The System of Profound Knowledge – Deming’s Four Lenses [Part 2]

Variation is at the heart of every process.

Whilst organisations rely on the use of data to generate metrics that support performance monitoring and decision-making, the understanding of variation within a statistical context is often overlooked.

How processes and their possible variations are understood alongside data can have a direct impact on decision-making and the way results are interpreted.

This mindset shift transforms how leaders ask questions, moving away from blame and towards meaningful learning and systemic improvement.

This is the second article of the fortnightly six-part mini-series:

1. Why the System of Profound Knowledge matters

2. Knowledge of Variation

3. Appreciation for a System

4. Theory of Knowledge

5. Psychology

6. How SoPK helps leaders think differently about management and improvement

I cover the key aspects of one of the 4 elements of Deming’s System of Profound Knowledge: the Knowledge of Variation.

Let’s explore:

• Key characteristics of the Knowledge of Variation

• The types of variation through use cases

• Management actions and improvement strategies

Let’s dive in! 🤿

Overreacting to natural variation creates more waste, not improvement.

To manage effectively, leaders must understand where variation comes from, how to distinguish its types, and what actions truly improve performance without falling into the trap of overreacting.

Key characteristics of the Knowledge of Variation are:

• Every process shows variation in products, services or performance.

• There are two types of variations:

  • Common-cause: inherent, predictable, within system boundaries.

  • Special-cause: unusual, external, or unexpected events.

• Treating common-cause as special-cause (or vice versa) leads to wrong solutions and wasted effort.

• Statistical tools help visualize variation, separate signal from noise, and guide decision-making. Examples are the Plan-Do-Study-Act (PDSA) Cycle or Control Charts.

• Human perception alone is unreliable. Leaders often misinterpret normal fluctuation as anomalies.

1. Common-Cause Variation

• Definition: It is the natural, expected fluctuation built into a stable system. It is predictable within a range and only reduced by improving the system itself.

• Analogy: It is like the slight daily changes in your commute time depending on traffic lights or weather, which it’s expected and within normal limits.

• Use Cases:

  • Aerospace manufacturing: Small, predictable differences in rivet hole diameters (e.g., 5.01 mm vs 4.99 mm) due to machine tolerances.

  • Airline operations: Flight departure times can vary by a few minutes depending on passenger boarding speed, but usually remain within the schedule buffer.

2. Special-Cause Variation

• Definition: It is unusual, unexpected variation caused by specific, identifiable factors. It falls outside the normal process range and signals something different happened.

• Analogy: Like arriving at work two hours late because of a road accident. It’s not normal traffic fluctuation, but an extraordinary event.

• Use Cases:

  • Aerospace manufacturing: A tool misalignment suddenly produces holes of 5.5 mm , whoch is clearly outside the expected tolerance.

  • Airline operations: A flight delayed by 3 hours due to a severe storm or a mechanical failure, which is not part of routine variation.

1. Common-Cause Variation

As it is part of the system design, common-cause variation require systemic improvements, not blaming individuals.

Management actions:

1. Don’t blame individuals for common-cause variation.

2. Instead, improve the system through strategies such as:

Redesign the process/system: Improve equipment, tools, methods, or workflows.

Standardization: Reduce variability by applying best practices consistently.

Training and capability building: Ensure people follow processes consistently.

Preventive maintenance: Keep machines and systems reliable.

Continuous improvement (Kaizen, Lean, Six Sigma): Systematic efforts to shrink variability and move the mean closer to target.

👉 Example: the turnaround times vary between 38–42 minutes. To reduce variation, improve ground crew coordination and standardize checklists.

2. Special-Cause Variation

As it is unpredictable, caused by a specific event or factor, Special-cause variation should be studied for lessons, not just “fixed.”

Management actions:

• Investigate the specific cause.

• Learn from it, and act to prevent recurrence if possible.

Examples of strategies are:

Detect quickly using control charts (spot points outside limits or patterns).

Investigate the root cause.

Correct the specific issue.

Share learning so similar causes don’t recur.

Contain or mitigate impact if the cause is uncontrollable.

👉 Example: one turnaround takes 60 minutes due to unexpected fuel truck breakdown. Action: investigate why the truck failed, fix it, and review preventive maintenance.

3.In short…

Common-cause → Requires improving the system.

Special-cause → Requires investigating and learning from the event.

Understanding variation means leading with awareness rather than reaction. Instead of blaming people for outcomes, leaders look at the system with curiosity, respond with clarity and turn variation into opportunities for learning and growth.

In the next article, I will explore the second element: Appreciation for a System.

See you next week. 👋

Senge, P. M. (2006). The Fifth Discipline: The Art and Practice of the Learning Organization (Rev. and updated ed.). Random House.

Chartered Quality Institute. Applying Deming’s System of Profound Knowledge. Retrieved from https://www.quality.org/knowledge/applying-deming-system-profound-knowledge

School Performance Institute. What Is the System of Profound Knowledge?. Retrieved from https://www.schoolperformanceinstitute.org/blog/2022/6/29/what-is-the-system-of-profound-knowledge

The Deming Institute. (n.d.). Knowledge of variation. Retrieved August 16, 2025, from https://deming.org/knowledge-of-variation/

Disclaimer: The information provided in this newsletter and related resources is intended for informational and educational purposes only. It reflects both researched facts and my personal views. It does not constitute professional advice. Any actions taken based on the content of this newsletter are at the reader's discretion.