Photography is a high-churn discipline. Shoots land in bursts, archives grow linearly, and the gear that holds your work accumulates the way any tool does — some of it fails silently, some of it only betrays you on the day you need it. This checklist is the version I’d give a working photographer after an evening’s conversation, not the glossy one a vendor produces.
It is deliberately short. Twelve items, grouped by cadence. Half of them take five minutes. The other half take longer, but compound: run them for a year and you have a practice rather than a hope.
Daily
1. Two copies, from the camera. Shoot to CF Express and SD simultaneously if your body has both slots. If it doesn’t, use ShotPut Pro, Hedge, or Photo Mechanic to duplicate cards on ingest — they verify the bytes match, which is the part that actually matters. The ingest tool checksum is the first-line-of-defense that most photographers skip because it adds three minutes. It’s worth three minutes.
2. Don’t format a card until its contents are in two separate Mac-attached destinations and one off-site-bound destination. The temptation, after a long wedding, is to format the second card after the first one has “clearly” copied fine. Don’t. The rule is: two local copies plus a pending off-site copy before the card goes back in the bag. If you shot all day and your upload to off-site hasn’t finished yet, that’s fine — the pending queue counts as long as it’s running. Formatting is the one-way door; treat it like one.
Weekly
3. Verify last week’s shoot opens in Lightroom without “offline” warnings. If the external drive that holds the RAWs is offline, the catalog is pointing at something that isn’t there. This is where silent archive drift begins — a drive letter changes, a volume gets remounted under a new path, and suddenly your catalog is a beautiful map of nowhere. Open the shoot, look for the question marks on previews, fix them while the context is fresh.
4. Check SMART status on any drive you rotated that week. DriveDx is the tool most photographers land on; whichever you use, look at reallocated-sector counts and pending-sector counts. A drive that grew either value this week is a drive you retire this month, not a drive you retire after it fails. Backblaze’s published drive stats have shown for a decade that SMART signals — especially SMART 5 and SMART 197 — precede actual failures with real predictive power.
Monthly
5. Run a test restore of a random file from the last shoot. Not a synthetic “hello world” file — a real RAW you care about, from a real session, restored from your cloud destination to a scratch folder. Open it in Lightroom. If it doesn’t open, the restore is theater. This is the single most neglected step in every backup workflow I’ve audited, and it’s the one that separates a working backup from a working-looking backup.
6. Rotate the off-site drive. Whichever of your two archive drives is in the studio, take it to the box in the car, the family house, or the locker, and swap. If a fire, a flood, or a thief visits your studio tonight, the drive that’s elsewhere is the drive that saves your career. Monthly rotation is the minimum; weekly is better if the work merits it.
7. Review snapshot sizes on the cloud destination. An unusual spike — a shoot that’s suddenly five times the size of last month’s — is usually a new cache, scratch, or sidecar directory that needs an exclusion rule. Left alone, these quietly eat storage for years. Cloud backup tools that chart snapshot deltas make this a thirty-second scan.
Quarterly
8. Replace a drive that’s passed year 5. Don’t wait for it to die; the math says the annualized failure rate climbs fast after year 4, and Backblaze’s Drive Stats show the fifth-year cohort failing at three to five times the rate of the third-year cohort. We’ve written the full data story at /backup/drive-failure. The short version: mechanical drives are not built for a decade of service, and the moment they start drifting toward that age is the moment to transplant what’s on them.
9. Check the Lightroom catalog’s integrity (File > Check Catalog). If it reports issues, restore the catalog from backup before you touch anything else. A corrupt catalog that gets “fixed” in place can lose keyword history, star ratings, and develop settings silently — the kind of loss you notice a month later when a client asks for a re-export. The catalog is software and therefore more fragile than the images themselves. Treat it that way.
10. Re-verify your retention policy matches your contract. Still holding deliverables for clients who wrapped last quarter? Fine. Holding shoots from 2019 that a client would rather you forget? Less fine. Retention is a business decision and a privacy decision, not a technical default. If you haven’t thought about it since you set up your backup, you haven’t thought about it enough.
Once a year
11. Do a full end-to-end disaster drill. Pretend your Mac got stolen on a shoot day. Borrow a second machine. How fast can you be shooting again — delivering edits, opening catalogs, reconnecting archives? In real-world recoveries, the bottleneck is almost never the RAWs; it’s the Lightroom catalog rebuild and the sidecar reconciliation. A single drill per year will surface gaps you didn’t know you had, and it will turn a future bad day into a drill you’ve already run.
12. Review the archive structure. By-year folders rot. Are you still finding things quickly? Consolidate, prune, re-import catalogs if needed. An annual archive review is the moment to delete the “to-sort” folder that’s been to-sorting itself since 2023 and to promote the one naming convention that actually survived contact with your shooting year.
This isn’t about being a compliance machine. It’s about never having the conversation I’ve had with a photographer in a parking lot at 11pm about an email to a client who doesn’t have their wedding album. A month of discipline is worth a career of peace.
For the deeper treatment — directory structure, catalog-backup cadence, and the full strategy for a 4-TB-a-year shooter — read the backup strategy for photographers guide. For the why behind item 8, read the drive-failure essay.