The Twelve-Minute Post-Mortem: Give a Failure Four Lines and a Check Date

When something goes wrong — a post flops, a generation comes out smeared, a launch underperforms — most builders do one of two things.

They brood about it for a week, replaying it in the background while they do everything else badly. Or they shrug and move on, which feels productive and guarantees they pay for the same mistake again next month.

The post-mortem is the third option. Pen, timer, one thing that broke. Twelve minutes, and when the timer rings, you are done thinking about it until the check date.

The four lines

Line one: what happened, in one sentence, with a number if you have one.

Not "the post did badly." Instead: "The Tuesday post got 40 views and zero clicks." Not "the song was bad." Instead: "The regeneration kept the bubblegum delivery after a genre change to Death Metal." One sentence, one fact.

Line two: one cause you control.

Not the algorithm, not the weather, not the customer. One cause inside your own decisions. "I posted at midnight with no hook in the first three seconds." "I changed the genre but left the old delivery tags on the lyrics." If the cause genuinely isn't yours, write that too — then the fix is a different kind of fix, like a refund, an apology, or a boundary.

Line three: one change you'll test.

One, not five. A change is something you do, not something you hope. "Try harder" is not a change. "Next post leads with the chorus in the first three seconds" is. It needs a verb and a moment: I will do this, at this step, next time.

Line four: a check date.

Pick a day one to two weeks out and write it down. "Check November 17: did the next three posts lead with the chorus, and did clicks move?" Without a check date, a post-mortem is a diary entry. With one, it's an experiment.

A worked example

Line one: "The song post got 60 views across three platforms and zero link clicks in 48 hours."

Line two: "Every post was the song title plus a link. No lyric, no story, no reason to stop scrolling."

Line three: "Next release, the first post is one lyric line on screen over the chorus, and the link goes in the second post, not the first."

Line four: "Check in two weeks, after the next release: did the lyric-first post beat 60 views and get at least one click?"

Notice what happened. The failure stopped being a verdict on the music and became a verdict on the post. The music was never tested — the packaging was. Most failures are not about the thing you made. They're about the step around the thing you made.

Three rules to keep it honest

1. One failure per post-mortem. Autopsy the whole month and you'll write an essay and learn nothing.

2. The cause has to be an action, not a trait. "I'm bad at marketing" is a mood. "I posted without a hook" is a cause.

3. Never post-mortem someone else's failure. You can't fix their cause. That's just gossip with a timer.

Two more guardrails: don't do it while the wound is still open or when you're exhausted — you'll write the defensive sentence, not the honest one. Give it a night. And don't post-mortem the same failure twice. If you wrote it, tested the change, and checked the date, the case is closed either way.

Keep them in one place

One notebook, one document, one list in date order — four lines each. After a month you'll have four or five entries, and reading them back takes three minutes. That's where the real payoff is: your repeat. Everybody has one. Posting without a hook. Skipping the lyric check. Promising a date before the work is ready. You won't see it until the entries sit next to each other.

One post-mortem a week is 52 a year. If even a quarter produce a change that sticks, that's 13 fixes that would otherwise have stayed as background brooding.

Do it on the worst thing this week, not the easiest thing. The worst thing is the one still charging you rent.