DMAIC stands for Define, Measure, Analyze, Improve, and Control. It is the core problem-solving methodology of Six Sigma and the structured framework that every DMAIC project follows from start to finish. Each letter represents one phase. All five phases run in sequence. No phase is optional. The framework guides a team from clearly defining a problem through confirming root causes, implementing solutions, and sustaining the improvement through ongoing monitoring.
DMAIC is not exclusively a Six Sigma tool. Organizations use it for any structured process improvement effort. But it is the methodology that defines how Six Sigma practitioners approach problems, and it is what distinguishes Six Sigma from less rigorous improvement methods that jump directly from problem identification to solution without the analytical rigor in between.
Table of contents
What is DMAIC?
DMAIC is a five-phase, data-driven improvement framework used in Six Sigma to reduce process defects, eliminate variation, and sustain gains over time.
The five phases are: Define (clearly state the problem and project scope), Measure (collect baseline performance data), Analyze (identify and confirm root causes), Improve (implement and test solutions), and Control (monitor to sustain the improvement).
According to ASQ, DMAIC provides a structured, rigorous path from problem identification to confirmed, lasting improvement, distinguishing it from ad hoc change efforts that frequently fail to hold. The framework is most effective when the problem is measurable, has an identifiable root cause, and requires statistical confirmation before any solution is implemented.
Key Takeaways
- DMAIC stands for Define, Measure, Analyze, Improve, and Control. All five phases are required and always run in sequence.
- DMAIC is the core improvement methodology of Six Sigma. It was developed at Motorola in the 1980s and formalized as a systematic framework during Six Sigma’s adoption at General Electric in the 1990s.
- The framework’s defining discipline is that each phase has a gate: the team must complete specific deliverables before moving to the next phase. This prevents teams from skipping root cause analysis and jumping to solutions.
- DMAIC is distinct from DMADV. Use DMAIC when an existing process needs improvement. Use DMADV (Define, Measure, Analyze, Design, Verify) when no adequate process exists and one must be built from scratch.
- Key tools by phase: Define uses project charter, SIPOC, VOC/CTQ; Measure uses data collection plan, MSA/Gauge R&R, process capability; Analyze uses fishbone diagram, Pareto chart, hypothesis testing; Improve uses DOE, piloting, FMEA; Control uses SPC charts, control plan, response plan.
- A DMAIC project typically takes three to six months from Define through Control. Projects with narrowly defined scope and readily available data can run faster. Broader, more complex projects take longer.
- Six Sigma Development Solutions offers Green Belt and Black Belt training that covers the full DMAIC framework, all phase tools, and live project application in onsite, live virtual, and online formats.
Public, Onsite, Virtual, and Online Six Sigma Certification Training!
- We are accredited by the IASSC.
- Live Public Training at 52 Sites.
- Live Virtual Training.
- Onsite Training (at your organization).
- Interactive Online (self-paced) training,
What Is DMAIC?
DMAIC is a structured, data-driven improvement process. It organizes a Six Sigma project into five defined phases, each with a specific purpose, specific tools, and specific deliverables. The structure prevents the two most common improvement failures: acting on assumed root causes rather than confirmed ones, and achieving short-term results that erode when the project team moves on.
The ASQ (American Society for Quality) describes DMAIC as the methodology that “drives Six Sigma projects” and is applicable whenever an existing process needs to be improved. It is the answer to the question every quality improvement team faces: how do we solve this problem in a way that produces confirmed, lasting results instead of temporary fixes?
Every phase answers a specific question:
- Define: What exactly is the problem, and why does it matter?
- Measure: How is the process currently performing?
- Analyze: What is causing the performance gap?
- Improve: What changes will close the gap, and do they work?
- Control: How do we ensure the improvement does not erode?
Also Read: Top Ways SSDSI Simplifies DMAIC for Beginner Students
The Origin of DMAIC
Six Sigma as a statistical quality standard was developed at Motorola in 1986 by engineer Bill Smith and executive Mikel Harry. The goal was to achieve no more than 3.4 defects per million opportunities in manufacturing processes.
General Electric’s CEO Jack Welch adopted Six Sigma as a company-wide initiative in 1995. Under GE’s deployment, DMAIC was formalized as the specific project improvement framework. GE trained thousands of practitioners across its businesses and documented the five-phase structure as the standard approach. The structured phase-gate discipline of DMAIC, where each phase requires specific deliverables before the team advances, became the defining feature that distinguished Six Sigma from earlier quality approaches.
The IASSC (International Association for Six Sigma Certification) now maintains the standard body of knowledge for DMAIC, which forms the basis of Green Belt and Black Belt certification exams globally.
Phase 1: Define
Purpose: Clearly state the business problem, confirm the project is worth solving, and organize the team.
