On this page
Use the headings in the article to move through the guide.
A bottleneck is the point in a process where work accumulates because that step cannot keep up with the flow of demand. A constraint is the factor that most limits the performance of the whole system. Sometimes the constraint is a machine or person; sometimes it is a policy, approval rule, supplier, system, market demand or missing information.
The direct answer is this: identify bottlenecks and constraints by following the work, looking for queues, measuring flow and asking which factor most limits the process outcome. Improving a non-constraint may make one area look better without improving the whole business.
Bottleneck vs Constraint
| Term | Meaning | Example |
| Bottleneck | A visible point where work waits or piles up. | All quotes wait for one estimator. |
| Constraint | The factor currently limiting the system goal. | Estimator capacity, approval policy, missing customer data or market demand. |
| Symptom | What people notice. | Customers chase for updates. |
| Root limitation | What actually restricts flow. | Incomplete inputs require repeated clarification. |
Constraint Detection Checklist
- Where is work waiting longest?
- Which step is always busy while others wait?
- Which person, system or approval controls release of work?
- Where do urgent jobs queue behind routine work?
- Which step causes downstream starvation or upstream pile-up?
- What would increase total output if improved?
Five Focusing Steps Adapted for SMEs
| Step | SME question |
| 1. Identify | What is limiting the whole process now? |
| 2. Exploit | How can we use the existing constraint better before spending money? |
| 3. Subordinate | How should other work be aligned to protect the constraint? |
| 4. Elevate | What extra capacity, skill, system or rule change is needed? |
| 5. Repeat | What becomes the next constraint after improvement? |
Decision Tree: Is This the Real Constraint?
| Question | If yes | If no |
| Does work pile up before this point? | Potential bottleneck. | Look elsewhere. |
| Would improving this step increase total process output or reliability? | Likely constraint. | It may be local inefficiency only. |
| Is the step blocked by policy, information or authority rather than capacity? | Investigate non-resource constraint. | Measure physical or people capacity. |
| After improvement, would the queue move somewhere else? | Plan next constraint review. | Recheck process boundary and demand. |
Panith Illustrative Case Study
Panith Illustrative Case Study: A small metal fabrication workshop believes welding is the bottleneck because parts often wait near the welding bay. Observation shows that welders frequently stop because drawings are unclear and materials are missing. The visible bottleneck is welding, but the constraint is upstream information and material readiness.
Common Mistakes
- Improving the busiest-looking area without checking system impact.
- Confusing a queue with the true cause of the queue.
- Adding people before fixing missing inputs or poor rules.
- Keeping every department fully utilised even when it overloads the constraint.
- Failing to repeat the analysis after the constraint moves.
References and Further Reading
- Goldratt Research Labs: Introduction to Theory of Constraints
- ASQ: Quality and the Theory of Constraints
- Lean Enterprise Institute: Theory of Constraints and Lean Thinking
Next Lesson
Continue with Lesson 16: Finding Delays, Rework and Process Gaps.
SEO improvement pass: expanded depth and practical application
Expanded Practical Guidance
The real constraint is the point where improvement would most increase the performance of the whole process. It may not be the loudest problem or the busiest person. A visible queue is a clue, not a final diagnosis.
Before adding capacity, check whether the constraint is being protected. A specialist who is the constraint should not spend time searching for information, chasing approvals or correcting avoidable errors. Use the constraint for the work only they can do.
Constraint Protection Matrix
| Constraint waste | Example | Countermeasure |
| Starved constraint | Estimator waits for missing customer details. | Improve intake completeness. |
| Blocked constraint | Manager approval waits in email. | Create approval rule and deadline. |
| Misused constraint | Skilled technician does admin that others can do. | Move admin away from the constraint. |
| Overloaded constraint | All work treated as equally urgent. | Create triage and scheduling rules. |
Practical Next Step
Identify one suspected constraint and list every non-essential activity it performs. Removing those activities may improve flow before hiring, buying equipment or adding software.
Visual Learning Aids
The following visual structures translate this lesson into practical management tools. They are designed for SME use and can be copied into a workshop, team meeting or improvement plan.
1. Process Diagram
| Input | Work system | Output | Owner | Main risk |
|---|---|---|---|---|
| Customer need or business trigger | Identifying Bottlenecks and Constraints applied to the business process | Reliable customer or internal outcome | Process owner | Rework, delay or quality failure |
2. Flowchart
[Start] Trigger received | v [Understand] Work is mapped | v [Decide] Bottleneck or waste is identified | v [Act] Improvement option is selected | v [Review] New way of working is tested | v [Improve] KPI review confirms next action
3. Decision Tree
Is the issue clear and evidence-based? - No: clarify the problem, collect examples and involve the people closest to the work. - Yes: does it affect customers, cost, quality, risk or growth? - No: document the learning and monitor lightly. - Yes: assign an owner, choose a practical method, define success measures and review progress.
4. Priority Matrix
| Low effort | High effort | |
|---|---|---|
| High impact | Do first: quick operational improvement | Plan carefully: strategic improvement project |
| Low impact | Do only if it removes friction | Avoid or defer unless required for risk, compliance or learning |
5. Implementation Timeline
| Stage | Typical SME timing | Output |
|---|---|---|
| Diagnose | Week 1 | Problem statement and evidence |
| Design | Week 2 | Chosen method, owner and measures |
| Test | Weeks 3-4 | Small pilot or controlled trial |
| Standardise | Weeks 5-6 | Checklist, SOP, dashboard or decision rule |
| Improve | Monthly | Review notes and next improvement action |
6. Illustrative Performance Chart
Illustrative example: replace these sample scores with real business data before using the chart for decisions.
| Measure | Current score | Target score | Visual gap |
|---|---|---|---|
| Clarity | 2/5 | 4/5 | 2 blocks to 4 blocks |
| Ownership | 2/5 | 5/5 | 2 blocks to 5 blocks |
| Measurement | 1/5 | 4/5 | 1 block to 4 blocks |
| Consistency | 3/5 | 4/5 | 3 blocks to 4 blocks |
7. Maturity Model
| Level | Maturity stage | What it looks like | Next improvement |
|---|---|---|---|
| 1 | Informal work | Informal work is visible in the way the business manages this topic. | Documented process |
| 2 | Documented process | Documented process is visible in the way the business manages this topic. | Measured process |
| 3 | Measured process | Measured process is visible in the way the business manages this topic. | Controlled process |
| 4 | Controlled process | Controlled process is visible in the way the business manages this topic. | Continuously improved process |
| 5 | Continuously improved process | Continuously improved process is visible in the way the business manages this topic. | Sustain and refine through periodic review |
8. Comparison Table
| Dimension | Current-state process | Future-state process |
|---|---|---|
| Decision basis | Experience, urgency or individual memory | Evidence, agreed criteria and visible trade-offs |
| Accountability | Unclear or dependent on the founder | Named owner with clear responsibility and review rhythm |
| Measurement | Discussed only when something goes wrong | Tracked using cycle time, first-time-right rate, cost and customer impact |
| Scalability | Works only while the team is small | Can be taught, repeated and improved as the business grows |