Lean Six SigmaOperations Excellence

How to Actually Run a DMAIC Root-Cause Analysis (Without the Buzzwords)

·2 min read

Most DMAIC write-ups read like a certification exam: five tidy letters, a slide per phase, done. In practice, the project lives or dies on two phases: how honestly you define the problem, and how rigorously you prove the root cause before you touch the process.

Define: write the problem statement like a bet, not a complaint

A usable problem statement has a number, a timeframe, and a boundary. "Turnaround time is too slow" is a complaint. "Average MPR preparation time has been 60 minutes against a 20-minute target for the last two quarters, across the Mumbai reporting team" is a bet you can actually lose or win.

Measure: trust the baseline more than your instinct

Before touching U-control charts or Pareto analysis, pull at least four to six weeks of real transaction-level data. On one project this meant pulling 1,500 daily transactions: enough volume that a staffing-rotation pattern, invisible in a week of data, became the dominant signal once charted properly.

Analyze: 5-Whys only works if you keep going after it gets uncomfortable

The failure mode I see most often: teams stop at the third "why" because the answer points at a policy someone in the room owns. Push to the fourth and fifth why anyway. The root cause is rarely a person. It is almost always a rule, a handoff, or an incentive that nobody has revisited since it was written.

Improve: ship a roadmap, not a wish list

A 6-point workflow redesign roadmap beats a 20-point brainstorm every time, because a short list is one people will actually execute. Rank by effort-to-impact, not by how interesting the fix is.

Control: the control plan is the part everyone skips

If there is no owner, no review cadence, and no trigger for re-opening the investigation, the gains decay within two quarters. Write the control plan before you present the results, not after.

DMAIC is not a reporting template. It is a discipline for not fooling yourself about what actually caused the problem. Everything else is paperwork.

Asmi Surya Prakash
Asmi Surya Prakash
MBA, Indian Institute of Management, Kashipur · Project Management Professional (PMP) · Lean Six Sigma Black Belt · Operations Analyst, A.P. Moller - Maersk

Related posts

A dashboard of dimmed metric tiles beside one highlighted on-time performance KPI leading to a decision
Operations Excellence

The KPI Dashboard Nobody Looks At: Designing Metrics Executives Actually Use

Most operations dashboards fail for the same reason: they report activity instead of decisions. Here is the test I use before a metric earns a spot on an executive dashboard.

·2 min read