Topic 07 · The Choice
Reality is simpler than the argument.
Not another factory. A conversation between Goldratt and his daughter about the beliefs under the tools. A few causes, not a cloud of them. A conflict rests on an assumption you can challenge. People are trying to do something sensible. The moment you are sure, look again.
Last of the set, after Obvious. This is why the Cloud and the five steps in Flow work when you actually use them.
01
A few causes
Complexity is usually a sign you have not found the core yet. Keep asking why until the number of causes shrinks.
- Flaky deploys
- Dates slip
- Flat velocity
- Review never clears
- Planning fights
- Heroics
Once understood
More is started than the constraint can finish
What people are saying
- Flaky deploys
- Review never clears
- Dates slip
- Context switching
- Quality arguments
- Flat velocity
- Planning fights
- Heroics on call
Eight problems. It feels irreducibly complex. That feeling usually means the cause has not been found yet.
02
You do not have to pick a side
Speed and quality, build and buy, standards and a short queue. The conflict feels necessary because of at least one assumption that is not solid.
Ship faster, and do not lower quality.
Assumption holding it in place: Quality can only be protected by a gate after the coding is finished.
The same move as an Evaporating Cloud in Thinking. Here it is a standing belief: a win that keeps both needs is the ordinary result of looking, not a rare negotiation.
03
People are good
When the behaviour looks bad, the cause is almost always the system, the measure, or what they currently believe. Not a flaw you can do nothing about.
A developer will not pick up reviews.
04
Never say you know
The moment the explanation feels finished, you stop looking. Especially when the results are not matching it.
“The tests are flaky because the environment is bad.”
Intuition usually notices the odd result first. The feeling tells you it matters. Logic is what you use after that, on the assumption, not instead of the other two.
05
The system had more room
When the conflict evaporates, moves that were “impossible” are just available. That is why the answer feels obvious afterwards.
A policy everyone defended
Optional, once the assumption under it is named.
A trade-off with two bad ends
A third move appears: smaller batches, a flag, a different boundary.
No capacity
Some of it was multitasking and a local-efficiency target, not a missing team.
06
One fight, written down
A recurring conflict. Both needs. The assumptions. The weakest one. Then the sentence you were too sure of.
- A debate that has lasted more than a few cycles. Draw the two sides and the shared aim.
- At least three assumptions keeping it alive.
- The weakest, and a small injection. Not a compromise.
- Where someone, including you, said they already knew. One question that could prove it wrong.
- After the change: did Throughput move, and did a new conflict appear? Inertia is the usual next constraint.
This week
0/4
Cheat sheet
- Simple
- A few causes govern what looks tangled. Few is not the same as easy to find.
- Conflicts
- At least one assumption is soft. Challenge it and you stop choosing between two bad options.
- People
- Unhelpful behaviour is usually the system, the measure, or a belief. Start there.
- I know
- Certainty is the signal to write the assumption and the fact that would kill it.
- Partners
- Intuition finds the odd result. Feeling marks what matters. Logic examines the assumption.
- Freedom
- When the conflict goes, moves that looked impossible are just sitting there.
Check yourself
5 questions. No timer.