loader image

Friday the 13th, and the day the inbox went quiet

Posted on by William
Timeline of the outage: Microsoft 365 failing from Thursday 4.15pm, Claude's Microsoft link failing at 5.52pm, a separate Claude sign-in problem Friday morning, Claude slowing from 1.09pm, and Microsoft reporting improvements at 3.05pm, with handwritten notes beside each.

It’s Friday the 13th. Microsoft 365 has been down since yesterday afternoon, Claude has been having a wobble of its own, and I’m writing this with one eye on the window for black cats. Coincidence? Almost certainly. But it’s been an interesting day, so here goes.

What happened

Just after quarter past four on Thursday afternoon, Microsoft 365 started falling over. Outlook, Teams, OneDrive and Copilot all went with it. By Friday afternoon it had been going for more than 20 hours. The timeline is in the picture at the top of this post (open it full size).

Microsoft says there’s no sign of an attack. Some of its servers became unavailable, the rest got swamped picking up the slack, and it snowballed from there. The BBC has been running a live page all day if you want the full picture.

Around 430 million people use Microsoft’s tools, so this wasn’t a small blip. Schools, GP surgeries, councils and police forces were all asking people to pick up the phone instead of emailing. One former Skype and Uber engineer reckoned it could be the “most disruptive email-related business outage of all time”. Microsoft’s share price, meanwhile, barely flinched. Nice for some.

The day our inbox went quiet

We run on Google Workspace, so our own email kept working. What changed was everyone else. Friday is usually when the inbox goes mad, with everyone trying to get things signed off before the weekend. This Friday, a whole bunch of clients just fell silent.

Illustration comparing a busy normal Friday inbox with an empty one this Friday, and a note about a few relaxed phone catch-ups instead.

So I rang them. And honestly? It was lovely. A few proper, relaxed catch-ups, the kind you never get round to when everything’s flying about by email. A couple of things didn’t quite make it out of the door, which is a shame, but nobody was cross about it. And I didn’t get invited to a single Teams call on a Friday afternoon, which I’m counting as a win.

Before I get too smug

Here’s the confession. I lean on Claude, the AI assistant, a lot. It reads and drafts emails, helps run projects and keeps track of what’s in flight across the studio. It’s brilliant, and it’s also a big chunk of my business resting on one company I don’t control.

Today it had its own bad day. Its link to Microsoft 365 broke along with Microsoft, and separately it had sign-in problems in the morning and slowed right down after lunch. We got a good morning’s work done before it went quiet, but by the afternoon I couldn’t ask it to check or reply to my email. I had to type them myself. Like an animal.

And to be fair to Microsoft, Google has bad days too. So does Apple. So does everyone. Which is really the point.

It’s not about picking the “right” company

There’s a name for this: concentration risk. When your email, files, calendar, video calls and now your AI all come from one place, one fault takes them all out at once.

A business consultant quoted by the BBC made a sensible point. Using one big provider isn’t the problem in itself. The problem is not knowing what depends on it, and having no plan B for the things that really matter.

Email is also more than email now. As the security expert Troy Hunt pointed out, it’s how you get login codes, confirm orders and reset passwords. Lose it, and a load of other things quietly stop working too.

Five things worth doing this week

Checklist of five ways to protect a small business from outages: know what you depend on, have a plan B, keep a second way to reach people, keep copies elsewhere, and watch for fake IT support.

  1. Know what you depend on. Spend ten minutes listing the tools your business can’t run without. You’ll probably be surprised how short, and how wobbly, that list is.
  2. Have a plan B for the two or three that really matter. Not for everything. Just the things that would stop you trading for a day.
  3. Keep a second way to reach people. A phone number clients know, a WhatsApp group, a line on your website saying “if email’s down, ring us”. Everyone who coped today had one.
  4. Keep copies of the important stuff somewhere else. Client contacts, current quotes, the files you’re working on this week. If they only live in one cloud, they’re only as available as that cloud.
  5. Be extra careful while everyone’s rattled. Banks and universities were warning people about fake “IT support” messages offering to fix the outage. Outages are prime time for scammers. If you missed our post about the inflatable giraffe, it’s the same trick in a different hat.

And finally

By Friday afternoon Microsoft said email queues were getting back to normal, which means a lot of people are about to receive a whole day’s worth of email at once. Spare a thought for them on Monday.

If you’re not sure how exposed your business is, or you’d just like a second pair of eyes on your setup, give us a ring. We’ll pick up. Promise.