What Is the Theory of Constraints?
The Theory of Constraints (TOC) is a management philosophy developed by physicist Eliyahu Goldratt and introduced through his 1984 business novel The Goal. Its founding insight is simple but profound: every system has at least one constraint that limits its performance, just as a chain is only as strong as its weakest link. Pouring improvement effort anywhere other than the constraint produces little or no gain in overall output — and may even make things worse by building inventory.
The implication reverses conventional wisdom. Local efficiency (keeping every machine busy) is not the goal; throughput of the whole system is. A non-constraint running at 100% utilization just manufactures inventory that piles up in front of the bottleneck.
The Five Focusing Steps
TOC provides a repeatable cycle for managing any constraint:
- Identify the constraint. Find the single resource or policy that limits throughput — usually the step with the largest queue of work in front of it.
- Exploit the constraint. Squeeze maximum output from it without major investment: eliminate its idle time, ensure it never works on defective material, offload setups, and keep it running through breaks.
- Subordinate everything else. Pace every other resource to the constraint. Non-constraints should produce only what the constraint can consume — no more — even if that means they sit idle.
- Elevate the constraint. If the first three steps haven't broken it, add capacity: another machine, another shift, outsourcing. This step costs money, which is why it comes after exploitation.
- Repeat — don't let inertia set in. Once a constraint is broken, a new one appears elsewhere. Return to step 1 and never let a policy from the old constraint outlive its usefulness.
Drum-Buffer-Rope Scheduling
TOC's production-scheduling mechanism is drum-buffer-rope (DBR), an elegant way to subordinate the system to its constraint:
- Drum — the constraint sets the beat. The whole plant marches to the rate the bottleneck can sustain.
- Buffer — a time buffer of work placed just upstream of the constraint so it never starves for material, even when upstream processes hiccup. Buffers are also placed at shipping to protect due dates.
- Rope — a communication signal that releases raw material into the front of the line only at the rate the drum consumes it. The rope prevents the floor from drowning in WIP.
DBR delivers reliable throughput with far less inventory than traditional scheduling, because work is released in step with the bottleneck rather than pushed in based on a forecast.
Throughput Accounting
TOC rejects traditional cost accounting's emphasis on allocating overhead and maximizing local efficiency. Instead it uses throughput accounting with three global measures:
| Measure | Definition | Goal |
|---|---|---|
| Throughput (T) | Rate the system generates money through sales, minus truly variable cost | Increase |
| Investment / Inventory (I) | Money tied up in things the system intends to sell | Decrease |
| Operating Expense (OE) | Money spent turning inventory into throughput | Decrease |
Net profit relates to these as Profit = T − OE, and Return on Investment as (T − OE) ÷ I. The priority order matters: TOC argues that increasing throughput has unlimited upside, whereas cutting OE and I have natural floors. Every proposed action is judged by its effect on T, I, and OE together rather than on a misleading per-unit "cost."
Little's Law: WIP, Throughput, and Flow Time
One law underpins the quantitative side of flow management. Little's Law states:
L = λW — in operations terms, WIP = Throughput × Flow Time.
Here WIP is the average amount of work-in-process in the system, throughput is the average completion rate, and flow time is the average time a unit spends in the system. The law is remarkably general — it holds for any stable system regardless of the distribution of arrivals or service times.
A worked example: a workstation completes 20 units per hour (throughput) and there are, on average, 5 units in process (WIP). Then flow time = WIP ÷ throughput = 5 ÷ 20 = 0.25 hours = 15 minutes per unit. The practical consequence for TOC is powerful: for a fixed throughput, cutting WIP directly cuts flow time. This is exactly why the rope caps WIP — less inventory means shorter, more predictable lead times. Explore the relationship on the Little's Law calculator.
TOC and the Broader Lean World
TOC, lean, and Six Sigma are complementary rather than competing. Lean removes waste across the whole stream; Six Sigma reduces variation; TOC focuses scarce improvement effort on the one constraint that matters most right now. A common hybrid sequence is to use TOC to find where to improve, then deploy lean and Six Sigma tools at that constraint for maximum leverage. A value stream map is often the fastest way to spot the constraint — it appears as the largest pile of WIP.