Impulse Launch Video
Target runtime: 8:00–8:45.
Factual corrections baked in (don't undo these): - The servo locks the anchor's own strap so the anchor can't be moved — it does NOT lock the watch to your wrist. Two of your variations claimed the watch-strap version; a reviewer would disprove that in one video. - The anchor battery line is now honest-but-punchy ("battery-powered version of the same noise") instead of implying it runs forever. Put the actual runtime (~1 hr now, 2–3 hr target) on the preorder page, not in narration. - Everything else in the script is claims the firmware actually supports as of spec v0.9, including the clock-change protection and the elapsed-time timers.
1. COLD OPEN — no narration (0:00–0:50)¶
Black. White text: 6:00 AM.
A dark bedroom. The anchor on the nightstand starts sounding. You, asleep. You reach toward the nightstand — and there's nothing to press. You fumble, find the watch, put it on. The room goes quiet.
Text: 6:05 AM. You're still in bed. Obviously going back to sleep.
The watch starts. Then the anchor joins it. It gets worse. Keep rolling until you are physically through the bedroom door — one unbroken take if you can get it.
Hard cut: kitchen, lights on, you holding coffee, looking awake and mildly annoyed.
so that's an alarm you can't snooze. i'll show you how it works in a minute, but first i want to tell you why i built it.
Production note: this open is a gamble — 35+ seconds with no words. It only works if the footage is real and slightly unflattering. Film it on an actual morning, not a staged afternoon with the curtains closed. The fumbling is the content.
2. THE PROBLEM (0:50–1:45)¶
the human brain was evolved for a world that no longer exists. a world where distractions were few, and where our psychology worked for us.
now it works against us. modern life is full of stresses ancient humans never had to deal with, and our wiring hasn't caught up. tech companies have spent billions of dollars making our phones as addictive as possible, and they're winning, because they're up against hardware that was hasnt changed for about forty thousand years.
(B-roll: your real screen time report. Hold on the number.)
[YOUR ACTUAL NUMBER: "last year i was averaging ___ hours a day on my phone. i did not want to be." — use your real screen time, not the national average. your number, said plainly, beats any statistic.]and it's not just phones. there are so many goals we wish we had the commitment to achieve, and we don't — not for any real reason. we're just too worn out from the necessities to get to the nice-to-haves.
this has been a problem for hundreds of years. you'd think someone would have invented a solution by now.
(Beat. Hold on that line. It's the pivot of the whole video.)
3. ORIGIN (1:45–2:35)¶
at least, that's what i thought ten months ago, when all i wanted was two things: an alarm clock i couldn't just snooze and go back to bed, and a way to keep me off my phone that wasn't a screen time blocker i could disable in four taps.
so i went looking. the alarm clocks that exist either roll away from you, which i can catch, or they're an app, which i can close. and every phone blocker has the same fatal flaw: the thing doing the blocking lives on the phone it's supposed to be policing. so the second you actually want to scroll, you win. you always win. that's not a bug in those apps. that's just what they are.
[WRITE THIS YOURSELF, 2–3 sentences: the specific moment you decided to build it. Which blocker you deleted, what time it was, what you were doing instead of the thing. The more embarrassing and specific, the better this lands. Do not write "i realized there had to be a better way."]so i built it myself.
(B-roll: the ugly first prototype. Breadboard, tape, whatever the earliest version was. Show the bad version — this shot buys you more credibility than anything you can say.)
4. THE SYSTEM, FAST (2:35–3:20)¶
what i ended up with is two pieces of hardware. this is the watch, and this is an anchor.
anchors go in the places that matter — nightstand, desk, front door. they plug into the wall and quietly broadcast who they are, so the watch always knows which room it's in. the watch is the brain. it holds your schedule, it knows how far it is from every anchor in the house, and it decides whether you're actually doing the thing you said you'd do.
there's an app too, but the app is only for setup. once the day starts, your phone has nothing to do with any of this. that turns out to matter a lot, and i'll come back to it.
