⛓️Five focusing steps

Goldratt's procedure for improving a system that has a 🧩 Bottleneck. It is the method the plant manager stumbles into over the course of 📕 The Goal.

1. Identify the constraint. Look for the pile of waiting work.
2. Exploit it. Get everything out of it as it stands: no idle time, no work on things that will be thrown away, nothing defective fed into it.
3. Subordinate everything else to it. The other steps run at the constraint's pace, even if that leaves them idle.
4. Elevate it. Only now spend money: more capacity, more people, a second machine.
5. Repeat. The constraint has moved. Find it again, and do not let habit become the new one.

The order is the lesson. Everyone wants to begin at step 4. Steps 2 and 3 are free and feel like doing less.

Use when: work is visibly queueing in one place.

Not for: a system whose constraint is a decision nobody will take. There is no pile in front of a decision.

In the launch: step 1 took me until 9 February, and the answer was QA. Step 3 would have meant not starting design work that testing could not receive. I started it anyway.

What I decided afterwards: 💎 A bottleneck sets the pace for everything upstream of it.