The Theory of Constraints (TOC) is a management approach that says every system is limited by a single biggest constraint — a bottleneck — and that improving the whole system requires focusing all improvement effort there. Developed by Eliyahu Goldratt, TOC provides five focusing steps to systematically identify and elevate constraints.
The Theory of Constraints identifies that every system has a limiting factor, and by pinpointing and managing this single bottleneck, organizations can significantly improve performance and efficiency. Applying the Five Focusing Steps—identify, exploit, subordinate, elevate, and repeat—offers a structured method for continuous improvement and maximizing throughput.
What Is the Theory of Constraints?
The Theory of Constraints is a management and continuous-improvement philosophy based on a simple systems idea: the performance of a system is limited by one or a very small number of constraints. Improving a non-constraint may make that local process look better without increasing the output of the system as a whole.
A familiar analogy is a chain. Strengthening a link that is already stronger than the weakest link does not increase the strength of the chain. TOC therefore directs improvement toward the limiting factor first.
In manufacturing, the constraint may be a machine, production step, skilled labor resource, inspection operation, or material supply. But TOC is not limited to factory bottlenecks. A constraint can also be a policy, management rule, sales capability, market demand, project resource, distribution rule, or another factor that limits progress toward the organization’s goal.
The TOC Institute describes TOC as a practical body of knowledge encompassing the five focusing steps, throughput accounting, thinking processes, and applications for production, supply chains, projects, sales, marketing, and other functions.
For a for-profit organization, traditional TOC literature frames the overarching goal in financial terms: making more money now and in the future. Operational improvement is therefore evaluated by whether it improves the performance of the whole system, not merely whether one department becomes more efficient.
This systems perspective explains why TOC can produce counterintuitive decisions. A non-constraint machine may deliberately sit idle rather than create excess work-in-process. A company may spend extra money protecting the constraint from defects. Local efficiency can fall while total throughput improves.
Who Developed the Theory of Constraints? (Goldratt, “The Goal”)
The Theory of Constraints was developed by Dr. Eliyahu M. Goldratt, an Israeli physicist, author, educator, and management thinker. TOC became widely known through his business novel The Goal, coauthored with Jeff Cox and first published in 1984.
The Goal follows plant manager Alex Rogo as he tries to save a struggling manufacturing facility. Through the story, readers learn to distinguish bottlenecks from non-bottlenecks, protect constrained resources, reduce unnecessary inventory, and manage the factory as an interconnected system rather than a collection of independently optimized departments.
Lean Enterprise Institute notes that TOC was first published in The Goal and describes its central concept as the existence of at least one factor that limits an organization from achieving more of what it strives for.
Goldratt’s work expanded beyond the original manufacturing application. TOC later developed methods for project management, distribution, supply chain, accounting, sales, strategy, and organizational problem-solving.
One reason The Goal remains influential is that it teaches the concepts through cause and effect rather than presenting TOC as a checklist. The reader sees how traditional measures such as high machine utilization and large batches can create inventory without improving the plant’s actual output.
The Five Focusing Steps of TOC
The five focusing steps are the core process of ongoing improvement in TOC. They create a repeating cycle for finding the system constraint, maximizing its contribution, aligning other resources with it, increasing its capacity when necessary, and then beginning again when the constraint changes.
| Step | TOC action | Management question |
| 1 | Identify the system’s constraint | What currently limits the system from achieving more of its goal? |
| 2 | Exploit the constraint | How can we get more from the constraint using what we already have? |
| 3 | Subordinate everything else | How should non-constraints operate to support the constraint? |
| 4 | Elevate the constraint | What additional capacity or structural change is justified? |
| 5 | Repeat and prevent inertia | Has the constraint moved, and are old rules now limiting the system? |
Step 1: Identify the constraint
Find the factor that most limits total system performance. In a plant, this may be the workstation whose effective capacity is below the demand placed on it. In a service process, it could be a specialist approval step. In a business with excess operating capacity, the market or sales system may be the constraint.
The key is to identify the system constraint rather than simply the busiest-looking process. Long queues, persistent backlog, missed due dates, high utilization, and work waiting before a resource can provide clues, but data should confirm the diagnosis.
Step 2: Exploit the constraint
Exploit means use the existing constraint as effectively as possible before buying more capacity. Eliminate avoidable downtime, protect its schedule, ensure trained staff are available, perform preventive maintenance intelligently, keep necessary materials ready, and prevent defective work from consuming scarce constraint time.
