Having a backup is not the same as being able to work. A backup protects your data. It does nothing to help you send an invoice, answer a customer or join a call while Microsoft’s servers are down, as they have been since Thursday afternoon in this week’s outage.
What keeps a small business working is a short plan in four layers. The first costs nothing but about an hour of setup, and the second little or nothing. Layers 1 to 3 keep you working during an outage; Layer 4 protects your data. Below we cost each one, work out when the paid layer pays for itself, and give you a one-page plan to copy.
Layer 1: offline copies, alerts and a one-page plan (free)
This is an hour’s work for whoever looks after your IT.
- Make sure Outlook keeps a local copy. Classic desktop Outlook stores recent mail on the computer, so staff can keep reading it offline. By default that is the last 12 months on most PCs, less on computers with small disks (Microsoft Learn). New Outlook for Windows saves fewer days by default; raise the number under Settings, General, Offline (Microsoft).
- Pin the files you cannot work without. In OneDrive or a synced SharePoint library, right-click the folder and choose Always keep on this device. The files then stay on the laptop “even when you’re offline” (Microsoft). Think price lists, live job folders, templates and the current month’s invoices.
- Keep key contacts outside Microsoft. Your top customers, suppliers, bank, IT support and insurer, with phone numbers, in a printed sheet or on a phone. If Outlook cannot connect, contacts saved only in Microsoft 365 may be out of reach too.
- Set up alerts that reach you. In the Microsoft 365 admin centre, Health > Service health > Customize > Email sends incident alerts to up to two addresses per admin (Microsoft Learn). Make one of them an address that is not on Microsoft 365.
- Write the one-page plan at the end of this article, print it and, if you have staff, walk through it once.
Layer 2: a second way to reach people (free to low cost)
The real damage in an email outage is silence. Decide now what you will use instead, so nobody improvises on the day:
- For customers: the phone, a notice on your website, and the social accounts they follow. If you would use WhatsApp Business or texts, set them up on business-controlled accounts now, not on someone’s personal phone in a panic. Data protection rules apply either way.
- For staff: a group chat outside Microsoft, or simply a phone tree. If Teams is your only internal channel, a Teams problem leaves you with nothing.
- For customers who rely on you daily: a simple status page hosted away from Microsoft.
Layer 3: an email continuity service (from about £3.75 a user a month)
This is the layer that actually keeps email working. A continuity service sits in front of your Microsoft 365 mailbox, usually as part of an email security product you might already need. When Microsoft cannot accept mail, it holds incoming messages and gives staff an emergency inbox to read and reply from, then passes everything back when Microsoft recovers:
| Service | What it does in an outage | Indicative cost |
|---|---|---|
| Barracuda Email Protection (Advanced, Premium and Premium Plus tiers) | Spools mail for up to 96 hours with an emergency mailbox | £45 to £95 per user per year on G-Cloud, about £3.75 to £7.92 a month |
| Proofpoint Essentials | Emergency Inbox is “a standard product feature, and is always on”; holds mail while your server is unreachable | Varies by reseller |
| Mimecast | Admin triggers continuity; users send and receive through Mimecast’s apps and web portal | Varies by bundle |
Sources: Barracuda and its G-Cloud price list (10-user minimum, excluding VAT), Proofpoint, Mimecast. Barracuda’s G-Cloud listing, a public-sector catalogue, was the only published UK list price we found for continuity; Proofpoint and Mimecast sell through resellers, so ask for a quote.
Two-minute check: do you already have it? Look up your domain’s MX record, the address other mail servers deliver to, with any free online “MX lookup” tool. If it ends in mail.protection.outlook.com (Microsoft Learn), or mx.microsoft on some newer setups (Microsoft Learn), your mail goes straight to Microsoft, with no gateway in front. If it points to a Mimecast address (mimecast.com), Proofpoint Essentials (ppe-hosted.com) (Proofpoint) or Barracuda (barracudanetworks.com) (Barracuda), you already pay for a gateway, probably through your IT provider. Ask which tier you are on and whether it includes continuity before you buy anything new.
Would it have helped this week? It depends on how the service switches on. This outage was mostly about reaching mailboxes, not about Microsoft refusing mail: Microsoft’s own updates describe it working through “email queues and backlogs” and extending its email timeout to 48 hours “to help protect delayed messages” (Microsoft, Microsoft 365 Status), which points to mail being accepted and held rather than refused. A gateway that only steps in when Microsoft stops accepting mail, as Proofpoint’s Emergency Inbox does when your server is unreachable, may not have kicked in for much of it. A continuity event that an administrator starts, as with Mimecast, opens the emergency inbox as soon as it is started, whether or not Microsoft is still accepting mail (Mimecast). Ask your provider which model you have, and who on your side can start it at 6pm on a Thursday.
