This week's Microsoft outage and what is worth having sorted out before it happens
Since Thursday afternoon, Argentina time, Outlook, Teams and much of Microsoft 365 have stopped working in dozens of countries. Thousands of companies were left without email, without meetings and without the files they kept in the cloud. Microsoft has ruled out an attack and puts it down to a problem in its own infrastructure.
For most of those companies there was nothing to do: the problem wasn't in their systems or their equipment. That's the uncomfortable side of working with cloud services. We hand a provider something we used to run ourselves, and it's almost always a good decision, because they do it better and cheaper. But when it fails, the company stops all the same.
What does depend on each company is what happens during those hours.
What is worth having sorted out beforehand
- Knowing what you depend on. A short list of the services you can't work without (email, chat, invoicing, files, the management system) and who provides each one. Often nobody has it written down.
- An agreed backup channel. If internal chat goes down, how does the team talk? If that's decided on the day of the outage, everyone improvises a different one.
- The critical things, reachable from elsewhere. Customer contacts, the day's schedule and the documents used every day, with a copy that doesn't depend on the same provider.
- What can keep going and what waits. Some tasks don't need the service that's down. Having them identified keeps the whole team from sitting and waiting.
- A message for customers. If the outage affects service, a short, timely notice keeps the provider's problem from looking like yours.
You don't need to duplicate everything
Having a second provider for every service is expensive and, for most small businesses, it doesn't make sense. It's enough to look at which part of the operation stops if each service fails and prepare something simple for that part. A list, an agreement and a copy of what matters usually cover most cases.
This outage will be fixed in hours or days. The next one, from this provider or another, will come too. Better that it finds the company with this already thought through.
Does your company have a contingency plan for when a provider fails? If it doesn't, or you're not sure it's enough, get in touch and we'll review it together.




