after we build — closing
The Loop
Five volumes. One cycle. The understanding travelling from person to code and back again — with a little less lost at each crossing.
The loop does not close at the retrospective meeting. It closes when the model update is written — when what was learned is somewhere permanent that survives the conversation.
The initiative brief for the next cycle starts from that updated state. The discovery questions are sharper. The prediction is better calibrated. The amigos template has one more prompt.
What the client experiences
None of this is visible to the client as process. They experience it as something simpler: the team understands us better than they did last quarter. The software fits more precisely. The conversations are shorter. The surprises are fewer.
That is what compounding looks like from the outside.
From the inside, it is a set of honest habits — writing the prediction before the cycle, checking it after, updating the model, carrying the update forward. Each one costs almost nothing. Together, over time, they produce the condition under which genuinely good work becomes possible.
The system does not get better because people try harder. It gets better because every cycle leaves behind a more accurate version of reality.
If the chain holds
If the chain holds, the team learns.If the chain breaks, the team repeats.
And the difference between those two is the difference between a team that scales and a team that stays busy.
Five volumes. One discipline. The understanding travels from the person who has the problem to the code that changes their situation — and back again, carrying what reality taught. Each crossing loses a little less than the last. Each cycle starts from a less wrong place. Each prediction aims closer.
That is the whole practice. That is why all of this matters.
Back to Volume V cover → · Back to the five volumes → · The map →