If a constrained machine loses 30 minutes because material is missing, the system may lose 30 minutes of potential throughput. The same 30-minute delay at a resource with abundant spare capacity may have little effect on total output.
Step 3: Subordinate everything else
Subordination means aligning non-constraints with the decision made for the constraint. Upstream processes should not flood the constraint with unnecessary WIP, and downstream resources should be prepared to receive its output.
This is where TOC challenges local utilization metrics. A non-constraint does not need to run at 100% simply because it can. Producing more than the constraint can process increases inventory and lead time without increasing system throughput.
Step 4: Elevate the constraint
If exploitation and subordination are not enough, increase the constraint’s capacity. Elevation may involve buying equipment, adding a shift, hiring specialized labor, outsourcing some work, redesigning the process, changing the product mix, or making another investment that raises the limiting capacity.
TOC intentionally places elevation after exploitation and subordination. The organization should first recover hidden capacity before spending capital to create more.
Step 5: Repeat and prevent inertia
When a constraint is broken, another factor becomes the new constraint. Return to Step 1 and repeat the process. Goldratt’s formulation also warns against allowing inertia to become the constraint: rules that made sense under the old constraint may become harmful after the system changes.
The five steps are therefore not a one-time bottleneck project. They are a focusing mechanism for ongoing improvement.
The Three Types of Constraints (Physical, Policy, Market)
For practical diagnosis, constraints are often grouped into three broad categories: physical constraints, policy constraints, and market constraints. The categories are useful, but real systems can contain overlapping causes, and TOC practitioners may use more detailed classifications.
Physical constraints
A physical constraint is a tangible resource or capacity limitation. Examples include a machine, test station, production line, warehouse dock, vehicle fleet, specialized technician, raw material, or physical space.
Physical constraints are often easiest to see because work accumulates in front of them. However, the visible queue may be a symptom of a policy or scheduling problem rather than true insufficient capacity.
Policy constraints
A policy constraint is a rule, assumption, measurement system, procedure, or management practice that limits performance. Examples include releasing work too early, rewarding every machine for maximum utilization, enforcing unnecessarily large batches, prioritizing orders by departmental preference, or using a purchasing rule that creates shortages at the constraint.
Policy constraints can be powerful because they shape behavior across many resources. Adding capacity will not solve a bad rule if the rule simply recreates the same problem.
Market constraints
A market constraint exists when the organization has more capacity to sell or deliver than the market is currently buying. The limiting factor may be demand, the value proposition, pricing, sales effectiveness, distribution reach, or the organization’s ability to convert demand into profitable orders.
When the market is the constraint, maximizing factory output is not the answer. TOC’s focusing logic shifts attention toward actions that increase profitable sales while preserving reliable delivery capability.
| Constraint type | Typical example | What limits the system? | Possible response |
| Physical | Heat-treatment furnace | Insufficient effective capacity at a resource | Exploit, protect, then elevate capacity if needed |
| Policy | Large-batch scheduling rule | Management rule creates excess WIP and delay | Change the rule and supporting measures |
| Market | Plant can make 1,000/day but profitable demand is 750/day | Insufficient market demand or sales conversion | Improve offer, market reach, sales, or demand generation |
Constraint Management: Identifying the Bottleneck
Constraint management begins by distinguishing a true system constraint from a temporary bottleneck. A bottleneck is commonly a resource whose demand exceeds its available capacity for a period. The system constraint is the factor currently limiting the system’s ability to achieve more of its goal.
In many manufacturing environments the terms are used interchangeably, but the distinction is useful. A machine may become temporarily overloaded because of a short-term product mix, while a different resource remains the strategic constraint over time.
Look for persistent evidence
Useful indicators include a sustained queue before a resource, chronic overtime, high utilization, frequent expediting, missed schedules linked to the same step, and downstream starvation when that resource stops. Capacity analysis should confirm whether its effective output is genuinely limiting the system.
Measure flow, not just utilization
A resource can report high utilization without being the constraint. Likewise, poor scheduling can create a queue at a non-constraint. Map the flow of work, quantify demand and effective capacity, and observe what happens to total throughput when the suspected constraint gains or loses productive time.
Protect constraint time
Once identified, every minute at the constraint deserves special attention. Perform quality checks before scarce constraint capacity when practical, stage materials in advance, avoid unnecessary changeovers, ensure tooling and information are ready, and schedule maintenance to reduce unexpected loss.
