EXIN LSSA Lean Six Sigma Yellow Belt study materials
EXIN LSSA Lean Six Sigma Yellow Belt study resources

EXIN LSSA Lean Six Sigma Yellow Belt: The DMAIC improvement cycle

EXIN LSSA Lean Six Sigma Yellow Belt · Work through your exam, one topic at a time.

All guide sections

The DMAIC improvement cycle

DMAIC - Define, Measure, Analyze, Improve, Control - is the structured problem-solving cycle at the centre of Six Sigma. The exam tests what each phase must produce before the next begins, and whether you can spot a team that has skipped a step. In one line: Define frames the problem, Measure establishes a trustworthy baseline, Analyze finds and verifies causes, Improve tests changes, Control locks the gain in.

Phase purpose and exit deliverables

  • Define: charter, problem and goal statements, scope, VOC/CTQ, team and sponsor sign-off.
  • Measure: operational definitions, data collection plan, measurement system analysis, baseline capability or DPMO, detailed process map.
  • Analyze: root causes verified with data, prioritised, with waste identified.
  • Improve: shortlisted and tested solutions, pilot results, implementation plan, updated risk assessment.
  • Control: control plan, standard work, control charts, reaction plan, handover to the process owner, verified benefits.

Gates and iteration

Each phase ends in a tollgate review with the sponsor; failing a gate sends the team back, not forward. The cycle is not strictly linear - a failed measurement system analysis in Measure can force a Define rewrite - but a phase is only finished when its deliverable is accepted.

DMAIC versus alternatives

Use DMAIC when a process exists but underperforms. Use DMADV (Define, Measure, Analyze, Design, Verify), part of Design for Six Sigma, when the product or process does not yet exist. A short, contained problem is often better served by a kaizen burst: Plan-Do-Check-Act maps loosely onto the cycle - Plan covers Define and Analyze, Do is the pilot, Check verifies results, Act standardises.

Worked example

A charter reads: "Install a new CRM because 30% of callback requests are missed." There is no baseline, and the goal names a solution. Diagnosis: Define is contaminated and Measure was skipped. A compliant path: charter the 30% miss rate over Q1 as the problem, agree an operational definition of "missed callback," validate the data source, Pareto the request types, verify the dominant cause, then pilot a scheduling change before any system purchase is considered.

Traps that cost marks

  • Jumping to Improve because someone already has a fix in mind.
  • Skipping Measure, then arguing about causes with no baseline.
  • Assuming Control means freezing the process; it means monitoring and responding.
  • Confusing DMADV with DMAIC - the design variant applies to processes that do not yet exist.