In one line: Every system has at least one constraint that limits its output. Improving anything else is an illusion, so find the constraint, get the most from it, align everything around it, then raise it, and repeat.
Core idea
Eliyahu Goldratt introduced the Theory of Constraints through a novel, The Goal [1]. A plant manager, guided by a mentor who teaches only by asking questions (see Socratic Method), discovers that his factory’s goal is simply to make money. That goal is measured with three numbers:
- Throughput: the rate at which the system generates money through sales, not just production.
- Inventory: money tied up in things the system intends to sell.
- Operating expense: money spent turning inventory into throughput.
The aim is to raise throughput while lowering inventory and operating expense. The main obstacle is the constraint, or bottleneck: the resource whose capacity limits the whole system. An hour lost at the bottleneck is an hour lost for the entire system; an hour saved anywhere else is a mirage.
How it works
The five focusing steps:
- Identify the system’s constraint.
- Exploit it: get the most out of it with existing resources. Never let it sit idle, and never let it work on defective parts.
- Subordinate everything else to it: pace the rest of the system to the constraint’s rate, not to their own maximum.
- Elevate it: if it’s still limiting, add capacity, such as another machine, more people or an alternative route.
- Repeat. Once the constraint moves, start over, and don’t let inertia become the new constraint.
Drum-buffer-rope puts this into practice:
- the constraint sets the pace (the drum);
- a small inventory protects it from upstream hiccups (the buffer);
- material release is tied to the constraint’s consumption (the rope).
In the novel, the scout troop’s slowest hiker is a memorable illustration: the troop moves at the pace of its slowest member, so he goes to the front and his load is shared out.
In practice
TOC applies far beyond factories. An engineering project’s constraint might be:
- a single test chamber
- one expert reviewer
- a supplier
- or an ill-defined process that keeps changing, which stalls everyone downstream and saps morale
It applies personally too. With no car, for instance, the constraint is mobility. You exploit it with planned public-transit routes, subordinate other choices to it (where you live and shop), elevate it by saving for a car, and then look for the next constraint. The program’s questions push the same habit: name a bottleneck you’ve seen, then walk it through the five steps.
Questions to ask yourself
- What is the constraint on my team’s output right now? (Hint: where does work pile up?)
- Are we keeping the constraint busy on the right work, or letting it wait?
- Which “efficiency” improvements are we making at non-constraints that don’t change the output?
Connections
- Why it happens: Dependent Events and Statistical Fluctuation. How it’s taught: Socratic Method.
- Related methods: Lean, Throughput and Timing Analysis. Systems view: Leverage Points.
- Capabilities exercised: Decision Making, Resourcefulness, Responsibility and Urgency to Deliver.