This does not mean unsafe operation or eliminating necessary maintenance. It means recognizing that constraint time has a different economic impact from idle time at a resource with spare capacity.
Avoid shifting the problem with excess WIP
Pushing more work into the plant does not increase constraint capacity. It creates longer queues. The TOC Institute specifically warns that unchecked non-constraints can create bloated WIP, elongated lead times, and expediting when they produce more than the constraint can handle.
That insight leads directly to Drum-Buffer-Rope scheduling.
Drum-Buffer-Rope (DBR): TOC Scheduling Explained
Drum-Buffer-Rope (DBR) is the classic TOC production scheduling method for synchronizing work with the system constraint. The names are metaphors: the drum establishes the pace, the buffer protects that pace from disruption, and the rope controls the release of new work into the system.
Drum: the constraint sets the pace
The drum is the schedule of the capacity-constrained resource. Because the constraint governs the system’s maximum output, its planned sequence and rate become the rhythm around which the rest of production is coordinated.
If a constrained finishing operation can reliably complete 40 units per hour, releasing work upstream at a sustained rate of 60 units per hour does not create 60 finished units. It creates a growing queue.
Buffer: protect throughput from variability
A buffer protects the constraint or customer due date from normal variation. In manufacturing DBR, this is often conceptualized as time protection rather than simply a pile of inventory. Work is released early enough that normal upstream variation should not starve the constraint.
Buffers also create management priorities. Work that is consuming too much of its protective buffer deserves attention before work that remains safely on schedule. Buffer management therefore helps focus expediting on the few orders that actually threaten throughput or delivery.
Rope: control material release
The rope is the communication or control mechanism that ties release of new work to the drum. Instead of releasing every order to the floor immediately, the system releases work according to the constraint schedule and required buffer.
The effect is less WIP, shorter queues, and a production system whose input is synchronized with its limiting capacity.
Simple DBR example
Suppose a four-step process has hourly capacities of 60, 55, 40, and 70 units. Step 3 is the constraint at 40 units per hour. Under conventional push scheduling, Step 1 may release 60 units each hour, causing inventory to accumulate before Step 3.
With DBR, the 40-unit-per-hour capability of Step 3 becomes the drum. A protective buffer is established so Step 3 is not starved by normal upstream variation. The rope regulates release at the front of the process so material enters at a rate consistent with the constraint schedule rather than the 60-unit capacity of Step 1.
| DBR element | In the example | Purpose |
| Drum | Step 3 schedule: 40 units/hour | Sets the system’s production rhythm |
| Buffer | Protective time/work ahead of Step 3 | Prevents normal variation from starving the constraint |
| Rope | Release signal tied to Step 3 schedule | Stops upstream resources from flooding the system |
DBR does not make every resource operate at the same utilization. Non-constraints need protective capacity so they can recover from variation and support the drum.
TOC Thinking Processes (Evaporating Cloud, Current Reality Tree)
Not every constraint is a machine. When the limiting factor is a policy, conflict, assumption, or complex organizational problem, TOC uses a family of logical problem-solving tools called the Thinking Processes.
The TOC Institute describes these tools as cause-and-effect methods designed to answer three questions: What to change? What to change to? How to cause the change?
Current Reality Tree (CRT)
A Current Reality Tree maps undesirable effects and traces their cause-and-effect relationships toward underlying root causes or a core conflict. Instead of treating late orders, excess WIP, overtime, and priority changes as unrelated problems, a CRT asks whether they emerge from the same deeper system condition.
The value of the CRT is focus. If several visible problems share a common cause, resolving that cause can improve multiple outcomes at once.
Evaporating Cloud (EC)
The Evaporating Cloud, also called a conflict cloud, structures a conflict in which two apparently necessary actions or requirements seem incompatible. The tool surfaces the assumptions connecting each side to the desired objective.
The goal is not to split the difference automatically. It is to identify an assumption that can be challenged so the apparent conflict “evaporates” and a win-win solution becomes possible.
For example, production may believe it must run large batches to keep unit costs low, while customer service needs small batches to shorten lead time. The cloud helps expose assumptions behind both requirements, such as setup time being fixed or local utilization being equivalent to profitability.
