How to organize files from old computers and backup drives

By Michael, founder of DocAttic · September 2026 · 7 min read

Most people don't have one place their documents live — they have a drawer. A couple of external drives, an old laptop still in a cupboard, a USB stick, and a big Backup of backup folder someone made years ago. Everything important is in there somewhere, several times over. Here's how to turn that pile into one archive you can actually trust, without copying the same file a dozen times.

The familiar pile

It's rarely one mess. It's layers of them, each made with good intentions at the time:

• A drawer of external drives — the disk from the computer before this one, and the one before that, each a near-complete copy of a life you've mostly forgotten the shape of.
• An old laptop — still boots, still has a full Documents folder you never fully migrated when you got the new machine.
• A "Backup of backup" folder — someone once copied everything into one place to be safe, then copied that place too. Now the safe copy is part of the problem.
• A big zip from years ago — archive-2018.zip, made before a reinstall, with hundreds of documents sealed inside it that no search on your PC can see.

The trouble isn't that anything is lost. It's that everything is present many times, in slightly different arrangements, and no two piles agree with each other.

Why merging by hand goes wrong

The instinct is to plug in a drive, open it next to your Documents folder, and start dragging. It never ends well. Do it carefully and you're comparing folders by name and date for hours, second- guessing whether Taxes-final or Taxes-final2 is the one to keep. Do it quickly and you either miss files — a folder you didn't open, a drive you didn't get to — or you duplicate them, dropping another full copy of the 2016 backup on top of the 2016 backup you already merged last month.

By hand, those are the only two outcomes: an archive with holes in it, or an archive that's bigger than all the originals put together because every document is in it five times. Neither is the tidy single copy you actually wanted.

It compounds, too. Each drive was itself a snapshot of an earlier machine, so the same 2015 invoice can sit on four different disks, in three differently named folders, under two different filenames. There's no arrangement you can hold in your head that untangles all of that reliably, which is why the by-hand approach quietly stalls halfway through and the drives go back in the drawer.

Documents hiding inside archives

The zips and tarballs are their own trap. A .zip, .tar.gz, .7z or .rar is a sealed box: to the rest of your computer it's one file, and everything inside it is invisible to search and to any drag-and-drop cleanup you do. So the documents in that old backup archive quietly get left out of every attempt to organize — or, if you do unzip it, you now have a fresh sprawl of files that overlap heavily with what's already on the drive. Either way the archive makes the duplication worse, not better.

The risk of deleting the wrong copy

The most nervous part of merging old drives by hand is the deleting. Once you believe you've gathered everything, the temptation is to wipe the sources to reclaim the space. But if your merge missed a folder, or one "duplicate" was actually a newer revision, that delete is permanent — and you often won't notice for months.

File by year, keep each document once

Once everything is gathered, the durable way to file it is the plainest: a folder per year. A 2014 contract goes under 2014, a 2019 tax return under 2019. It needs no app to make sense and no discipline to maintain, because the year a document belongs to never changes.

And the one rule that makes a multi-drive merge worth doing is this: each unique document is kept exactly once. The reliable test isn't the file's name or its folder — those lie after years of copying — it's its content. Two files that are byte-for-byte identical are the same document, no matter what they're called or which drive they came from, and you only need one. Anything that genuinely differs — a newer revision, a re-scan — is a distinct file and stays.