Your Backup Probably Wouldn't Survive a Real Failure: Lessons from a Computer Repair Workshop
Every week, someone walks into a computer repair workshop and says the same sentence: "I thought everything was backed up." Sometimes it was. More often, the backup existed on paper but failed in practice. After years of diagnosing dead laptops, failed drives and accidental formats in Geneva, I see the same six patterns repeat. None of them need expensive tools to fix, only a little attention.
1. A copy on the same computer is not a backup
The most common "backup" I see is a folder called Backup on the same drive, or a second partition on the same disk. It feels safe, but it shares every risk with the original: the same drive failure, the same malware, the same spilled coffee. A real backup has to live on separate hardware.
2. The external drive that never leaves the desk
An external disk that is always plugged in is better than nothing, but it is exposed to the same power surge, theft, fire or ransomware as the computer next to it. Malware can encrypt every connected drive in the same sweep. Disks also age, and many people discover only during a crisis that the drive stopped working months ago.
3. Cloud sync is not the same as backup
Services like OneDrive, Google Drive and iCloud mirror your files across devices. That is convenient, but it also means a deleted or corrupted file is deleted or corrupted everywhere. Many of these services offer version history or a recycle bin, but only for a limited time, and the limits differ by provider and plan. Check yours before you need it. An account lockout can also cut you off from your own files at the worst possible moment.
4. Encryption without a saved recovery key
Disk encryption such as BitLocker on Windows or FileVault on Mac is good security practice. The problem appears after a hardware repair or a firmware change, when the system may ask for a recovery key. If that key was never saved, the data can be effectively unreachable, even for the owner. Before you rely on encryption, find your recovery key and store it somewhere other than the encrypted computer. For Windows, this may be your Microsoft account, a printout or a password manager.
5. Formatted by mistake: what you do next matters most
Accidentally formatting a drive happens more often than people admit, usually when someone selects the wrong disk. The single most important rule is to stop using the drive immediately, because every new file written can overwrite what you want back. On traditional hard drives, recovery is often possible if nothing has been written since. On SSDs the picture is less favourable: the TRIM function can make deleted data much harder or impossible to recover soon after deletion. If the data matters, switch the computer off and get advice before trying recovery software yourself.
6. Liquid damage and the "dead laptop" myth
A laptop that will not turn on after a spill is not necessarily a lost laptop. In many cases the storage is intact even when the motherboard is not, and the data can be recovered. But two habits make things worse: trying to charge or restart the machine, and delaying. Disconnect power, do not press the power button repeatedly, and bring it in quickly. Be aware that on many modern devices the storage is soldered to the board, which makes recovery more complex and sometimes impossible.
The rule that covers most of this: 3-2-1
The widely recommended 3-2-1 approach is simple:
3 copies of your data (the original plus two backups)
2 different types of storage (for example, an internal disk and an external drive or cloud)
1 copy kept off-site or offline, away from the computer
Add two habits that matter as much as the rule itself. First, keep at least one copy disconnected when not in use, so malware cannot reach it. Second, test your restore. A backup you have never restored from is only a hope.
A 30-minute backup audit for any small business
If you run a small company, set aside half an hour and answer these questions:
What would hurt most to lose? List client files, accounting data and email. Back those up first.
Where do the copies live? Confirm at least one is off-site or offline.
When did it last run? Look at the last successful backup date, not the schedule.
Can you restore? Pick one file and restore it. Time how long it takes.
Where are your recovery keys and passwords? Store them separately from the devices they protect.
Who is responsible? Name one person, and name who covers when they are away.
Switzerland adds a legal angle. The revised Federal Act on Data Protection (nFADP), in force since September 2023, requires appropriate technical and organisational measures to protect personal data. Reliable backups and tested recovery are a practical part of that obligation if you hold client data. This is general information, not legal advice.
Conclusion
Most data loss I see is not caused by exotic hacking or rare hardware faults. It comes from ordinary events, like a spill, a wrong click, a drive that quietly aged, or a key nobody saved, meeting a backup that was never really tested. The fix is cheap and boring: separate copies, one of them offline, a saved recovery key, and a restore test once in a while. Do the 30-minute audit this week, and you will probably never need the workshop for the thing that matters most.
