Exploit the Constraint
also Optimize the Constraint · Exploiting the Constraint
Wring the most useful output from the constraint as it currently is, before spending anything to add capacity to it.
Exploiting the constraint is the second of Goldratt’s five focusing steps, coming right after identifying the system’s constraint (or bottleneck). To exploit means to get the most useful output from the constraint exactly as it stands, without yet buying more capacity. The standard example: if a machine is the bottleneck, make sure it never sits idle — don’t let everyone break for lunch at once, don’t starve it of work, don’t run scrap or defective material through it. Every hour the constraint is idle or wasted is throughput the whole system loses and can never recover, since by definition nothing else in the system can make up the difference.
CF treats this as a model of cheap-before-expensive prioritization that generalizes far beyond factories. The ordering matters: you exploit, then subordinate everything else to the constraint, and only then elevate it by adding capacity. Skipping ahead to spend money is a common error, because adding capacity elsewhere — or even at the constraint while it’s still being squandered — cannot raise throughput. First extract the free gains.
This connects to CF’s broader claim that under one percent of things are worth focusing on: in a chain of dependent events, performance is set by the weakest link, so improving anything else is irrelevant (local optimization). Exploiting the constraint concentrates effort on the one factor that actually governs the result, mirroring how CF attacks the decisive factor in decision making generally.