The Define phase does not jump into data collection. It first establishes that the team is solving the right problem in the right way. The most common Define phase failure is producing a problem statement that contains an assumed solution. A problem statement must describe the problem — its nature, location, timing, and magnitude — without naming a cause or prescribing a fix.
Key Define Phase Tools
Project Charter: Documents the problem statement, goal statement, project scope, team members, timeline, and quantified business impact (Cost of Poor Quality). The charter is the project’s governing document. It gets signed by the project sponsor before any data is collected.
SIPOC Diagram: A high-level map showing Suppliers, Inputs, Process steps, Outputs, and Customers. A SIPOC built in the Define phase confirms the process boundaries and prevents scope creep throughout the project.
Voice of the Customer (VOC) and CTQ Tree: VOC captures what customers care about in their own words. The CTQ (Critical-to-Quality) tree translates those words into specific, measurable performance requirements. The project Y — the primary metric — comes from the CTQ.
Define Phase Deliverable
A signed project charter with a specific, quantified problem statement, a measurable goal, defined scope boundaries, and a confirmed business impact calculation.
Phase 2: Measure
Purpose: Collect reliable baseline data on the process being investigated. Establish current performance before any changes are made.
The Measure phase produces two outputs: a validated measurement system and a confirmed baseline capability. The baseline answers the question “how bad is the problem right now?” and gives the team a reference point against which all future improvement will be compared.
The most dangerous Measure phase mistake is skipping measurement system validation. The team collects data with an unvalidated gauge, calculates baseline performance, and proceeds to the Analyze phase — only to discover much later that the data reflected instrument error rather than process performance.
Key Measure Phase Tools
Data Collection Plan: Documents what data to collect, how, from where, by whom, and in what sample size. Includes an operational definition for every measure.
Measurement System Analysis (MSA) / Gauge R&R: Validates that the measurement instruments and operators used in the project produce reliable, repeatable results. A gauge contributing more than 10% of tolerance variation is marginal. More than 30% is unacceptable and must be corrected before data collection proceeds.
Process Capability Analysis: Calculates Cp, Cpk, and sigma level from baseline data. These metrics quantify how well the current process meets the customer’s specification limits. A Cpk below 1.0 means the process is actively producing defects. A Cpk above 1.33 meets most customer minimum requirements.
Run Charts and Control Charts: Plot process data over time to determine whether the process is stable (in statistical control) before capability is calculated. Capability analysis on an unstable process produces unreliable results.
Measure Phase Deliverable
A validated measurement system, confirmed baseline data, and a documented capability index (Cpk and sigma level) representing current process performance.
Phase 3: Analyze
Purpose: Identify, validate, and prioritize the root causes driving the performance gap measured in the Measure phase.
The Analyze phase is where most process improvement programs fail to differentiate themselves from opinion-based approaches. The DMAIC Analyze phase requires statistical confirmation of root causes — not just team brainstorming. A cause that has not been statistically validated is a hypothesis, not a confirmed root cause.
The Analyze phase moves through three steps: generate hypotheses about potential root causes, collect data to test them, and confirm statistically which causes are real drivers of the outcome.
Key Analyze Phase Tools
Fishbone Diagram (Ishikawa / Cause-and-Effect Diagram): Maps potential root causes across six categories: Man, Machine, Method, Material, Measurement, and Environment. Generates the hypothesis list for data-driven testing.
Pareto Chart: Ranks defects or causes by frequency. Identifies the vital few causes that account for the majority of the problem. Based on the Pareto principle: roughly 80% of defects typically come from 20% of causes.
Hypothesis Testing: Statistical tests (two-sample t-test, ANOVA, chi-square, regression) confirm whether suspected root causes actually produce a statistically significant effect on the output Y. A p-value below 0.05 confirms the relationship is real, not the result of sampling variation.
Value Stream Mapping: Maps the complete process flow including value-added and non-value-added steps, wait times, and inventory. Identifies waste that contributes to the performance gap.
Scatter Plot and Regression Analysis: Visualizes and quantifies the relationship between a potential X variable and the project Y. Confirms whether the relationship is linear, strong, and statistically significant.
Analyze Phase Deliverable
A prioritized, statistically confirmed list of root causes (vital few Xs) that drive the project Y metric.
Also Read: DMAIC vs. PDCA: Which is Better?
Phase 4: Improve
Purpose: Design, test, and implement solutions that directly address the confirmed root causes from the Analyze phase.
The Improve phase is where solutions are developed — but only after root causes have been confirmed. Solutions must target the Analyze phase’s validated root causes. Solutions that address symptoms instead of root causes produce temporary improvements that erode.
