Notes

Why your filing system keeps failing

· 3 min read

You have built a filing system before. Possibly several. They start well, they last a few weeks, and then there is a folder called Misc with four hundred things in it and you are back to searching by memory.

The usual explanation is that you did not stick with it. That explanation is wrong, and believing it means you build the same thing again.

The real failure is a timing problem

A folder structure asks you to answer a question at the moment you save a file: where does this belong?

At that moment you are doing something else. You are in the middle of a task and the file is a side effect of it. You do not know yet whether the thing is important. You often do not know what it is, because you have not read it, because you just downloaded it.

So you put it somewhere provisional. Downloads, the desktop, Misc. Which is the correct decision given what you knew, and which is also how the system dies.

The second failure is that categories are not stable

You make a folder called Work. Then you change jobs, so now there are two kinds of work. You make a folder for a client, then the client has three projects. You make a folder for the house, then you buy a different house.

Every structure is a snapshot of how your life was arranged on the day you made it. Life moves and the structure does not, so you either keep reorganising, which nobody does, or the structure slowly stops matching and you start bypassing it.

The third failure is that you cannot search a name that says nothing

Even when filing goes well, most of what you file is called something useless. Scan_0042.pdf. IMG_6268.jpg. document (3).pdf. invoice.pdf, of which you have nineteen.

Folders were meant to compensate for bad names. Put invoice.pdf in the right folder and the folder supplies the meaning. That works until the folder has two hundred files in it, and then you are back to opening things one at a time to see what they are.

What to do instead

Three changes, and they can be made independently.

Move the decision later. Let things land anywhere, and sort in bulk when you have attention to spare. A weekly pass over one pile is far more likely to happen than a decision on every file.

Fix the names, not the folders. A file called Riot Digital invoice, March is findable from anywhere. A file called invoice.pdf is findable only from its folder. Names travel, folder context does not.

Keep the structure shallow. Five or six top level groups that match how you actually think, not thirty nested ones that match an org chart.

Where software helps and where it does not

Software cannot make you care about filing. What it can do is remove the two tasks you will not do by hand: reading every file to decide where it goes, and renaming every file to something meaningful.

That is the whole design of Stash. You make a handful of stashes and describe each in a sentence. Everything you drop in gets read and put in one of them, with a name taken from what is actually inside it. When it is not confident, it does not guess: the thing waits in a pile called Needs you with a suggestion, and you tap the right one.

The corrections matter more than they look. When you move something, that gets remembered and fed back in, so the sorting drifts towards how you file rather than how a stranger would.

The test of a filing system

Not whether it looks tidy. Whether you can find a specific thing in under a minute six months later, without remembering where you put it.

By that test, most folder structures fail and a flat pile with good names and good search passes. Which is worth thinking about before you build the tree again.

Stash does the filing

Drop anything in and it gets read, named for what is actually inside it, and put where it belongs. Ask for it back in plain words. A hundred things free.

Open Stash