5. BACK TO THE MORNING — sunrise lock, explained (3:20–4:10)¶
so. this morning.
the watch charges on the nightstand overnight, right next to its anchor. at 6, the anchor sounds. the alarm isn't a device on my nightstand — it's the room. there's nothing to hit.
putting the watch on quiets it and starts a grace period i set myself. mine's five minutes, which is enough to sit up and become a person. when it's up, if i'm still in the bedroom, the watch and the anchor go together, and they escalate until i'm out of the room. the moment i'm out: silence.
and every one of those numbers is something i chose the night before, when i was calm and fully in agreement that getting up at 6 was a great idea. past me runs the morning. present me just complies.
6. 6:30 AM — THE DOCK (4:10–5:10)¶
(Text card: 6:30 AM. You dock the phone on the kitchen anchor.)
from 6:30am to 9am , my phone lives here.
this is the other thing i originally wanted, and it works completely differently from every screen time app, because it doesn't block anything. it measures distance. the anchor knows the phone is sitting on it, and the watch knows how far i am from that anchor. there's no app on my phone to delete, because the enforcement was never on my phone. the phone is just... over there.
(Demo: walk over, pick it up, stand there scrolling. Put a timer on screen. At 60 seconds, it goes off.)
there's a tolerance, so a quick check is fine — glance at a text, put it back, nothing happens. stand there scrolling for a minute, and the watch and the anchor both start, and they won't stop until the phone goes back down.
honestly, this is the part i'm proudest of.
7. "OKAY, BUT—" — every way out, in one block (5:10–6:40)¶
(This is Variation 3, compressed. Deliver it fast — you're anticipating the comments, not lecturing.)
now, i know what you're thinking, because i thought all of it too. so let's just go through it.
"just take the watch off." the watch knows when it's on my wrist. taking it off during a commitment doesn't end the commitment — it hands it to the anchors, which start sounding in whichever rooms i told them to. so taking it off doesn't make it quiet. it makes it louder, in a room i'm not standing in.
"just unplug the anchor." there's a battery inside. unplugging it gets you a battery-powered version of the same noise. and during a commitment, the anchor physically locks itself to whatever it's strapped to, so it can't be quietly carried out of the room or shoved in a drawer.
"just change the schedule." the watch owns the schedule, not the app. the watch won't accept any schedule changes that are obviously you trying to get out of something you shouldn't. and you can make any commitment harder instantly, whenever you want — but making one easier takes about a day to go through. the version of me that wants out at 6am doesn't get a vote. the version of me from yesterday afternoon already voted, and he was in a much better mood.
"just change the clock." tried it. the watch refuses time changes that would end a window you're currently in, and everything inside counts elapsed time, not clock time. winding the clock forward ages nothing.
and if something real comes up — because life happens — you get emergency passes, on an allowance you set for yourself. mine's
[YOUR NUMBER]a week.i'm not going to tell you this thing is impossible to cheat, because it isn't. this is probably one of the easiest ways to get around the system slow-mo shot of destroying an anchor with a hammer you can also just put a pillow over an anchor. although from my testing if you just move the anchor to a harder-to-block place, this solves it. but nobody wakes up at 6am and executes a plan. what happens at 6am is you reach out and press a button, everything i just showed you exists to make sure there is no one-second version of giving up. and it turns out, when the easy way out is gone, i just... do the thing.
(That last line is the most important beat in the video. Slow down. Mean it.)
8. 7:30 AM — THE WIFI TURN (6:40–7:20)¶
(Text card: 7:30 AM. Front door.)
here's where it stopped being the thing i set out to build. at this point i'd achieved my original goals. but then i realized the radio in my watch doesn't just see bluetooth — it sees wifi too.
which means i can use wifi networks the same way i use anchors. i can commit myself to places outside my own house. anywhere with a network. which is everywhere.
i've been using this to get to the gym. the watch checks that i'm out the door by 7:30 — if it can still see my home wifi, it knows i haven't left, and it gets progressively harder to ignore until i do. i've had a consistent gym schedule for
[HOW LONG, honestly]now. first time in my life.
(Gym b-roll. Ideally the real 7:30 shot — door, cold, whatever it actually looks like.)