Other TOC Thinking Process tools
The broader classical set includes the Future Reality Tree, Negative Branch Reservation, Prerequisite Tree, and Transition Tree. The Future Reality Tree tests whether proposed changes will produce the desired effects; Negative Branch Reservation anticipates harmful side effects; the Prerequisite Tree identifies obstacles and intermediate objectives; and the Transition Tree sequences actions from the current state toward the desired future state.
TOCICO’s published dictionary identifies these tools as part of the Thinking Processes used to address the TOC change sequence.
TOC vs. Lean vs. Six Sigma (Comparison)
TOC, Lean, and Six Sigma all aim to improve organizational performance, but they focus attention differently. TOC asks where the system constraint is and concentrates improvement there. Lean emphasizes value, flow, pull, and the removal of waste. Six Sigma emphasizes reducing variation and defects through disciplined, data-driven process improvement.
ASQ’s comparison of the three approaches describes TOC as using a systems approach, Lean as emphasizing flow and waste reduction, and Six Sigma as emphasizing analytical studies and variation reduction. In practice, the methods are often complementary rather than mutually exclusive.
| Dimension | Theory of Constraints | Lean | Six Sigma |
| Primary focus | System constraint and throughput | Waste reduction, flow, pull, value | Variation, defects, process capability |
| Core question | What limits the system’s goal? | What prevents value from flowing efficiently? | What causes unacceptable variation or defects? |
| Typical roadmap | Five focusing steps | Lean principles, value-stream improvement, kaizen | DMAIC / statistical problem solving |
| Representative tools | DBR, buffer management, thinking processes | 5S, Kanban, takt, standardized work, VSM, SMED | Control charts, capability, hypothesis tests, DOE, measurement analysis |
| Optimization level | System first | Value stream / system flow | Process and critical outputs, integrated to business goals |
| Best use | Prioritizing where improvement has the greatest system effect | Removing waste and improving flow | Solving variation- and defect-driven problems |
A powerful combination is to let TOC determine where to focus and then use Lean or Six Sigma tools at that constraint. If the constraint loses capacity because of long changeovers, SMED may be the right Lean tool. If it loses throughput because of process variation and defects, Six Sigma analysis may be appropriate.
ASQ has documented combined approaches in which TOC identifies restrictive bottlenecks and Lean Six Sigma tools are then used to exploit or eliminate them. The important principle is to avoid spreading improvement resources equally across processes that do not limit system performance.
Real-World TOC Examples
Manufacturing bottleneck
A plant has cutting capacity of 90 units per hour, machining capacity of 65, heat treatment of 45, and assembly capacity of 80. Heat treatment is the system constraint.
The team first exploits it by staging loads in advance, moving quality inspection upstream, reducing avoidable downtime, and ensuring operator coverage during breaks. Other departments subordinate by releasing only enough work to keep heat treatment protected without building excessive queues. If demand still exceeds 45 units per hour, the company may elevate by adding furnace capacity or outsourcing selected loads.
Healthcare diagnostic imaging
A hospital may discover that MRI availability limits patient throughput for a particular diagnostic pathway. Exploitation could include reducing unused appointment slots, improving patient preparation, standardizing room turnover, and preventing avoidable rescheduling. Other departments subordinate by ensuring patients and required information are ready when scanner time becomes available.
Only after recovering hidden capacity would management evaluate an additional scanner or extended operating hours.
Software delivery
A software organization may have plenty of developers but only a small security-review team authorized to approve releases. Adding developers increases work waiting for security without increasing releases.
TOC would focus on the security-review constraint: improve submission quality, automate repeatable checks, reserve specialist time for work that requires judgment, and limit upstream work-in-process. If the constraint remains after exploitation and subordination, additional qualified review capacity may be justified.
Sales as the market constraint
A factory may have capacity for 10,000 units per month but profitable demand of only 7,000. The plant is not the constraint; the market or sales system is. Increasing machine speed would create more unused capacity.
Improvement instead focuses on increasing profitable demand, strengthening the offer, expanding channels, improving sales conversion, or identifying market segments that value the organization’s capabilities.
Project environment
In multi-project environments, a scarce specialist or shared resource can constrain project completion. TOC’s project-management application, Critical Chain, focuses on resource dependencies and protective buffers rather than padding each task independently. The same systems principle applies: protect the factor that governs completion and control the release of work.
Common TOC Implementation Mistakes
Mistake 1: Treating every bottleneck as equally important
Organizations often launch improvement projects everywhere they see a queue. TOC requires identifying the system constraint – the factor whose improvement changes total system performance. Fixing a local inconvenience at a non-constraint may produce no additional throughput.
