Turns launch-day results into the second launch - a metrics checklist filled during the day itself (hourly rank, referrer split, signup count), a public day-plus-2 retrospective shipped within 48-72 hours as an X post and a LinkedIn document post, permanent placement of earned badges and coverage on the site, and the conversion-ratio readout that calibrates the next launch cycle. Use when someone asks "write our launch retrospective", "we launched two days ago, what should we post now", "turn our launch numbers into a follow-up post", "how do we get a second wave of attention from the launch", or any product launch ended within the last three days and the numbers are still fresh. Do NOT use for an incident or outage postmortem - use postmortem-writer instead - or for the social post mechanics themselves - use launch-thread-anatomy instead; for adapting the writeup across more platforms, use content-repurposing.
---
name: launch-retrospective
description: Turns launch-day results into the second launch - a metrics checklist filled during the day itself (hourly rank, referrer split, signup count), a public day-plus-2 retrospective shipped within 48-72 hours as an X post and a LinkedIn document post, permanent placement of earned badges and coverage on the site, and the conversion-ratio readout that calibrates the next launch cycle. Use when someone asks "write our launch retrospective", "we launched two days ago, what should we post now", "turn our launch numbers into a follow-up post", "how do we get a second wave of attention from the launch", or any product launch ended within the last three days and the numbers are still fresh. Do NOT use for an incident or outage postmortem - use postmortem-writer instead - or for the social post mechanics themselves - use launch-thread-anatomy instead; for adapting the writeup across more platforms, use content-repurposing.
---
# launch-retrospective
The cheapest story a launch creates is the launch itself, and most teams abandon it within a day. The audience that saw the launch but did not act is far larger than the audience that acted, and the retrospective reaches them with fresh social proof - it outperforms the announcement because it is generosity-coded (teaching) rather than ask-coded (selling). Launches come in cycles: the retrospective's conversion ratios are what calibrate the next one, so a one-and-done launch leaves both the second conversion touch and the calibration on the table.
## Operating procedure
Run the steps in order. The capture checklist must exist before launch day because hourly rank data cannot be reconstructed afterward; the numbers readout precedes the narrative because the narrative is built around whichever number is screenshot-able; publication and badge placement close because they consume the finished piece.
### Step 1: Gather the launch record
If the launch has not happened yet, this skill has exactly one job today: print the launch-metrics-checklist template below and hand it to whoever owns launch day. Everything else waits for data. Otherwise, gather:
1. **Charter**: the client's brand charter - the published piece's register and any badge styling on the site inherit from it. Paste the charter block if one exists. If the client has a charter but the foundations pack is not installed, elicit the needed values directly and label them provisional. If no charter exists at all, halt and route to premium-design-foundations - never substitute worked-example values.
2. **The filled metrics checklist**: hourly rank, closing upvote count, referrer split, signup count, conversion figures, and what broke. If the team launched without one, reconstruct what is still verifiable (closing counts, analytics referrers, badge emails) and label everything else a guess.
3. **Pre-launch expectations**: waitlist size and the play verdict locked at date lock - VELOCITY-PLAY or FEEDBACK-PLAY from the 200-signup channel-strategy fork owned by waitlist-and-teaser-mechanics (the fork calibrates the launch; it never blocks a date). The retrospective judges the launch against the play it ran, not against someone else's play.
4. **Badges and coverage earned**: every badge, ranking, and press mention, each with its source URL.
5. **The honest what-went-wrong list**: at least one genuine miss, from the people who ran the day.
If any input is a guess, label it a guess and move on.