A post-performance postmortem: a checklist for what actually happened

A post-performance postmortem: a checklist for what actually happened

The question: a post over- or under-performed — how do you diagnose why, instead of guessing and repeating the mistake?

Without a structured postmortem, teams attribute outcomes to whatever feels intuitive, confusing correlation (two things moving together) with causation (one causing the other).

The postmortem checklist, run within 72 hours:
— Compare reach-per-impression to your own 90-day median, not to a viral outlier.
— Isolate the variable: was format, topic, hook, or timing different from your baseline?
— Read the first 10 comments — did they engage the idea or miss the point? (Missing the point signals a weak hook.)
— Check the reach-to-impression ratio for re-surfacing; high ratio means dwell carried it.
— Note who engaged: buyers or peers? Peer-only engagement inflates vanity metrics.
— Resist single-cause stories; most outcomes are multi-factor.

Limitations: n=1 per post means you can never fully isolate cause — the algorithm, news cycle, and timing all confound. Treat every conclusion as a hypothesis to test on the next post, not a proven rule.

What it means for B2B: a disciplined postmortem turns scattered posts into a compounding learning system instead of repeated luck.

Bottom line: diagnose with your own median and resist the tidy single-cause story — that's where most teams fool themselves.