At larger companies, performance reviews are the primary time people reflect on what’s wrong.
At startups, there’s no time to keep making the same mistakes.
Here’s how I reflect on mistakes ❤️👇
I had never heard of a postmortem until I started building software.
Now I do them at least once a month 🤓
They let you objectively look at all the facts of a situation, identify what went wrong, and what can be done better next time.
Decisions.
New features.
Launches.
Processes.
Major bugs.
Investments.
Team member departures.
Sales wins and losses.
Meals.
Anything!
1. Purpose
2. The Situation
3. The results
4. What went wrong
5. What we can do better next time
Step 1. Start with the purpose
Why are you doing this analysis in the first place?
What are you looking to understand?
This will help guide the rest of your analysis.
Step 2: The situation. The unbiased facts about everything that happened.
Who was involved?
What happened?
What decisions were made?
Why were those decisions made?
All the facts, unbiased and without color.
Step 3: The results. What was the result? This is still an unbiased analysis of what happened.
For example, if you launched a new feature...
- How many new sign ups were there?
- What was the feedback?
- What were your goals and did you hit them?
Step 4: What went wrong. This is where you get to bring out all the 💩.
All the juicy, gruesome details of what didn’t go well.
Go as deep as you tolerate. Be brutally honest.
Step 5 (the last one!): What we can do better next time.
This is where you distill all your learnings from the other sections.
Based on what you learned, what can you do better?
Come up with clear, actionable changes that can be made next time.
The point isn’t to blame anyone, it’s to empower the team to do better.
If you’re worried about sharing, go back and make sure it’s unbiased.
They help you reflect and quickly shift.
@hnshah and I recently did a postmortem. Read about what we learned here on @hackernoon: hackernoon.com/what-winning-s…