Noise. Same information, private math, results that never meet. The room isn't as aligned as it thinks it is.

Bias. Judgment pulled the same direction by what you already expect. Optimism travels better than caution — every time.

Incentives. Positional weight writes the equation before anyone speaks. That's not cowardice. That's arithmetic.

Accumulation. Add and never subtract. Every addition looks reasonable, until the honest question has forty variables instead of four.

In your last big decision, did every function reading the same data get to the same conclusion? Or did you find out weeks later that people had been quietly reading it differently?

When soft evidence came back against a thesis you'd already committed to, did the strategy change — or just the story around it?

Who in your room can disagree with the most senior person without paying a price? If you can't name three people, that's your answer.

In the last twelve months, has your company subtracted anything of consequence? Not deprioritized. Stopped.

Have the problem framed by someone with no stake in it. Start the discussion with the most junior person in the room. Let the heaviest voice go last. Three rules. Install them Monday. Nobody anchors the room, because nobody upstream has told the room what to think yet.

Before the discussion opens, everyone writes their number, their pick, their forecast — privately, at the same time — then everybody shows. Ninety seconds. The most dangerous data in your company is already computed, sitting in the heads around the table.

Before the commitment, write down the evidence that would mean the thesis is failing. Put a date on the calendar. When the date arrives, you look — no re-litigating, no "let's give it one more sprint." The trigger was set by the version of you with clean incentives. A pre-committed trigger freezes the price of the honest question.

Once a quarter, the meeting that approves new initiatives has to name what stops — stopped, not "deprioritized." Cap the number of simultaneous bets. The number matters less than the existence of a number, because a cap forces the trade-off conversation that adding never does.