When this topic matters
Campaign ended. Results were below expectations. What happened? And mainly: what to do differently next time?
Post-mortem is systematic way to learn from failures (and successes).
What happens in practice
Typical "post-mortem": "It did not work because [simple cause]. Next time we will do it differently." No deeper analysis.
Problem: causes are often complex and repeat. Without systematic analysis you repeat same mistakes.
Why it fails
Blame culture: post-mortem becomes blame finding. People defend instead of being honest.
Too superficial: "Bad database" is not root cause. WHY was database bad?
No actions: analysis without concrete changes is waste of time.
How to think about it
Post-mortem structure: 1) What was the goal? 2) What actually happened? (numbers, not impressions). 3) Why? (5 whys — get to root cause). 4) What to do differently next time? (concrete actions). 5) Who will do it and when?
Rules: blameless (no blaming), data-driven (numbers, not impressions), action-oriented (concrete changes).
- Goal: what did we want to achieve?
- Reality: what actually happened (data)
- Root cause: why? (5 whys)
- Action: what specifically to do differently
What you gain and what you lose
Systematic post-mortems: you learn from mistakes, do not repeat them. But requires time and honesty.
Without post-mortems: save time. But probably repeat same mistakes.
When to apply
After every significant campaign or project — especially if results did not match expectations.
But also after successes: why did it work? What to replicate?
Post-mortem = goal + reality (data) + root cause (5 whys) + concrete actions. Blameless, data-driven, action-oriented. Do them after every significant campaign.