the short version
- Pick the date first. Leave at least a week for App Review, and set the release to manual so approval doesn’t mean ‘live at 3am’.
- Thirty days out you need a sentence, two or three channels, and a name for who is awake on the day.
- Twenty-one days out the hero moment is chosen. Fourteen days out the assets exist. Seven days out the build is submitted.
- If a gate isn’t true, move the date once, on purpose. Don’t slide it by going quiet.
In this post6 sections
Pick the date first, leave at least a week for App Review, and choose the two or three channels you will actually show up on. Then work backwards through four gates: 30 days, 21, 14 and 7. If a gate isn’t true on the day you reach it, the date moves once, out loud. A launch without a date is a mood. A date nobody checks is how “soon” happens.
This is the founder’s list: decisions and dependencies. It is not a second copy of the production calendar. What gets made in each week (teardown, polish, screenshots, page, film, UGC, the day) is the four-week launch plan. The last seven days, in more detail than a gate can hold, is before you submit. Use those to do the work. Use this to know whether the date is still real.
We haven’t run a client’s 30 days. Ignura is new. The gates are the ones we’d hold you to, and the ones you can hold yourself to.
Thirty days out
The goal of this week is a date and a sentence. No pixels yet.
Say the date to someone who will remember it. A Tuesday, Wednesday or Thursday, unless you have a reason. Check it against a bank holiday and anything you already know will eat the week. Write it where the whole team can see it. A date in one person’s head is not a date.
Leave room for review. App Store review is often quick, and “often” isn’t a plan. Submit the release build at least a week before launch day, and set it to manually release, so approval means “waiting” rather than “live while you were asleep”. If you’re on Android too, read the closed-test note in the four-week plan and start whatever wait the current Play rules require. Policies move. The point is that the wait is measured in weeks, so it belongs in this gate, not in the last one.
Write the one sentence. What the app does, and who it’s for, in words a stranger can repeat. The landing-page post is how to write it. If the team can’t agree the sentence, the screenshots and the film will each invent their own, and they will not match.
Pick two or three channels. Product Hunt, the App Store page, X, TikTok, a mailing list you actually have. Not seven. Name the owner of each one. “We’ll figure out Product Hunt” is not an owner. If Product Hunt is one of them, open a personal account now. The Product Hunt checklist is why a week-old account is the floor, not the plan.
Name the day. Who posts, who replies, who watches the server, who sleeps first. One name per job. A job named “the team” belongs to no one.
Twenty-one days out
The hero moment exists, on a screen, not as a sentence in a doc.
Use the app like a stranger. Delete it, reinstall, sign up with a new email, and write down every hesitation. The moment that makes you think “oh, nice” is the one that goes on screenshot 1, in the film, and at the top of the landing page. If you have three candidates, you haven’t chosen. Choose.
By the end of this gate:
Polish, if you need it, is two weeks from £4,000. Leaving it until the week of launch means you launch the hesitation you wrote down.
Fourteen days out
The assets exist as files, not as intentions. They can still be rough. They cannot still be theoretical.
| Asset | True means |
|---|---|
| Screenshots | Frame 1 would survive the phone test. The rest of the set is in progress, not unstarted |
| Listing words | Subtitle and the first lines of the description exist, and they agree with frame 1 |
| Landing page | A page you can send, with the sentence and one action. Pretty can wait. The link cannot |
| Film or preview | A rough cut, or a clear decision that you aren’t making one |
| UGC | Briefs out, rights line filled in, shoot date before launch week |
| Support | An inbox a human will read, and a second human if the day runs overnight |
This is the gate where founders discover they have a film script and no footage, or a page and no sentence. Both are recoverable at fourteen days. Neither is recoverable the night before, which is why that night is not the night you start.
The stop-scrolling test belongs here, on whatever cut you have. Two seconds, sound off, on a phone. If it fails, change the opening this week, while changing it is still cheap.
Seven days out
The build you mean is submitted. The listing you mean is the one in App Store Connect, not the one in a Figma file named “final-v7”.
The detailed list (manual release, privacy, the link on a phone that isn’t yours, the night-before ban on new ideas) is before you submit. Don’t keep a second copy of it here. The gate, in one breath:
- The release is in review, or approved and set to manual release
- The store link you’ll post is written down, and you’ve opened the thing it will point at
- The run sheet has a name on every hour that matters
- The Product Hunt maker comment is drafted, if you’re using Product Hunt
- Nobody is “just quickly” adding a feature
A new binary restarts the wait. A new headline, the week of launch, usually means the sentence was never agreed. If the sentence is wrong, it had to be wrong in public at the 21-day gate, when changing it was free.
If a gate fails
Move the date once. Tell the people you promised, the same day, with the new date. Don’t slide it by omission, and don’t move it four times in a whisper. One honest move is a plan. A date you quietly stopped mentioning is the thing your next investor will ask about, and you will not enjoy the answer.
When you move it, move it to a day that clears the gate you failed. “Next Thursday” is only a plan if review, or the film, or the creator’s shoot, can actually finish before it. Otherwise you’ve booked a second disappointment.
What you can cut, if the date can’t move: deep polish, a long film, a seventh channel. What you shouldn’t cut: frame 1, a page with one true sentence, and a link that works. The four-week plan has a two-week version for when the date is already close. It is a cut list, not a wish.
Questions
How far ahead should I submit to the App Store?
At least a week before the day you want to be live, with manual release turned on. Review is often faster. Often isn’t something you can put on a calendar. The before-you-submit post is the list of what you hand in with the binary.
What if I only have two weeks?
Use the two-week cut in the four-week plan: screenshots, a one-page site, a small batch of videos, a tight launch day. Shrink the film or skip it. Don’t skip the sentence, and don’t skip the link check.
Do I need all of this if I’m not doing Product Hunt?
No. Product Hunt is a channel, not a law. The gates are the date, the sentence, the store, the page and a person who is awake. Drop any channel you won’t staff. A quiet channel you meant to “also do” just makes the day longer.
When should I ask for help?
At the 30-day gate, if the sentence or the screens aren’t something you’d show a stranger, or if nobody on the team has time to be the person who’s awake. Ignite is the whole launch, four weeks from £16,000. One piece on its own starts at £1,500. Our first 3 clients get up to 50% off, for a testimonial, a public before and after, and one intro to another founder.
If your date is in the next month and you want the gates read against it, book a free 15-minute call. Bring the date. We’ll tell you which gate is the one that will slip.
