Launch week, hour by hour
Launches don't fail on day one. They fail on day three, when the founder stops posting because day two was quiet — which is exactly when the second wave of attention would have arrived.
Day 0 — the day before
- Submit to the pile-A directories from the submission sprint. They take days to appear, so they need a head start.
- Line up three people who've said they'd take a look. Not for votes — for the first three comments, so the post isn't empty when strangers arrive.
- Write your replies to the two objections you already know are coming. You'll be answering them all week.
- Turn off anything that emails you. You want to be looking at one screen tomorrow, not eleven.
Day 1 — post early, then stay in the room
1Morning: post to your single biggest channel
One channel, not five. Whichever community you've actually been part of. Everything else this week radiates out from how this one goes.
2First three hours: reply to everything within minutes
This is the window that decides distribution nearly everywhere. A post with six real replies in the first hour outperforms a better post with none. Clear your calendar for it.
3Afternoon: personal messages, not broadcasts
Message the twenty people who'd genuinely want this, individually, with a sentence about why them. Twenty DMs beats a thousand impressions.
4Evening: write down what people actually asked
The questions you got today are your homepage copy, your FAQ, and the next three posts. This is the most valuable thing day one produces and almost nobody records it.
Day 2 — the quiet day
Day two is always quieter than day one and it means nothing. Do not read it as a verdict. Spend it on the second tier of channels and on the thing you learned yesterday.
- Post to your second and third communities, rewritten for each — same story, different opening.
- Fix the one thing three or more people got confused by. Then say publicly that you fixed it; that post outperforms the launch post surprisingly often.
- Finish the pile-B directory submissions.
Day 3 — where most launches stop
- Send your newsletter pitches. Editors work on a lead time — pitched on day three, you appear in two weeks, exactly when the launch spike has died.
- Post the first 'what happened' update somewhere. Real numbers, including the disappointing ones.
- Start commenting — not posting — in the two communities you want to post in next month.
Days 4 and 5 — convert attention into something durable
- 1Email everyone who signed up, personally, asking one question: what were you hoping this would do? The answers are worth more than the signups.
- 2Publish the launch retrospective. What you expected, what happened, what you'd do differently. This is the post that gets shared by people who didn't care about the launch itself.
- 3Check your sources. Whatever you're using to measure traffic, look at it properly now — you have five days of data and it will tell you which two channels deserve the next month.
- 4Pick next week's one thing. Not five. One channel, one post, one experiment.
The week after
The compounding channels — newsletters, creators, podcasts, SEO — all pay out between two weeks and three months after you pitch them. Launch week is when you plant those, not when you harvest them. If week two feels like nothing is happening, that's because week two is the gap between the two, and it's the most normal thing in this entire process.
Common questions
What if day one goes badly?
A launch post that gets no traction is information about the post, not the product. Rewrite the opening line, wait a week, post to a different community. Nobody remembers a quiet launch except you.
Should I launch on a Tuesday?
It matters far less than being available to reply for three hours after you post. Pick the day you can genuinely clear.
Can I launch twice?
Yes, and most successful products do. A major feature, a rewrite, a new audience — each is a legitimate reason to run this whole week again.