DMAIC is a five-phase, data-driven method for improving an existing process: Define, Measure, Analyze, Improve and Control. It is the core roadmap of Six Sigma and Lean Six Sigma projects, and it works because each phase must be completed with evidence before the team moves on.
Define: agree the problem and the goal
The team writes a project charter that states the problem, the goal, the scope and the business case. It captures the voice of the customer and draws a high-level SIPOC map so everyone sees where the process starts and ends.
- Tools: project charter, voice of the customer, CTQ tree, SIPOC
Measure: establish the baseline
The team collects data on how the process performs today and checks that the measurement system can be trusted. Without a reliable baseline, nobody can prove an improvement later.
- Tools: data collection plan, process map, measurement system analysis, control charts, process capability
Analyze: find and verify root causes
Instead of jumping to solutions, the team lists possible causes and tests them with data. Only causes confirmed by evidence move forward.
- Tools: fishbone diagram, 5 whys, Pareto chart, hypothesis tests, regression, value stream map
Improve: design and test solutions
Solutions are generated for the verified root causes, risk-assessed and piloted on a small scale before full roll-out.
- Tools: brainstorming, FMEA, quick changeover, error proofing, design of experiments, pilot plans
Control: hold the gains
The new way of working is standardized and monitored so performance does not slip back. The process owner takes over from the project team.
- Tools: control plan, standard work, visual management, control charts, response plan
When should you use DMAIC?
Use DMAIC when an existing process has a measurable problem whose cause is not obvious. If the solution is already known, a Kaizen event or a simple action plan is faster. For a process that does not yet exist, Design for Six Sigma (DMADV) is more suitable.
