The Analyze Phase: Separating the Real Driver From the Noise
Lean Six Sigma improvement projects follow DMAIC: Define, Measure, Analyze, Improve, Control. The Analyze phase sits in the middle, and it is where many beginners stumble — not because the tools are hard, but because the temptation to skip ahead is strong. After measuring a problem, everyone has a theory about the cause, and it feels efficient to start fixing. Analyze exists to slow that instinct down and replace opinion with evidence.
The phase has one job: identify and verify the root cause or causes of the problem you defined and measured. The key word is verify. A suspected cause is a hypothesis, not a conclusion. In early 2021, with processes strained by remote handoffs and unreliable supply, it is especially easy to blame whatever broke most visibly — the late shipment, the missed approval — when the real driver is something quieter and upstream.
From symptom to suspected cause
Start by widening the field before you narrow it. A few inexpensive tools, used in order, keep a beginner honest:
Process map the reality. Walk the process as it actually runs, not as the manual describes it. Hidden rework loops, waiting, and duplicate checks often reveal where defects are born.
Brainstorm causes structurally. A fishbone (cause-and-effect) diagram groups possible causes — people, method, machine, material, measurement, environment — so the team considers the whole field rather than its favourite suspect.
Ask why, repeatedly. The 5 Whys drills past the symptom. 'The report was late' → 'because the data arrived late' → 'because the upstream team batches it weekly' — now you're near a real cause.
Verify with data, not conviction
Once you have candidate causes, the discipline of Analyze is to test them against the data you gathered in Measure. This is where Lean Six Sigma earns its reputation. A Pareto chart shows whether a small number of causes account for most of the defects, so you fix the vital few rather than the trivial many. A scatter plot reveals whether a suspected factor actually moves with the outcome. The goal is to distinguish a cause that correlates with the problem from one that merely happens nearby.
State each suspected cause as a testable statement, then look for data that could disprove it.
Beware confusing correlation with causation; two things rising together may share a hidden third cause.
Quantify the effect — knowing a cause drives 60% of defects is far more useful than knowing it 'contributes.'
Carry forward only the causes the data supports; let go of the rest, however popular.
Why the discipline pays off
Teams that skip verification end up in the Improve phase fixing the wrong thing — spending money and goodwill on a change that doesn't move the metric. A rigorous Analyze phase produces a short, evidence-backed list of the causes that genuinely drive the problem, which makes the Improve phase focused and the Control phase defensible. The payoff is simple: you change what matters, and you can show why.
If your organization keeps fixing symptoms while the underlying problem persists, XNM's strategic advisory can help you find and verify the real driver before you spend a dollar on the fix.