DMAIC — Define, Measure, Analyse, Improve, Control

DMAIC is the structured problem-solving roadmap at the heart of Six Sigma. Each phase has specific tools and deliverables: Define (problem statement, charter), Measure (baseline data), Analyse (root cause), Improve (implement solution), Control (sustain the improvement).

One-page infographic

Visual summary of DMAIC — Define, Measure, Analyse, Improve, Control · Free PDF

Download PDF

Why companies use it

·Prevents the common failure mode of implementing solutions before understanding root causes
·Creates a common language for improvement projects across the organisation
·The structured gate reviews at each phase ensure leadership oversight and resource prioritisation
·The Control phase explicitly addresses sustainability — the most common failure point of improvement projects

What hiring managers look for

·DMAIC is the language of structured improvement — engineers who know it can lead projects without constant supervision
·Being able to create a clear project charter (scope, goal, timeline, business case) is a key project management skill
·Understanding which statistical tools belong in each DMAIC phase shows practical knowledge, not just theoretical familiarity
·The ability to present a DMAIC tollgate review to senior management is valued in improvement engineering roles

Typical interview questions

What are the five phases of DMAIC and what is the key deliverable of each?

What tool would you use in the Measure phase to understand where in the process defects are occurring?

How do you transition from the Analyse phase to the Improve phase — what evidence do you need?

What does a good Control Plan include, and why is the Control phase often the weakest?

Have you used DMAIC on a real project? What phase was most challenging and why?

Common mistakes

·Rushing through Define — a poorly scoped project with a vague problem statement will fail regardless of the quality of analysis
·Measuring the wrong metric in Measure — the Y variable must be directly linked to the customer problem and measurable with an acceptable MSA
·Jumping to the Improve phase with a solution already decided — DMAIC is meant to find solutions, not confirm pre-existing beliefs
·Treating the Control phase as optional — without a control plan, the process reverts to the old state within 90 days
·Using DMAIC for every problem — simple issues should be handled with 5 Whys or kaizen, not a full DMAIC project

Real engineering example

An automotive stamping plant used DMAIC to reduce door panel dimensional variation. Define: customer was returning 1.2% of panels for fit issues. Measure: CMM data showed variation was concentrated at one specific feature. Analyse: fishbone and regression analysis identified die wear rate as the primary X. Improve: reduced die maintenance interval from 100k to 65k hits. Control: added die wear measurement to the PM checklist. Defect rate dropped to 0.08% within 12 weeks.

Topics covered

DMAICSix Sigmaproblem solvingimprovementqualityproject

Related interview guides

Manufacturing Engineer →Continuous Improvement Engineer →