There's a moment that happens to everyone who gets into productivity systems. You discover capture — the idea that you should write things down instead of holding them in your head — and it feels like a revelation. Everything goes in. Ideas, tasks, articles, half-formed thoughts, receipts, meeting notes. Into the inbox. Get it out of your head.

Then the inbox becomes unmanageable. And you either abandon the system or start spending hours processing it. And you tell yourself the problem is discipline — if you were more rigorous, the system would work.

The problem isn't discipline. It's that you were never clear on what capture was for.

The Three-Bucket Test

Before anything goes into your capture layer, run it through three questions:

Is this pointing at a future action? Not "interesting," not "might be useful someday," but: will I act on this? Will I do something because of this? If no, it's not a task or a project — it's a reference. Put it somewhere you'll find it when you're looking for it, not in your action inbox.

Is this pointing at a decision? Not a vague "think about this more later," but: there's a specific decision I need to make, and this is part of the input. Capture it with the question it answers, not just the information itself.

Is this ephemeral? A passing thought, a random idea, something you'll think about naturally when the time comes. Let it pass. Your brain holds ephemera just fine. The capture layer is for commitments, not curiosity.

Most things fail at least two of these tests. They get captured anyway, because capturing feels productive. It isn't.

Designing a Capture Layer You Trust

A capture system you'll actually use has two properties: it's fast at the point of capture, and it leads somewhere.

Fast at the point of capture means one gesture. A keyboard shortcut. A voice memo. A note app with a widget. If capturing requires more than three seconds of friction, you won't do it when you need to.

Leads somewhere means every capture ends with a destination: project, next action, reference, or trash. Not "inbox." The inbox is a processing queue, not a home. If your system allows things to live in the inbox permanently, it isn't a system — it's a pile.

The practical design: capture to inbox, process the inbox daily, and make a decision on every item before it turns one week old. If an item sits in the inbox longer than that, either it's not important enough to capture, or you're not processing your inbox often enough.

The Only Rule That Matters

Capture is not collecting. It's committing.

If you capture it, you owe yourself a future action. If you don't owe yourself a future action, don't capture it. Let it go. Your brain is fine holding ephemeral thoughts. The capture layer is for things that matter enough to act on, decide on, or refer back to.

Capture less. Trust more. Process what you capture. That's the whole thing.