The Backup That Doesn’t Exist Until It’s Tested
The majority of people who lose important data had some form of backup. The backup was to a drive that failed before the restore was needed, was to a cloud service that didn’t back up the specific folder containing the important files, was outdated by months (backing up photos to a service that was paused when storage filled up), or was a backup that was never tested for successful restore and turned out to be corrupted. A backup strategy that isn’t complete, current, and tested isn’t a backup strategy — it’s a false sense of security.
The 3-2-1 backup rule is the standard framework that addresses the most common backup failure modes: keep 3 copies of data, on 2 different types of media, with 1 copy offsite. This redundancy ensures that no single failure — a drive failure, a ransomware attack, a fire, a theft — destroys all copies simultaneously.
The Three Copies: What They Are
The first copy is the primary data on your working computer — the files you’re actively using and editing. The second copy is a local backup on a separate physical device from the primary: an external hard drive or NAS (network-attached storage) in your home or office. The third copy is an offsite backup: cloud storage, an external drive at a family member’s home, or an offsite backup service. If your primary drive fails, the local backup provides immediate restore. If your home is burglarized or damaged, the offsite copy is unaffected.
The ‘different types of media’ requirement addresses the failure mode where both copies are on the same type of media that fails simultaneously: SSDs have a specific failure mode profile different from spinning HDDs; cloud storage has different failure characteristics than local storage. Having local and cloud copies ensures that a specific technology’s failure mode doesn’t eliminate both copies.
Software for Local Backup
macOS Time Machine (built-in, free) provides continuous local backup to an external drive with version history — the ability to restore files as they existed at any point in the backup window. For Mac users with an external drive, enabling Time Machine is the easiest complete local backup available. Windows File History (built-in, free) provides the equivalent for Windows, backing up libraries, desktop, contacts, and favorites to an external drive with version history.
Acronis Cyber Protect Home Office (formerly Acronis True Image, $50/year) provides full system image backup alongside file backup for Windows and Mac — the capability to restore an entire system including operating system and applications to a new drive, rather than just restoring files to a working system. Full system image backup is particularly valuable when the goal is rapid recovery from a complete system failure rather than just recovering specific files.
Cloud Backup vs. Cloud Sync: The Critical Distinction
Cloud sync (Dropbox, Google Drive, OneDrive, iCloud Drive) synchronizes files between devices — what’s on one device is available on others, and changes on one device propagate to all. This is useful for file access across devices but is not a backup strategy. If a file is deleted on one device, the deletion syncs to all devices and to the cloud copy. If files are encrypted by ransomware, the encrypted versions sync to the cloud, potentially overwriting the clean copies.
Cloud backup (Backblaze Personal Backup, Carbonite, CrashPlan) is specifically designed for backup: it continuously backs up all files on the computer, maintains version history for deleted and changed files, and doesn’t propagate deletions or changes in ways that destroy the backup. Backblaze Personal Backup at $99/year (unlimited storage) is consistently recommended as the best value cloud backup for its combination of unlimited storage, continuous backup, and 30-day file version history.
Testing Restores: The Step That Validates Everything Else
A backup that’s never been tested for successful restore may not actually work — backup software can run without errors while producing corrupted or incomplete backups. The annual backup restore test is the practice that transforms a backup strategy from assumed reliability to verified reliability: restore a small set of files from each backup system (local backup, cloud backup) and verify they open correctly and match their expected content.
For businesses and anyone with genuinely irreplaceable data (business records, original creative work, irreplaceable photos), the restore test should be done quarterly rather than annually. The psychological resistance to testing restores (‘I don’t want to deal with restoring right now’) is the same resistance that leaves backup failures undiscovered until a real restore is urgently needed. A planned test restore is far less stressful than an emergency restore that discovers the backup wasn’t working.