Murder Board

Thinking Tools5.3 Validating Impact and Execution

From United States military and government briefing practice; the term is Pentagon usage, and NASA credits Hans Mark with building it out of the tiger team.

Puts a single brief, estimate, or plan in front of a panel chosen to question it hard, so every weak point surfaces in rehearsal before the real audience sees it.

A market-entry brief in front of a panel chosen to break it

The reader sits on the panel with twenty minutes on the clock. At each turn, pick the question: the one that tests what the brief rests on, or the friendly one the board will not ask. The authors' answer shows, and the bar counts the hard questions asked before the board gets its turn. Illustrative.

Panel

Brief's authors

Panel

Brief's authors

Select a stage to read it, or play the sequence.The brief. A grocery chain's strategy team proposes entering the next state: twelve stores in two years, a 30 percent share in its towns, payback in four years. The board sees it in ten days.First question. Pick the panel's first question.Pick a move to see the reply.Of course. The region has 1.4 million households, and...A tour: the authors repeat the brief, and nothing is tested.One consultant's report. Nothing else, to be honest.A weak point: the main number rests on one unverified source.Weak point logged. The authors write it down without arguing. The fix is a second source, or a lower share.Second question. Pick the next question.Pick a move to see the reply.We didn't model a response. Roughly seven years, not four.A weak point: the plan assumes the rival does nothing.Next September, if permits go smoothly.A fact check: useful, and the board will not stop there.Second weak point. Logged: a case with a rival price cut, and what payback looks like under it.Third question. Pick the next question.Pick a move to see the reply.Mostly permitting, which we've covered on slide 14.An invitation: the authors name the risks they already handled.It rises 15 percent on a schedule already in law. We used today's rate.A weak point: a known cost increase left out of the model.Third weak point. Logged: rerun labor cost at the scheduled rates.Last question. Pick the panel's last question.Pick a move to see the reply.The share figure. If that falls, so does everything after it.The authors name their own weakest point, in their words.Thank you. We feel good about it.A warm close: the panel leaves the authors more confident than they should be.The revisions. A second source for share, a rival price-cut case, and scheduled wage rates. A fresh panel reviews the revised brief before the board does.

Hard questions asked0of 4

  1. The brief (Brief's authors). A grocery chain's strategy team proposes entering the next state: twelve stores in two years, a 30 percent share in its towns, payback in four years. The board sees it in ten days.
  2. First question (Panel). Pick the panel's first question.
    • Can you walk us through the market size slide? Of course. The region has 1.4 million households, and... A tour: the authors repeat the brief, and nothing is tested.
    • The 30 percent share: where does it come from, and what else confirms it? One consultant's report. Nothing else, to be honest. A weak point: the main number rests on one unverified source.
  3. Weak point logged (Brief's authors). The authors write it down without arguing. The fix is a second source, or a lower share.
  4. Second question (Panel). Pick the next question.
    • What happens to payback if the leading local chain cuts prices 5 percent when you open? We didn't model a response. Roughly seven years, not four. A weak point: the plan assumes the rival does nothing.
    • When does the first store open? Next September, if permits go smoothly. A fact check: useful, and the board will not stop there.
  5. Second weak point (Brief's authors). Logged: a case with a rival price cut, and what payback looks like under it.
  6. Third question (Panel). Pick the next question.
    • Great store design. Any risks you'd flag? Mostly permitting, which we've covered on slide 14. An invitation: the authors name the risks they already handled.
    • Labor assumes today's wages. What is the minimum wage there in two years? It rises 15 percent on a schedule already in law. We used today's rate. A weak point: a known cost increase left out of the model.
  7. Third weak point (Brief's authors). Logged: rerun labor cost at the scheduled rates.
  8. Last question (Panel). Pick the panel's last question.
    • If the board asks one question, which answer are you least sure of? The share figure. If that falls, so does everything after it. The authors name their own weakest point, in their words.
    • Thanks, this is very thorough work. Thank you. We feel good about it. A warm close: the panel leaves the authors more confident than they should be.
  9. The revisions (Brief's authors). A second source for share, a rival price-cut case, and scheduled wage rates. A fresh panel reviews the revised brief before the board does.

An example

A regional grocery chain's strategy team had a brief ready for its board: twelve stores in the next state over two years, a 30 percent share in the towns it entered, payback in four years. The team had reviewed it three times, always among themselves.

Ten days before the board meeting, the chief financial officer convened four people who had not seen it, among them an outside lender and the chain's labor relations manager. They had a day with the brief, twenty minutes with the authors, and one job: break it. They asked where the share figure came from (one consultant's report, nothing else). They asked about payback if the local leader cut prices when the first store opened (seven years; nobody had modeled it). They asked what the minimum wage there would be in two years (15 percent higher, on a schedule already in law).

The authors argued with none of it and wrote it all down. The brief that went to the board carried a second share source, a price-war case, and the new wage rates.

Why it works

Authors cannot see the holes in their own work, because the assumptions that matter most are the ones they stopped noticing. Reviews by the same team confirm the brief, and friendly reviewers ask the questions the authors already answered. A murder board changes who asks and why: outsiders, chosen to be skeptical, with the single goal of finding the weak point, under time pressure that mimics the real audience.

The rehearsal moves the damage to a cheaper moment. A weak point found ten days before the board is a revision; the same point found by a board member is a lost decision and a damaged reputation. The discipline on the authors' side matters as much as the panel's: answering a hard question by arguing protects the brief for twenty minutes, while writing it down protects it in the meeting that counts.

The thinking behind it

Military and government briefing practice The murder board comes from United States military and government briefing rehearsal, and no book or author ever codified it. The term is Pentagon usage: the writer William Safire traced it there in 1987. NASA credits Hans Mark, who ran its Ames center from 1969 to 1977, with building the practice out of the tiger team. Using it on a board paper or an investment case borrows that practice for business.

How to run it

  1. Pick the panel. Recruit three to six skeptical people who have not seen the work, ideally ones who think like the real audience.
  2. Give them the product. Send the brief a day ahead, with one instruction: find what will not survive.
  3. Set the clock. Hold the session to the time the real audience will give, or less.
  4. Let them question everything. Sources, assumptions, numbers, and conclusions are all fair game, in any order.
  5. Write down, don't argue. The authors log each weak point the panel finds, with no defense beyond a factual answer.
  6. Revise against the log. Fix each weak point, or prepare a short answer for the ones that cannot be fixed.
  7. Repeat with fresh eyes. If the revision is large, run a second panel; the first one knows the answers now.

Where it goes wrong

The panel is too friendly. Colleagues who like the authors, or report to them, ask polite questions and leave the brief stronger only in its authors' minds. Choose people with no stake in its approval, and tell them their job is to break it.

The authors defend. Every argued answer teaches the panel to stop pushing on that point, and the weakness is still there on the day. Ask the authors to write each challenge down and save their answers for the revision.

The log is never acted on. A session that finds three weak points and leaves the brief unchanged has only rehearsed the embarrassment. Set a date for the revision before the panel meets.

When to use something else

  • Red Teaming. When the plan needs an adversary acting against it over days or weeks, not a panel questioning a document.
  • Premortem. When the team itself should imagine the plan has failed and explain why, before anyone outside sees it.
  • Critical Response Process. When the work is early and its maker needs feedback that improves it, not an attack.

Where it sits in SOLVE eX

Go to the source

Back to the catalog