Every March 31st, the IT world celebrates World Backup Day. Companies post about the importance of backing up your data. Everyone nods along.
Then April 1st arrives and nothing changes.
This year, let's do something different. Let's ask the question that actually matters.
The Only Question That Matters
When did you last successfully restore data from your backup?
Not 'when did your backup last run successfully.' Not 'do you have a backup solution.' When did you actually test recovering something?
If you can't answer this question with a specific date and result, you don't have a backup. You have hope.
Why Backup Testing Matters
We've seen it countless times:
- Backups running for years... to an empty folder because someone moved the source data
- Backup drives connected to servers... and encrypted along with everything else when ransomware hit
- Cloud backup configured perfectly... except for the critical legacy application that got missed
- Backup tapes stored securely... with no compatible drive to read them
Every one of these businesses thought they had backups. They had backup software running. They had 'successful' notifications.
They didn't have a recovery capability.
The 5-Minute Backup Audit
Right now, answer these questions:
- What is being backed up? Can you list every system, every data source, every location? Including cloud services like Microsoft 365?
- Where are backups stored? If your office burns down, are your backups gone too? Is there an offsite or cloud copy?
- Can ransomware reach your backups? Are they connected to the same network as production systems? If your server gets encrypted, can the attacker encrypt the backups too?
- How old would restored data be? If you had to restore right now, would you lose one day of work? One week? One month?
- Who knows how to restore? Is it documented? Does more than one person know the process? What if that person is on holiday when disaster strikes?
What Good Looks Like
The 3-2-1 rule:
- 3 copies of your data
- On 2 different types of media
- With 1 copy offsite
Regular testing:
- Automated daily verification (file integrity checks)
- Monthly random file restores (pick something, recover it, confirm it works)
- Annual full recovery test (simulate total loss, time the recovery)
Documented recovery:
- Step-by-step restoration procedures
- Contact details for key people
- Prioritised recovery order (what needs to come back first?)
- Tested and updated regularly
The Homework
This World Backup Day, do one thing: restore a random file from backup.
Pick any file from any time in the past 3 months. Ask your IT provider (or your internal IT person) to restore it. Time how long it takes. Check that the restored file is correct and usable.
If this takes more than an hour, or can't be done at all, you've learned something valuable. Better to learn it now than when you desperately need a restore.
What About Microsoft 365?
Here's a surprise for many businesses: Microsoft doesn't back up your data.
Microsoft 365 provides availability (their service stays running) but not point-in-time recovery. If someone deletes files, or ransomware encrypts your SharePoint, or a leaver's mailbox gets purged, Microsoft can't get it back.
Microsoft themselves recommend third-party backup. This is in their shared responsibility documentation.
If you use Microsoft 365 without additional backup, you're at risk.
Getting This Right
Backup isn't glamorous. Nobody gets excited about it. But it's the difference between 'that was a scary day but we recovered' and 'that was the end of the business.'
Our Backup & Disaster Recovery service includes:
- Proper 3-2-1 architecture
- Automated verification with actual test restores
- Microsoft 365 backup included
- Documented, tested recovery procedures
- Known recovery time objectives