Profound Knowledge Pt. 2: Understanding Variation.

A Better Way to Read Performance

Variation is universal. Literally. It is the way that nature behaves all around us.

Many organizations measure performance using simple metrics, clear numbers, one-time snapshots, or summary charts. But without tracking how those numbers behave over time, we miss the full picture and can be misled.

Consider this example: four teams (A through D), four periods of performance. At first glance, it looks like some people are doing well, while others are not.

The temptation? React. Fix it. Push someone harder.

But now look at the same data shown over time — with a line chart and a simple average added.

Suddenly, a new story appears: all the teams are fluctuating around the same average. The system is stable — the ups and downs are expected. This is common cause variation, the natural rhythm of a process.

Here’s the insight:

• Don’t overreact to normal variation.

• Don’t punish a stable system.

• Change the process if the outcomes aren’t acceptable.

By contrast, when something unexpected happens — a sudden spike, a drop, a change in pattern — we’re seeing special cause variation. That’s your signal to investigate, isolate, and correct the anomaly.

Think of a thermostat that fluctuates 1–2 degrees. That’s common cause. But if it suddenly jumps 10 degrees? That’s a special cause worth investigating.

Understanding the difference is the key to successful data analysis and decision making.

What Variation Should You Focus On?

While every organization is different, one principle applies universally: focus on the process.

All processes have inputs, at least one value added activity, and an output. These three elements will each have variation. Where are the pain points? That would be the best starting point.

One common pain point is rework — repeating a process step because it wasn’t done right the first time. Rework eats time, money, and morale.

To fix it, you need to understand the variation behind it. Start by defining what “normal” looks like for the process — then use that baseline to detect causes and reduce unnecessary variation.

Tip: To correct the outputs of a process, examine the inputs.

Every output is the result of inputs + process behavior. If you’re getting inconsistent or unsatisfactory results, the first place to look is upstream — the inputs. So, we look to “input variation” for some guidance. Some key inputs are likely:

•     Supply Chain - inconsistent materials or delivery delays

•     Procedures and Work Instructions - unclear or outdated documentation

•     Training - gaps in skill, knowledge, or expectations

•     Workspace environment - temperature, lighting, cleanliness, layout

•     Tools and equipment - calibration, wear, usability

Of course, not all variation impacts performance. The key is to identify which variations are meaningful — the ones that affect cost, quality, or delivery. That’s where your focus should be.

Tip: The Fishbone Diagram

If you’re having trouble identifying your variables. Try using a Fishbone Diagram (Ishikawa) — typically used for root cause analysis, it’s also great for identifying sources of input variation. Simply list your potential variables under each category. It may just reveal what you’ve been overlooking.

Why is Understanding Variation So Important?

When we are beginning to understand variation, what do we do with it?

At the very least we can stop being reactive. Understanding what the process is doing can prevent us from overreacting to a “blip” and help us identify common cause variation from special cause influence.

As we advance, we begin to develop “control” of variation. This leads to defect and rework prevention (as opposed to reaction), and will have a positive financial impact.

As Dr. Deming famously put it: 

“Inspection does not improve quality, nor guarantee quality. Inspection is too late. The quality, good or bad, is already in the product. Quality cannot be inspected into a product or service. It must be built into it.”

The message is clear: prevention beats detection. And the path to prevention begins with understanding variation.