What Siri and dictation already do well
Start with the built-in doors, because for simple things they work. Say Hey Siri, remind me to send the deck on Thursday, and a reminder appears in the Reminders app with Thursday attached. Nothing to install, nothing to open, and it works from a locked phone or through the car. If your needs end at one clean task per sentence, use it and be glad.
Dictation is the second door. Tap the mic on the keyboard and speak into Notes, Reminders, or any text field. It transcribes faithfully. The catch is that transcription is all it does. You get your words back as a paragraph, and a paragraph is not a task.
The third door is Voice Memos, the honest workhorse. Record the ramble and nothing is lost. But nothing is caught, either. A voice memo is a promise to listen to yourself later, which makes it a second inbox, and second inboxes are where commitments go to hide.
Where voice capture on iPhone breaks down
Notice what all three doors assume: that you speak in tasks. One commitment per sentence, cleanly bounded, with the date said out loud in a form a parser expects.
That is not how thoughts arrive. They arrive tangled. The lease reply and the photos for Jess and a half-memory about the dentist, all in one breath, with the dates buried mid-sentence and the people implied. Siri is built for one task per utterance, so giving it the tangle turns your drive into a negotiation. Dictation accepts the tangle and hands it straight back, word for word.
So the real work is still waiting for you at the next desk: split the transcript, pick the dates, decide where each piece lives. Psychologists call the pull of unfinished business the Zeigarnik effect. Open loops keep tugging at attention until they close, and a transcript does not close a loop. It just moves the loop somewhere you now have to visit.
One rambling take, read once, filed as tasks
This is the gap Matters was built to close. Open the app, hold the mic, and talk the way you actually think: pauses, tangents, three things at once. While you speak, the screen shows a timer and a sound level, not a rolling transcript. It is not trying to keep up with you live. It waits.
When you stop, the whole take is read in one pass, and every name, date, and promise in it becomes its own task. The lease reply lands on the fifteenth, tied to the landlord. The photos become a task with Jess on it. Before Thursday's board thing becomes Wednesday. An errand gets marked as an evening thing, so its reminder waits for the evening instead of interrupting your standup.
And the recording itself is gone. Your audio is transcribed, read once, and discarded, never stored. What survives is the useful part: structured tasks, on every device you sign into.
Going hands-free: Siri, the Action Button, the Lock Screen
The mic works fine from inside the app, but the whole point of voice is not having to navigate anywhere. Matters registers its Siri phrase the moment it is installed. Say Hey Siri, catch a thought in Matters, or just Add to Matters, and the app opens straight into listening. Talk, stop, pocket the phone. By the next red light the tangle is tasks.
If your iPhone has an Action Button, point it at the same door: assign it to the Catch a thought shortcut in Settings. One press and the phone is listening, which matters in the places where saying a Siri phrase out loud feels silly, like a quiet train carriage. The same shortcut can sit on the Lock Screen, one tap from a dark phone to an open ear.
There is a quieter trick for text. The Shortcuts app is allowed to read your clipboard without prompting, and Matters accepts text from a shortcut. Build a one-step shortcut that hands the clipboard to Matters, and a copy plus one press files the commitment without a single screen appearing. No paste alert, no app switch, no tap. It is the closest thing an iPhone has to capture with no interface at all.
When you cannot speak, the other doors still work
Some rooms do not allow thinking out loud. For those, the same catching works in other shapes. Share anything to Matters from any app: the email that ends with can you handle this by Friday, the screenshot, the listing, the school flyer photographed on the corkboard. Tap share, tap Matters, keep moving.
Mid-conversation, the Matters keyboard has one extra key. Someone types can you send the invoice tonight, you reply sure, and you tap the key. The message still sends, and the promise is now on your list too. And the plain box in the app takes typing as messy as your speech: three thoughts, no dates picked, no lists chosen. The reading is the same behind every door.
Is my voice recording stored anywhere?
No. The audio is transcribed, read once into tasks, and discarded. It is never stored, so there is no archive of takes to scroll back through, and no recording of your voice in the car exists anywhere. The same rule covers everything Matters reads: the task is kept, the raw material is not. The full promise is written down in the Trust Center, where you can hold it against us.
Do I have to speak in complete tasks?
No, and you should not try. Polishing a thought before saying it is just the filing moved into your mouth. Say the tangle. If the filing ever lands wrong, open the task and change it: a different date, another project, someone else's name. It is a strong first draft, not a verdict. Most of the time you will not touch it, which is how you know it is working.
What does voice capture cost?
Voice capture is part of Matters Pro, and Pro is free for 14 days, no card needed. The test worth running costs nothing: on tomorrow's drive, say the messy thing out loud instead of promising to remember it. Park, open your list, and see what came back. Matters is on iPhone, Android, Mac, and Chrome.