When One Drive Is a Prayer, Not a Plan

My twins were seven years old when they wandered into my office, found an open Lightroom catalog, and deleted a folder of client proofs. Four hundred selects from a corporate headshot session, gone. I had them restored in 90 seconds from a local backup. My kids barely registered it as an incident. I registered it as proof that the system worked.

That 90-second recovery didn’t happen by accident. It happened because I’ve built redundancy into every layer of my workflow, from the moment a card comes out of the camera to the final delivery of a client gallery. Most photographers treat backup as something they’ll get around to. I treat it as the actual job.

If you shoot professionally and you’re relying on a single hard drive, you’re not running a business. You’re gambling with your clients’ memories and your own reputation.

What “Corruption” Actually Means (And Why It Happens Without Warning)

Hard drives fail. Flash cards corrupt. RAID arrays, which people often assume are backup solutions, are not backups. They’re redundancy at the hardware level, which protects against a single drive dying, but not against accidental deletion, ransomware, or a power surge that takes out the whole enclosure.

The failure mode that worries me most is silent corruption. A drive can read and write fine for months while quietly degrading at the sector level. You won’t know until you try to open a file and it’s gone. This is exactly what happens with cheap, heavily written SD cards used past their rated write cycles. The card reports no errors. The data is simply not there anymore.

Solid-state drives are faster and more durable under physical stress, but they’re not immune to firmware failures or sudden complete data loss with no warning symptoms. Spinning hard drives fail more predictably, but “predictably” still means eventually and without a confirmed date on the calendar.

The answer is not better hardware. The answer is redundant copies in multiple physical locations.

The 3-2-1 Rule, Applied to an Actual Shooting Day

The 3-2-1 rule is old and it’s still correct: three copies of your data, on two different media types, with one copy offsite. Here’s how that maps to a real commercial shoot for me.

On location, I shoot dual-card when the camera supports it. For my Sony A7R V, I configure CFexpress Type A in slot one as primary and SD in slot two as an in-camera backup. The moment I’m back at my desk, I copy the primary card to two separate local drives before I format anything. I use a dedicated Thunderbolt 4 dock with two 4TB Samsung T7 Shield drives. The copy takes about eight minutes for a 1,200-raw-file portrait session at 60MB per file.

Before I sleep that night, those files are also syncing to Backblaze B2 cloud storage. My current plan runs about $7 per month for the storage I use, which hovers around 8TB. That’s a rounding error compared to a single client invoice.

For my website and business files, I use a separate workflow. My site runs on WordPress, and I back it up daily using UpdraftPlus Pro ($70 per year) to a dedicated Google Drive folder. Database and files, separate archives, retained for 30 days. I test a full restore every three months by spinning up a local staging environment using Local by Flywheel.

The Offsite Copy People Always Skip

Every photographer I know has local backup covered, at least partially. Almost none of them have a true offsite copy that updates regularly. “I back up to an external drive” does not mean you’re protected if your office burns down or gets robbed.

Cloud storage is the practical answer for most working photographers. Backblaze Personal Backup is $9 per month for unlimited storage from a single computer. If your primary editing machine is where all your RAW files live, that single subscription covers continuous versioned backup with 30-day file history. For studio-scale operations with a NAS, Backblaze B2 gives you S3-compatible object storage at $6 per TB per month.

A physical offsite option that actually works: a rotating set of drives kept at a second location. I use three 8TB WD MyBook drives. One stays at my home office, one goes to a rented safe deposit box at my bank, and one is in transit between them on a two-week rotation. This sounds elaborate. It takes about 20 minutes every two weeks.

Monthly Testing Is the Part That Makes Everything Else Real

I test my backup drives every month. Not just plug them in and confirm they mount. I actually open files. I check file counts against my Lightroom catalog. I verify the most recent offsite cloud upload timestamp. I run a quick S.M.A.R.T. check on my spinning drives using DriveDx on macOS.

This monthly ritual catches things. Last February, my offsite sync had silently stopped running after a macOS update changed a permission setting. The local copies were fine. The cloud backup had a six-week gap. Without the monthly check, I would never have known until I needed that copy.

No backup system runs itself indefinitely without supervision. Drives change firmware. Software updates break automations. Cloud services change API structures. Your job is to verify, not just set and forget.

The single most important thing you can do today is not buy another drive. It’s to open your most recent backup right now, verify a file actually opens, and confirm the timestamp is what you expect. If you can’t do that in two minutes, you don’t have a backup. You have the idea of one.