9. THE REST OF THE DAY, FAST (7:20–7:50)¶
(Text cards. Move quickly — these are proof of range, not new arguments.)
a few other things i've got it doing. afternoons, there's a window where i have to be at my desk — and my phone can't be, because those two rules stack. being at my desk with my phone in my hand was ninety percent of what "studying" meant to me last year.
and at night, besides weekends, my phone docks at 11 and i have to be in my bedroom by 11:30. i built this thing for the mornings, and the bedtime one has been surprisingly useful.
[OPTIONAL, 1–2 sentences: one true detail about what the bedtime change actually did for you. sleep, mood, mornings being easier — whatever's real. skip it if nothing's true yet.]
10. CLOSE (7:50–8:40)¶
everything i showed you is one idea: half of doing something consistently is just showing up to the right place at the right time. this makes sure you show up. i showed you five uses. there's no limit here.
if you're this far into the video, i'm going to assume you think this is at least moderately cool. so: it ships in october of 2027, and the kickstarter goes live on the first of march. (ship date corrected 2026-08-21 — the old line said "january to march 2027, maybe sooner", which disagreed with every other asset and is now wrong by three quarters.) ⚠️ the "available for pre-order at 40% off" half of this line is still wrong and still unresolved — pre-orders were removed from the site on 2026-08-09 and nothing replaced them. see
Todos.md; the replacement is email capture, a pre-launch follow, or a $15 reservation, and it depends on aGATE-DEPOSITSdecision that hasn't been made. and if you'd rather wait for launch, there's a link in the description to drop your email, and i'll tell you when it's fully out.
[CLOSE IT YOURSELF, 1–2 sentences, spoken to camera. Not a slogan — the thing you'd say to a friend who's on the fence. If you can't think of anything, end on: "stop negotiating with yourself." but yours will be better.]
---¶
What I changed and why¶
Structure. This is your hybrid: Var 2's cold open and one-day spine carry retention, and Var 3's objection-handling is compressed to ~90 seconds where it does the most work — right after the dock demo, which is when a skeptical viewer's "yeah but" is loudest. The one-day spine also quietly fixes Var 1's biggest weakness (a hardware tour before anything happens).
The strap-lock fix. Var 1 and Var 3 said the anchor can "lock the watch's strap so it can't be slipped off." Per the firmware spec, the servo locks the anchor's own strap so the anchor can't be moved or muffled mid-commitment. I rewrote that beat ("can't be quietly carried out of the room or shoved in a drawer") — it's still a strong answer to "just unplug it," and it's true.
The battery line. "A battery inside, so unplugging one buys you nothing" implied indefinite runtime; it's about an hour right now. "Unplugging it gets you a battery-powered version of the same noise" keeps the punch and survives a teardown video. Put the runtime number on the preorder page.
Cadence pass. Your own style guide warns about the triple-beat AI rhythm, and the variations had it in a few places — the escalation description appeared nearly verbatim four times ("...and they escalate, and they do not stop until..."), and "that's the entire product" showed up twice with different phrasing. I kept each of those moves exactly once, at its strongest location, and flattened the other instances into plainer sentences. Same with "past me runs the morning" — once, in the sunrise section, nowhere else.
Placeholders. Five spots are marked for you: your real screen-time number, the origin moment, your emergency-pass allowance, the honest gym duration, and the final line to camera. These are the beats where authenticity is the entire value, and where anything I wrote would be a generic version of your life. The rule of thumb for filling them: if it's slightly embarrassing and very specific, it's right.
Things I kept on purpose. "That's not a bug in those apps, that's just what they are." "You always win." "There's just nothing to press." "He was in a much better mood." The lowercase, the ellipses, the mid-sentence honesty ("which sounds annoying, and it is") — that's your voice and it's the best thing in the doc.
Two flags for the shoot: 1. If your prototype is the motor-enabled board revision, you can show wrist vibration. Just make sure whatever escalation you show on camera matches what production units will do — if there's any doubt, let the buzzer and the anchor carry the demo. 2. The phone-far-from-watch variant (as opposed to docked-at-anchor) is the noisy one — don't demo it live. Everything in this script uses the docked setup, which is verified on hardware. Keep it that way.