Shut Up?!: Talk Balance

Shut Up?! counts your share of a conversation while it happens. Your iPhone tells your voice from everyone else's and shows how the talking is split over the last minute, the last five, the last thirty, and the whole session. When your share drifts too far, your Apple Watch taps you. Nobody else at the table notices.

It works for meetings, sales calls, interviews, dinners, dates, family debates, and every story that somehow became a keynote. Solo mode measures a practice talk with nobody else in the room.

WHAT IT COUNTS

Talk time, yours against everyone else's, as a live tug of war
Turns taken, interruptions, and your longest monologue
Speaking pace, pauses, and filler words
Reading scores on your own words (Gunning Fog, Flesch-Kincaid, LIX), so you know when you have drifted into jargon
Tonight against your own past sessions, learned on the phone, so a loud room where everyone talked reads differently from a room with a lot of you in it
A reading for every number: which way is high, where the line worth acting on sits, and one worked example

WHEN YOU RUN ON

A tap on the wrist first. If the streak continues, a Live Activity, then a notification, then one of six dryly funny alarm voices, from The Butler to The Terminal. Snooze, mute, pause, and one-tap stop are always there.

When the session ends you get a receipt. Share it as an image, or print it on a receipt printer.

PRIVACY IS THE PRODUCT

No audio file is ever created. None.
No transcript is ever saved. A word turns into a few numbers inside the speech callback and is gone in the same line of code.
Speech recognition and voice matching run on your iPhone. If the on-device model is missing, the session refuses to start rather than send anything out.
Other people are one anonymous group. Never named, never read, never profiled.
Sync is off by default. Turn it on and it carries numeric totals and dates. Nothing else.

Your words stay in the room. Only the numbers survive.

WHERE THE MATH COMES FROM

I built Shut Up?! as a Wharton MBA student, mostly because I kept hearing myself talk. The personal baseline comes from a marketing course, and uses the model that learns how often one customer buys and flags the weeks that sit far from it. The reading scores follow the textual analysis material Professor Brian Bushee teaches in Financial Disclosure Analytics. Any mistakes in them are mine, and neither he nor Wharton had any hand in the app.

THE FINE PRINT

Shut Up?! jokes about your numbers, never your character. It is a mirror, not a recorder, not a surveillance tool, and not a substitute for professional advice. Everyone in the conversation should know a session is running. That is what opt-in means.

COMMON QUESTIONS

Does it record me? No. No audio file is ever created and no transcript is kept.
Does it need the internet? Not to count. Speech recognition and voice matching run on your iPhone. The one download is Apple's speech model, offered before your first session.
Does it work without an Apple Watch? Yes. The Watch adds the wrist tap. The iPhone shows everything else.
Does it name the other people? No. Everyone who is not you is one anonymous group.
Can I practise a speech alone? Yes. Solo mode measures pace, pauses, filler words and reading level with nobody else in the room.
Does it work in other languages? Counting and timing do. The reading scores and some filler words are written for English and are held back in other languages.
Does it start by itself? Never. You start every session by hand, and everyone at the table should know.

Start a session tonight. Find out who the keynote speaker really was.