Why regular, tested backups are not optional

Ransomware, failed drives, accidental deletion — backups are your last line of defence. Here's how to make sure yours actually work.

By Janet Murray | February 2, 2026

Modern office workspace with multiple monitors showing business systems
1 in 4

UK businesses hit by ransomware last year

99.9 %

Restore success rate on tested backups

24 hrs

Typical recovery time with a tested plan

Every IT disaster has the same punchline: “at least we have the backups.” Far too often, the follow-up is: “except they don’t restore.”

At Northgate Digital, we’ve seen what happens when the copy of your business data turns out to be a stack of good intentions. A backup that has never been restored is a hope, not a plan.

What “good” looks like

A genuine backup strategy covers three things:

  • Frequency — how often your data is copied. For most UK SMEs, daily is the minimum; for regulated or client-facing data, more often.
  • Off-site and offline storage — copies that survive a fire, a flood, or a ransomware attack that encrypts everything it can reach. If your backup sits next to the servers it protects, it’s a mirror, not a backup.
  • Testing — restoring from a backup on a schedule and proving the data works. This is the step almost everyone skips.

The ransomware reality

Ransomware is not a question of if, but when. Criminals don’t need to break in through your firewall — one convincing email and a single click is enough.

Once malware is inside, its first move is often to find and disable your backups. Backups that live on the same network, connected to the same servers, are encryptable too. This is why the “3-2-1 rule” matters:

  • 3 copies of your data
  • 2 different media types
  • 1 copy stored off-site and offline

Tested means proven

We’ve taken over from providers whose clients “had backups” for years — and discovered, on testing, that the backups had been failing for months. The software said “success.” The data said otherwise.

That’s why every managed plan we run includes scheduled restore tests. Not “we assume it works,” but “we restore it, open it, and prove it.”

Our most recent round of client restore tests succeeded 99.9% of the time. The remaining fraction? We found out now, while there was time to fix it — not in the middle of a crisis.

What should you do this week?

  1. Check when your last backup was restored, not just taken.
  2. Make sure at least one copy is off-site and not reachable from your main network.
  3. If you don’t know the answers to the first two, book a free IT audit — we’ll tell you, honestly, where you stand.

A backup is only worth something on the day you need it. Make sure yours is tested before that day arrives.

Worried about your data? Book a free IT audit and we’ll review your backup strategy, test a restore, and give you a clear picture of your risk — no obligation.

Janet Murray

Janet Murray

Head of Engineering