Four questions to ask before you buy:
- Is it already in place? Continuity depends on your incoming mail already being routed through the provider, which means changing your domain’s MX records. That takes planning and testing, so set it up now rather than during the next outage.
- How do staff sign in to the emergency inbox? If it uses their Microsoft 365 sign-in, ask what happens when Microsoft’s sign-in is part of the outage.
- How long does it hold mail? Barracuda’s continuity switches itself off after 96 hours. By Friday lunchtime this week’s outage had been running for almost 20 hours.
- Which tier includes it? Continuity is not in every bundle, so check the specific product, not the brand.
What not to buy: a plain “backup MX” mail server. Senders’ servers already hold your mail and retry for hours or days, so a second server that only stores mail you still cannot read adds little. The value is in the emergency inbox.
When an emergency inbox pays for itself
Work it out for your own firm with one line:
Cost of a day without email = staff x hours in the day x hourly cost x share of the work that needs email
For a 10-person office on the typical UK wage (median hourly pay for full-time employees, excluding overtime, was £19.67 in April 2025 (ONS)), where half the work depends on email:
| Cost | |
|---|---|
| A day without email: 10 x 7.5 hours x £19.67 x 50% | £738 |
| Barracuda continuity for 10 users for a year, cheapest tier (£45 each) | £450 |
| Barracuda continuity for 10 users for a year, dearest tier (£95 each) | £950 |
So if one such day a year is the norm, the cheapest tier more than pays for itself on staff time alone, before counting a single lost order. The dearest tier needs about 1.3 such days a year. Two adjustments make the sum honest. The wage figure is pay only: add employer’s National Insurance and pension and each hour costs you more, which favours buying. But an emergency inbox does not recover every lost minute, which favours waiting. If your share of email-dependent work is nearer a quarter, the cheapest tier needs about 1.2 lost days a year to pay its way.
Two more limits apply. If most of your customers phone or walk in, a day without email costs you far less than half its staff time, and Layers 1 and 2 may be enough. And an emergency inbox keeps email going, but it does nothing for Teams calendars or SharePoint sharing, which also failed this week.
Layer 4: back up the data (about £3.50 a user a month)
A backup does not help during an outage, but it is the answer to a different and nastier problem: mail or files deleted by mistake, by a departing employee or by ransomware. Microsoft’s own safety nets are time-limited: once it leaves Deleted Items, deleted email stays recoverable for 14 days by default, and an admin can raise that to 30; keeping it longer needs a retention policy or hold (Microsoft Learn). Exchange Online on its own “does not provide a way to perform a traditional backup of mailboxes” (Microsoft Learn), which is why Microsoft now sells Microsoft 365 Backup as a separate, paid service. The NCSC tells small organisations to back up essential data and “make backing up part of your everyday business” (NCSC).
BackupVault, a UK provider, quotes £3.50 per user per month for a 10-person firm, excluding VAT, falling to £1.75 at 500 users (BackupVault). Microsoft 365 Backup is priced by storage instead, at a list price of US$0.15 per gibibyte of protected content per month (Microsoft Learn).
Switching provider swaps one risk for another
Every provider fails sometimes. Microsoft 365 itself had a roughly 10-hour outage in North America in January 2026 (The Register). An Amazon Web Services fault in October 2025 knocked out apps and websites for about 15 hours (AWS), and Microsoft’s Azure Front Door failed later the same month, taking parts of Microsoft 365 with it (ThousandEyes). Moving everything to another provider swaps one single point of failure for another. A plan that assumes your main system will go down, and says what you do next, is worth more than a new supplier.
A one-page outage plan to copy
Print this, fill in the blanks and keep a copy somewhere that does not need email to open.
When email or Microsoft 365 goes down
Who decides it is an outage: [name], backup [name]. Check status.cloud.microsoft first.
Within 15 minutes: [name] changes the voicemail greeting and adds the website notice below. Nobody reinstalls Outlook or resets passwords.
Voicemail greeting: “Thanks for calling [business]. Our email is delayed by a Microsoft outage. For anything urgent, please leave a message or call [number] and we will ring you back today.”
Website and social notice: “Our email is affected by the Microsoft 365 outage. Messages should arrive late rather than be lost. For anything urgent, call [number].”
Who rings whom: [name] calls customers with live orders or payments due this week; [name] calls key suppliers. Contact sheet is kept [where].
Payments: no change of bank details is accepted during an outage without a call-back on a number we already hold.
Staff channel: [WhatsApp group, phone tree or other non-Microsoft channel].
Afterwards: save Microsoft’s incident report, log the hours lost, and check whether a service credit is due.
The NCSC’s free Cyber Action Toolkit covers the cyber security basics that sit alongside a plan like this.
If you are still dealing with this week’s outage, start with what to do when email is down and how to claim a service credit.



