Focusing on the Constraint
also Focus on the bottleneck · Optimize limiting factors
Directing optimization effort solely at the system's limiting factor, since improving anything else yields no gain toward the goal.
A system’s output is capped by its single weakest part, just as a chain’s strength is set by its weakest link. Goldratt’s Theory of Constraints draws the blunt conclusion: optimization away from the constraint is wasted. Speeding up a workstation that feeds a slower one produces no extra finished goods; the materials still wait at the bottleneck. So you find the limiting factor and pour your effort there, ignoring the rest. This is the core distinction between local and global optimization — a “local optimum” looks like progress up close but does nothing for throughput, the system’s actual success at its goal.
The flip side is excess capacity: in any working system, most factors are already more than good enough, so changing them moves nothing. This is why CF rejects a balanced plant where every part runs at equal capacity — and, by extension, why it rejects building one.
CF’s distinctive move is to carry this principle into epistemology and decision-making (see constraint applied to epistemology). It grounds CF’s refusal to weigh and sum 50 factors into a credence: most factors are non-constraints with margin to spare, so weighted-factor methods err by letting any small change shift the score. Against the Bayesian motto “always update,” CF says: optimize the use of your attention. Give non-constraints a fast pass/fail binary evaluation, reserve detailed analysis for the few factors near a breakpoint, and judge by good enough, not degree. The disciplined version is the five focusing steps.