The Murder Board
Before starting a project, invite smart outsiders whose only job is to tear your two-page plan apart
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 4
- Confidence
- 90%
A pre-mortem-style ritual Palantir ran at the start of every new project. You write a two-page plan (vision, goals, three-month tactics, principles), then invite three or four smart people who know nothing about the project and task them explicitly with destroying it. The design forces you to surface disagreement before you have sunk time into the work.
Origin
Qureshi attributes the concept originally to the army ('I think this is originally an army thing'). He ran it with his team at Palantir and advised newcomers on how to write the principles section.
Core principles
- 01The reviewers' job is adversarial: tear the plan apart, not validate it.
- 02Bring in people with no context — they question the frame you have stopped questioning.
- 03A principle only earns its place if a reasonable person could disagree with it.
- 04Keep the plan short (two pages) so the critique targets thinking, not prose.
How to run it
- 1
Write a two-page project plan
Capture the vision, the goals, the tactics for the next three months, and a section of principles you will follow for this project. Keep it to about two pages.
- 2
Recruit three or four smart, uninformed critics
Invite people you know are sharp but who know nothing about the project. Distance from the details is the feature, not a bug.
- 3
Task them explicitly with destroying the plan
Their assigned job is to tear the plan apart. Framing it as demolition (rather than 'feedback') gives them permission to be genuinely critical.
- 4
Stress-test every principle for disagreement
Reject any principle nobody could reasonably oppose ('move fast'). Replace it with one that makes some smart people say 'wait, why would you pick this? this seems wrong.'
Pro tip If everyone nods at your principle, it is decoration. A real principle draws a line someone would defend the other side of.
Watch out Vague, universally-agreeable principles feel productive but commit you to nothing and guide no decisions.
In the wild
Qureshi repeatedly saw new joiners write principles like 'move fast.' He pushed back that everyone likes move fast, so nobody can disagree with it, which makes it useless as a principle. He steered them toward principles a lot of people would actively object to.
→ Teams produced principles that were genuinely load-bearing and could be argued about, sharpening project direction before work began.
Common mistakes
Inviting reviewers who already know the project
Insiders share your blind spots and your investment in the plan, so they defend the frame instead of attacking it. The critique only works when the reviewers have no context to protect.
Writing unfalsifiable principles
Principles nobody could disagree with ('move fast,' 'be customer-obsessed') pass the murder board trivially and therefore teach you nothing. They give false confidence while providing zero decision-making guidance.
Is it for you?
Best for
Teams or founders about to commit resources to a new project or product bet who want disagreement surfaced early
Not ideal for
Tiny reversible decisions where the cost of assembling critics exceeds the cost of just trying it
From the transcript
“you basically had to organize what they called a murder board for it”
“You invite three or four smart folks you know who don't know anything about the project and their job is just to tear apart your…”
“everyone likes move fast like it's not a good principle actually because nobody can really disagree with this reasonably”
“so you need something that people can disagree with”
From the episode
How Palantir built the ultimate founder factory
Nabeel S. Qureshi (founder, writer, ex-Palantir)