The Improve phase tests solutions before full deployment. Piloting on a small scale lets the team confirm that the proposed solution actually moves the primary metric in the right direction before committing to organization-wide implementation.
Key Improve Phase Tools
Brainstorming and Solution Selection Matrix: Generates candidate solutions and evaluates them against criteria including impact on root cause, feasibility, cost, and implementation risk.
Design of Experiments (DOE): A statistical method for simultaneously testing multiple process variables to identify the combination that produces the optimal output. Used when the solution involves optimizing process settings with multiple interacting factors.
Piloting with PDCA: A pilot test runs the proposed solution at small scale, collects data, and compares results against the Measure phase baseline. The PDCA (Plan-Do-Check-Act) cycle structures the pilot systematically.
FMEA (Failure Mode and Effects Analysis): Identifies potential failure modes in the proposed solution before full deployment. Risk Priority Numbers (RPNs) prioritize which failure modes require preventive action before the solution goes live.
Error Proofing (Poka-Yoke): Designs the improved process so that errors physically cannot occur, or are automatically detected before they create defects.
Improve Phase Deliverable
A piloted, validated solution that statistically confirms improvement in the primary metric compared to the Measure phase baseline.
Phase 5: Control
Purpose: Sustain the improvement. Prevent the process from reverting to its previous performance once the project team disengages.
The Control phase is where most Six Sigma gains erode when teams do not execute it rigorously. An improved process that is not monitored and defended gradually drifts back toward the original performance level as conditions change, operators are replaced, and the urgency that drove the improvement project fades.
The Control phase makes the improvement permanent through documentation, monitoring, training, and a formal handoff to the process owner.
Key Control Phase Tools
Statistical Process Control (SPC) Charts: Monitor the primary metric over time. Alert the process owner when the metric drifts outside statistically defined control limits. The most common charts are the X-bar and R chart (for continuous data in subgroups) and the individuals chart (for continuous data collected one at a time).
Control Plan: Documents what is monitored, how often, by whom, with what instrument, and what action is taken when the monitored metric goes out of control. A control plan is the difference between a sustained improvement and a temporary one.
Response Plan: Specifies what the process owner does when an SPC chart signals an out-of-control condition. The response plan defines who investigates, what they check, and how quickly they must act.
Standard Operating Procedures (SOPs): Updated documentation of the improved process, ensuring that new operators learn the correct method and that the improvement does not depend on individual knowledge.
Training: Ensures operators and supervisors understand the new process standards and know how to respond to control chart signals.
Control Phase Deliverable
A running SPC chart on the primary metric, a signed control plan, updated SOPs, trained operators, and a formal project close-out confirming that the metric has sustained at the improved level.
DMAIC Phase-by-Phase Summary
| Phase | Question Answered | Key Tools | Primary Deliverable |
| Define | What is the problem and why does it matter? | Project charter, SIPOC, VOC, CTQ tree | Signed project charter with quantified business case |
| Measure | How is the process currently performing? | Data collection plan, Gauge R&R, process capability, run charts | Validated baseline Cpk and sigma level |
| Analyze | What is causing the performance gap? | Fishbone diagram, Pareto chart, hypothesis testing, regression | Statistically confirmed root cause list |
| Improve | What solution works and has been proven? | DOE, piloting, FMEA, error proofing | Piloted solution with confirmed metric improvement |
| Control | How do we keep the improvement? | SPC charts, control plan, SOPs, training | Running control chart and signed control plan |
DMAIC vs. DMADV
DMAIC and DMADV are both Six Sigma frameworks. They serve different purposes. Using the wrong one wastes the project team’s time.
Use DMAIC when: An existing process is not meeting customer requirements. The process exists, produces measurable data, and needs improvement.
Use DMADV when: No adequate process exists and one must be designed from scratch. Or when an existing process has been fully optimized through DMAIC but still cannot meet the required performance level. DMADV replaces Improve and Control with Design and Verify, because the team is designing something new rather than improving what exists.
| Dimension | DMAIC | DMADV |
| Applies when | Existing process needs improvement | New process must be designed |
| Framework | Define, Measure, Analyze, Improve, Control | Define, Measure, Analyze, Design, Verify |
| Starting condition | Process exists and produces data | Process does not exist or is beyond fixing |
| Also called | — | DFSS (Design for Six Sigma) |
How Long Does a DMAIC Project Take?
A focused DMAIC project addressing a specific, well-defined problem in one process typically runs three to six months from Define through Control. This timeline assumes that data is available or collectible within a few weeks and that the team dedicates consistent time to the project each week.
Project duration extends when scope is too broad, when baseline data requires an extended collection period (seasonal processes, low-frequency events), when the root cause investigation reveals complexity requiring additional data cycles, or when the Improve phase pilot requires a long validation period.
