is convenient on a good day, but on a bad day, it means failures cascade.
“The alternative is to connect independent products through open standards, rather than relying on one proprietary stack. Email is the original example of this. It was built as a federated system, where any provider can talk to any other. Your mail, calendar and contacts can live with one provider, be read in another company’s app, and connect to a chat tool or file store from a third. Each piece can be chosen on its merits, and each piece can fail without taking the others down.”
“A distributed ecosystem gives you a smaller blast radius on any failures, resilience through diversity of suppliers and your choice of tools rather than what one provider dictates. You also get real portability because your data sites in open formats behind open protocols, so you can move, back up or mirror it however you want”
Jens Hagel, Founder & Owner of Kjello
“I founded hagel IT-Services in Hamburg in 2004. Our 32 people run IT for more than 150 mid-sized firms, and in 2013 I decided we would move clients to Microsoft’s cloud instead of keeping Exchange servers in the basement.”
“I would make that call again. In my experience, a server run by a small firm fails more often than Microsoft does, and when it fails there is rarely anyone around who can fix it. Leaving Microsoft is not the answer for most companies.”
“What worries me more is that few firms have written down what they do while a service is down. In three recent infrastructure reviews we counted, not one company had a written emergency plan. One page is enough: who decides, how staff reach each other when Teams and email are gone, and which tasks have to keep running on paper or by phone.”
Paul Adams, Group Technical Director At Utilize
“The biggest risk for businesses isn’t necessarily that a provider like Microsoft experiences an outage; it’s that so many organisations have built their entire working day around the assumption that these services will always be available.”
“Cloud platforms have transformed how businesses operate, offering flexibility, collaboration and efficiency. But convenience can create complacency. When email, calendars, document storage and communication all sit within the same ecosystem, a disruption can quickly affect far more than one individual application.”
“I don’t think the answer is for businesses to abandon major cloud providers. It’s about recognising that no provider can guarantee uninterrupted service and planning accordingly. Every organisation should know which processes are business-critical, how employees will communicate if their usual tools go offline, and how they will access essential information.”
“Having alternative communication channels, secure backups and a tested business continuity plan doesn’t need to be complicated or expensive. The real mistake is waiting until an outage happens to work out what to do. Resilience should be part of a business’s IT strategy, not an afterthought.”
