4. Five principles

§Five rules recur at every gate of the collection. They appear in the Sprint as field discipline, in Strategy as rules for a room, in Experimentation as the anatomy of a loop. Read together they are the operating principles of the whole thing, and each one exists because of a moment where its absence cost something.

§ 4.1Write before you speak#

§Every exercise starts in silence, on paper or a board, then is read, then voted, then discussed. The fourth whitepaper made it a rule for the strategy room; the reason is the same everywhere.

§In the field sprint, the most talked-about theme was paying and splitting the bill. The most written-about theme was recognition, twenty-six notes against thirteen. Talking and writing produced different rankings, and only the writing was counted. The team that had ranked payments first out loud had, in silence, ranked recognition first all along.

§ 4.2Prototype before you ask#

§"Would you use an app that recognizes you when you arrive?" always gets a yes. It is a free compliment and costs the person nothing. Put something in their hands instead and watch where they get stuck.

§The customer who could not understand the loyalty unit would have said yes to loyalty in the abstract. She could not say yes to the screen in front of her, and the screen was the truth. This is the Sprint's ladder of fidelity in one sentence, and it is the reason the third whitepaper spends so long on the difference between what people say and what they do.

§ 4.3Only build what came from a question#

§AI makes it trivial to add a feature. That is the danger, not the benefit, and it is the principle this whole whitepaper is built around.

§The rule is that nothing enters the prototype unless it traces to a question the team wrote before anyone built. In the field sprint the three features that failed hardest were the three that came from nowhere: a loyalty unit that looked clever, a check-in gesture borrowed from an airport, a payment card lifted from a fintech screenshot. Each cost one prompt to add. Each cost an interview to discover it was wrong.

§ 4.4Same-day synthesis, or it did not happen#

§Interviews are synthesized before the team leaves the building. One page per person: verbatim quotes, where she got stuck and for how long, what she expected the unit of value to be, what she agreed to at the end and with what face. The sixth whitepaper carries the same rule into after-launch reading: the number is instant, the meaning is not, and the meaning can only be set by the person who was there, on the day.

§The next morning, the face someone made when she said yes is gone. So is the pause before she said it. Memory keeps the words and drops the evidence.

§ 4.5Open is a valid state; undecided is not#

§When a block runs over, close it with what you have and mark what is open, with an owner and a date. The fourth whitepaper made this a rule for the strategy room and the third for the sprint's close; both times it means the same thing.

§In the field sprint, demo day ended with six open decisions on the board. Each had a name next to it. None was resolved that day, and that was fine, because each was a task rather than a mood.

§ 4.6None of them is about AI#