Project duration shortens when the team is well-trained, the measurement system is already validated, the data is readily available, and the root cause is relatively isolated.
A common reason beginner projects take longer than expected is scope creep in the Define phase. The most important duration control in DMAIC is a narrowly written, specific problem statement and a clearly bounded SIPOC that prevents the project from expanding as the team learns more.
Frequently Asked Questions: DMAIC
Q: What does DMAIC stand for?
A: DMAIC stands for Define, Measure, Analyze, Improve, and Control. It is the five-phase, data-driven improvement framework used in Six Sigma to solve process problems. Define establishes the problem and project scope. Measure collects baseline performance data. Analyze identifies and confirms root causes. Improve tests and implements solutions. Control monitors the process to sustain the improvement over time.
Q: What is the purpose of DMAIC?
A: DMAIC provides a structured, data-driven path from problem identification to confirmed, sustained improvement. Its primary purpose is to prevent two common improvement failures: acting on assumed root causes rather than confirmed ones, and achieving short-term results that erode when the project team moves on. By requiring statistical confirmation of root causes before any solution is implemented, and by mandating a Control phase with ongoing monitoring, DMAIC produces improvements that hold.
Q: What is the difference between DMAIC and DMADV?
A: DMAIC is used when an existing process needs improvement. It improves what already exists. DMADV (Define, Measure, Analyze, Design, Verify) is used when no adequate process exists and one must be designed from scratch, or when an existing process has been fully optimized but still cannot meet customer requirements. DMADV replaces the Improve and Control phases with Design and Verify, reflecting that the team is creating rather than correcting.
Q: How long does a DMAIC project take?
A: A focused DMAIC project with a specific, well-scoped problem typically takes three to six months from Define through Control. Projects with readily available data and a clearly isolated root cause can close faster. Projects with broader scope, long data collection requirements, or complex root causes take longer. Scope control in the Define phase is the primary way teams keep project duration within the expected range.
Q: What is the most important phase in DMAIC?
A: Every phase is required, but the Analyze phase is where most improvement efforts differentiate between real Six Sigma and superficial problem-solving. The Analyze phase demands statistical confirmation of root causes. This prevents teams from implementing solutions that address symptoms rather than causes. Without a rigorous Analyze phase, the Improve and Control phases are built on assumptions rather than evidence, and improvement gains are frequently temporary.
Q: What tools are used in each DMAIC phase?
A: Define phase tools include the project charter, SIPOC diagram, VOC analysis, and CTQ tree. Measure phase tools include the data collection plan, Gauge R&R, process capability analysis, and run charts. Analyze phase tools include the fishbone diagram, Pareto chart, hypothesis testing, regression analysis, and value stream mapping. Improve phase tools include DOE, piloting with PDCA, FMEA, and error proofing. Control phase tools include SPC charts, the control plan, response plans, and updated SOPs.
Q: Is DMAIC only used in Six Sigma?
A: No. DMAIC is the core framework of Six Sigma, but ASQ confirms it can be used as the improvement framework for any structured problem-solving effort, not exclusively Six Sigma projects. Organizations that have not formally deployed Six Sigma still use DMAIC as a structured improvement approach. The framework’s five-phase structure and its requirement for data-driven root cause confirmation make it applicable to any process improvement context where measurable performance data exists.
DMAIC Training in Six Sigma
DMAIC is the foundational framework for both Green Belt and Black Belt certification. Green Belts learn to lead DMAIC projects as a significant part-time responsibility alongside their primary job. Black Belts lead more complex DMAIC projects full-time and mentor Green Belts through the process.
At Six Sigma Development Solutions Inc, DMAIC is not taught as a theoretical framework. Practitioners work through real project scenarios, apply each phase’s tools to actual process data, and complete toll gate reviews that mirror the experience of presenting a real project to a sponsor.
We offer Six Sigma DMAIC training in three formats:
- Onsite training — delivered at your facility, with your real process as the project context. Best for teams running live DMAIC projects who need hands-on coaching through each phase gate.
- Live virtual training — instructor-led sessions delivered online, covering the full DMAIC curriculum with real-time interaction, group exercises, and Minitab practice.
- Online training — self-paced Green Belt and Black Belt certification programs covering all IASSC-testable DMAIC content at your own pace.
Explore our Six Sigma DMAIC training programs or contact our team to find the right program for your goals.
About Six Sigma Development Solutions, Inc.
Six Sigma Development Solutions, Inc. offers onsite, public, and virtual Lean Six Sigma certification training. We are an Accredited Training Organization by the IASSC (International Association of Six Sigma Certification). We offer Lean Six Sigma Green Belt, Black Belt, and Yellow Belt, as well as LEAN certifications.
Book a Call and Let us know how we can help meet your training needs.
