Theory of Constraints (TOC)
Theory of Constraints holds that every system has exactly one bottleneck limiting its output at any time, and improving anything else is wasted effort until you fix that one constraint.
Five steps circle round in order, arriving back at the top the moment a fixed constraint gets replaced by a new one.
Reach for this when…
- You're improving efficiency everywhere and output still isn't moving.
- One step in the process is clearly the reason everything else waits.
- You need to decide where the next investment in capacity actually goes.
How to run it
- Identify the constraint limiting the whole system.
- Exploit it: get everything possible out of it before spending on it.
- Subordinate every other step to keep the constraint fed and never idle.
- Elevate the constraint: invest to increase its capacity.
- Repeat: once it moves, find the next constraint.
A worked example
Situation. Budi Santoso managed an electronics assembly plant near Surabaya, Indonesia that had spent a year improving efficiency on every line except one, while output barely moved.
Applied. He mapped the full process and found one soldering station, running at full capacity with a queue in front of it every shift, was the real constraint. He first stopped it ever sitting idle, then added a second shift there alone once that wasn't enough.
Result. Overall output rose within a month, more than a year of scattered efficiency drives had achieved. Budi now checks for the new bottleneck every quarter instead of assuming last year's fix still holds.
The catch
TOC works cleanly on a linear production line; it gets murkier in complex service or knowledge work where the constraint moves or is a person, not a machine, and naming it becomes political. Fixing the constraint also just moves the bottleneck elsewhere, which some teams treat as failure instead of the expected next step.
The moment you fix a constraint, another one appears somewhere else. That's not the process failing, that's the process working.
Origin: Eliyahu Goldratt