Voice notes and ADHD: catching the idea before it disappears

Most productivity advice is written for a brain that can hold a thought while it looks for somewhere to put it. If yours doesn't reliably do that, the advice doesn't just fail — it fails in a way that feels like a personal defect, which is worse.

This article is about the mechanics, not about willpower. If you have ADHD — or simply a brain that loses the thread between having the idea and writing it down — the problem is almost never that you don't care enough. It's that the standard capture system asks for six things in a row, and any one of them is enough to lose the whole chain.

This is a productivity piece, not medical advice. Tools help with friction; they don't replace diagnosis, treatment or professional support, and nothing here is meant to.

The gap where things vanish

There's a very short window between "I've thought of something" and "it's safely stored". In that window the idea is being held by working memory, which is exactly the resource that's in short supply.

Now count what a normal task app asks you to do inside that window:

Take out the phone
Unlock it
Find the app
Wait for it to load
Tap "new task"
Type it
Pick a project
Pick a date and a priority

Eight steps. And here's the part the advice never accounts for: the cost isn't the 40 seconds, it's what appears on the screen along the way. You unlock the phone and there are three notifications. One of them is interesting. Ten minutes later you close Instagram with a vague feeling that you'd opened the phone to do something.

The idea didn't get lost because you're disorganized. It got lost because the path to storing it goes through the most distracting object you own.

Why "just write it down" isn't the same instruction for everyone

Two extra costs stack up on top of the friction, and both hit hardest exactly where the standard system demands most.

Deciding is expensive, and capture apps demand decisions. Project, tag, priority, due date. Every one of those is a small choice, and a form with four choices at the worst possible moment doesn't get filled in — it gets abandoned. You close the app with the task half-written, which is the same as not having written it.

Perfectionism about the system eats the system. The elaborate Notion setup, colour-coded, with rollups and four linked databases. It's genuinely fun to build. It's also a project in its own right, and the day you're too tired to maintain it properly, you don't maintain it badly — you stop opening it. A system you can only use at 100% is a system that spends most of its life at 0%.

What voice actually changes

Voice removes the two expensive parts at once.

It collapses eight steps into one. Open a chat you already have pinned, hold the button, talk. You never really look at the screen, so the notifications never get their chance. The idea goes from your head to storage without passing through anything that can hijack it.

It removes the decisions. You say it however it comes out — messy, out of order, with three things mixed together — and the sorting happens afterwards, done by software instead of by you. This is the real shift, and it's more recent than people realise: telling a machine "note down that I need to talk to Marta about the contract, that's urgent, and check the domain thing, that's whenever" and getting back two clean, prioritized tasks is something that simply didn't work five years ago. If you're curious about what happens in between, we broke it down in how AI extracts tasks from a transcript.

There's also a smaller thing that matters more than it sounds: talking is faster than your own hesitation. Typing gives you time to reconsider, rephrase, decide the thought wasn't worth noting. Voice gets it out before that particular committee convenes.

Designing a system that survives you

The goal isn't a beautiful system. It's one that still works on a bad week.

One inbox, and it's the boring one

Not the app with the nicest onboarding. The one that's already on your home screen, already open, already synced everywhere. If capture lives in three places, you don't have three backups — you have three incomplete systems and no reason to trust any of them.

Zero categories at capture time

Don't choose a project. Don't set a priority. Don't decide whether it's important. Capture is dumping, not sorting. Anything you decide while capturing is a chance to stall.

If organizing has to happen — and it does — let it happen automatically, or let it wait for the review.

A review that's short and always at the same time

Five minutes, once a day, ideally attached to something you already do (coffee, the end of the workday). Not a "when I have time" weekly review — that's the appointment that never gets kept.

And crucially: the review is for skimming and deleting, not for reorganizing. If you find yourself restructuring the database, you've drifted into building the system again instead of using it.

The restart has to be free

This is the design rule everyone skips. You will abandon this for three days. Maybe two weeks. The question isn't whether — it's what you find when you come back.

A system with streaks, overdue counters and a red badge showing 47 punishes you for returning, so you don't return. A system that's just a list, where old entries can be deleted in one sweep with no ceremony, lets you pick it up on a Tuesday as if nothing happened. Design for the comeback, not for the streak.

Let the list be imperfect

Some of what you capture will be nonsense, half-thoughts, or things you'll never do. Good. That's the price of a capture system with no filter, and it's a low price. The alternative — filtering as you go — is what made you lose the good ones too.

What this doesn't fix

Being honest here matters more than usual, because this area is full of tools sold as solutions to something they don't touch.

Capturing isn't doing. A frictionless system can produce a list of 200 items you feel worse about, not better. Capture solves losing things. It doesn't solve starting them, and it's not supposed to.

The review still requires showing up. It's the one non-negotiable moment of screen and attention. Making it short and fixed is the whole strategy for protecting it.

Time blindness is untouched. Knowing the task exists is not the same as knowing it takes three hours and Friday is tomorrow. That's a different problem and needs different tools — calendars, alarms, someone expecting the thing.

A tool is not treatment. Reducing friction genuinely helps, and it's not a substitute for professional support. Anyone telling you an app fixes ADHD is selling something.

The realistic version of better

The change isn't that you become organized. It's smaller and more useful: the ideas you have stop depending on whether you happened to be near a keyboard, and the background hum of I know there was something gets quieter, because the things you said out loud are actually somewhere.

That's a real improvement, and it comes from removing steps — not from trying harder at the ones that were never the problem.

If you want to test it on the cheapest possible terms, try MicTask: one voice note, said however it comes out, and see it arrive in Notion already split into tasks. And if the part that trips you up is the destination rather than the capture, start with building a task manager in Notion — keeping it simpler than you think you need is most of the trick.