DMAIC is the core Six Sigma methodology for improving an existing process, structured as five sequential phases.
The idea
Each letter is a phase: Define the problem and goals; Measure current performance with real data; Analyze the data to find root causes of variation or defects; Improve the process by testing and implementing fixes; Control the improved process so gains stick, typically with monitoring and standard work. The cycle assumes the process already exists — unlike its cousin DMADV/DFSS, which designs a new one from scratch.
When to use it
- An existing process has a measurable quality, cost, or speed problem.
- Data can realistically be collected on the process.
- The organization needs a repeatable, disciplined structure rather than an ad-hoc fix.
How to apply it
- Define: write a problem statement, scope, and goal.
- Measure: map the process and collect baseline data.
- Analyze: use root-cause techniques to isolate the real drivers of the problem.
- Improve: generate, test, and pilot changes before rolling them out.
- Control: put monitoring, documentation, and ownership in place so the improvement holds.
Watch out for
Skipping straight to “Improve” without solid Measure and Analyze data is the most common failure — it produces fixes for the wrong problem. The Control phase is also often neglected, letting the process drift back to its old state.
Related models
- Dflss — the related Design for Lean Six Sigma approach for new processes.
- Root Cause Analysis — the technique underlying the Analyze phase.
- Kaizen — a complementary philosophy of continuous, incremental improvement.
Sources
Six Sigma methodology, originally developed at Motorola and popularized by General Electric.