Mechanism
Every relieved bottleneck reveals the next one, but investment continues to target the old constraint.
Enabling conditions
- Several resources are scarce
- Constraint maps update slowly
Failure modes
- Non-constraints are optimised
- Buffers disappear into routine demand
Early signals
- • Improvement does not change throughput
- • The same resource remains the investment target
Countermeasures
- • Reassess after each change
- • Protect critical buffers
Diagnostic questions
- 1.What limits the whole system now?
- 2.Where will the constraint move next?