Mistake 2: Jumping directly to more capacity
Buying another machine or hiring more people corresponds to elevation, Step 4. Skipping exploitation and subordination can waste capital because existing capacity may be lost to downtime, poor scheduling, defects, missing materials, or bad policies.
Mistake 3: Maximizing every resource’s utilization
Running non-constraints continuously creates excess WIP when their output exceeds what the constraint can process. TOC deliberately accepts idle time at non-constraints when producing more would not increase system throughput.
Mistake 4: Starving the constraint
The constraint should not wait unnecessarily for material, information, operators, tooling, maintenance response, or quality decisions. A small upstream disruption can become a system-level throughput loss when it starves the limiting resource.
Mistake 5: Letting defects consume constraint capacity
Whenever practical, detect defects before expensive constraint processing. A defective unit that uses scarce constraint time and is later scrapped has consumed capacity that could have produced a saleable unit.
Mistake 6: Using too much inventory as a buffer
A buffer is protection against variation, not permission to create unlimited WIP. Oversized buffers lengthen lead time and hide problems. Buffer management should distinguish useful protection from excess queue.
Mistake 7: Ignoring policy constraints
A visible machine queue may be caused by scheduling, batch-size, measurement, or release policies. Adding equipment without challenging those rules can leave the true constraint untouched.
Mistake 8: Failing to restart the five steps
Once an improvement breaks the original constraint, the organization must identify the new limiting factor. Continuing to optimize the old constraint after it has moved is the inertia Goldratt warned against.
Mistake 9: Treating TOC as incompatible with Lean or Six Sigma
TOC can provide the focus while Lean and Six Sigma provide specialized improvement tools. Rejecting useful methods because they come from another improvement tradition can slow progress. The question should be which tool best improves the current system constraint.
Conclusion
The Theory of Constraints changes the improvement question from “How can every department become more efficient?” to “What currently limits the performance of the whole system?” The five focusing steps provide the discipline to identify that constraint, exploit it, subordinate other resources, elevate it only when necessary, and repeat the process when the constraint moves. Drum-Buffer-Rope translates the same logic into production scheduling, while the TOC Thinking Processes address policy conflicts and complex cause-and-effect problems. Lean and Six Sigma can strengthen TOC implementation by supplying specialized tools for flow, waste, variation, and quality at the point where those improvements matter most. Used consistently, TOC turns bottleneck management into a system-wide process of ongoing improvement rather than a series of disconnected local projects.
Frequently Asked Questions
What is the Theory of Constraints?
The Theory of Constraints (TOC) is a management approach that says every system is limited by one
biggest constraint — a bottleneck — and that improving the whole system requires focusing effort there. Developed by
Eliyahu Goldratt in ‘The Goal,’ TOC provides five focusing steps for improvement.
Who developed the Theory of Constraints?
Israeli physicist and business consultant Eliyahu M. Goldratt developed the Theory of Constraints,
introduced in his 1984 business novel ‘The Goal.’ Goldratt expanded TOC through subsequent books and consulting
work until his death in 2011.
What are the five focusing steps of TOC?
1. Identify the system’s constraint (bottleneck). 2. Exploit the constraint (maximize its throughput). 3.
Subordinate everything else to the constraint’s pace. 4. Elevate the constraint (invest to increase capacity). 5. Return to
step 1 as a new constraint emerges.
What is Drum-Buffer-Rope (DBR)?
Drum-Buffer-Rope is TOC’s production-scheduling method. The ‘drum’ is the constraint (it sets the pace).
The ‘buffer’ is a small time-based inventory protecting the constraint from starvation. The ‘rope’ is the signal from the
drum back to material release, so materials enter at the drum’s pace.
What is the difference between TOC, Lean, and Six Sigma?
TOC focuses improvement on the single biggest bottleneck. Lean focuses on eliminating waste across the
whole value stream. Six Sigma focuses on reducing variation and defects. All three can complement each other —
Goldratt himself argued TOC should sequence Lean and Six Sigma efforts to focus them on constraints first.
What are the three types of constraints in TOC?
Physical constraints are equipment, workforce, or material limits. Policy constraints are internal rules,
procedures, or metrics that block flow (often more damaging than physical constraints). Market constraints exist when
demand — not capacity — is the limiter, requiring sales or product-strategy fixes.
