DMAIC Explained: The 5 Phases of Lean Six Sigma

DMAIC is the structured, data-driven method at the heart of Lean Six Sigma. It gives a team a proven way to solve a real business problem: define what is wrong, measure it, find the root cause, fix it, and make the fix stick. The name is an acronym for its five phases, worked in order: Define, Measure, Analyze, Improve, Control.

Used well, DMAIC stops teams from jumping straight to solutions that often don’t actually fix the problem. This guide walks through what DMAIC is, whether it is Lean or Six Sigma, each of the five phases with the mistake teams make most often, when to reach for a different method, and a real project where DMAIC took on-time delivery from 16% to 100%.

What is DMAIC?

DMAIC is a proven, highly effective problem-solving methodology that uses data-driven decision-making to solve organizational challenges. Instead of relying on opinion or gut feel, a team follows five phases in sequence, letting evidence guide each step. It is the default roadmap for improving an existing process, and the backbone of almost every Lean Six Sigma project.

The five phases are Define, Measure, Analyze, Improve and Control. Each has a clear goal and a clear output that feeds the next phase, which is why the order matters. Many of the core Lean Six Sigma tools map directly onto these phases.

Is DMAIC Lean or Six Sigma?

DMAIC comes from the Six Sigma problem-solving methodology. That said, plenty of organizations also use it for classically Lean challenges, such as improving cycle times and removing delays. In practice the phases are the same either way. What changes is the type of problem you point it at: reducing errors and defects leans on Six Sigma, while speeding up a slow process leans on Lean. DMAIC is the shared structure that holds both together.

The 5 phases of DMAIC

Here are the five phases in order, each with its goal and the mistake teams make most often in it.

Define

Define describes the problem through a clear problem statement. The most common mistake is that people jump straight to solutions inside the problem statement itself. When that happens, the “solutions” usually are not solving the actual problem, because the problem was never pinned down.

Measure

Measure quantifies the problem with data, so you know how big it really is and where to focus. The most common mistake is not collecting data, or collecting the wrong data, which leaves the team guessing where the problem actually sits.

Analyze

Analyze determines the root causes of the problem. The most common mistake is skipping this phase altogether. Skip it and the solutions often do not work, because they never eliminated the underlying cause.

Improve

Improve develops the solutions to the root causes found in Analyze. The most common mistake is jumping straight to a solution without understanding the problem or its causes. The result is a fix that does not hold, or a band-aid over the symptom rather than a cure.

Control

Control sustains the improvement through process controls, such as a control chart or an error-proofing method like Poka-Yoke. The most common mistake is skipping this step, after which the process quietly drifts back to what it was before.

When not to use DMAIC

DMAIC is built to improve a process that already exists. You would not use it when the process is so broken that the only real fix is to design a new one, or when there is no process to begin with. In that situation the right method is DMADV (Define, Measure, Analyze, Design, Verify), which designs a new process from scratch around customer needs.

For smaller, faster cycles of improvement, teams sometimes use PDCA (Plan, Do, Check, Act) instead. As a rule of thumb: DMAIC for a data-heavy problem on an existing process, DMADV to design something new, and PDCA for lighter, quick-turn continuous improvement. Lastly, teams might also use the Lean A3 problem-solving method, if the organization feels more comfortable with the Lean approach, which tends to be less data-rigorous than DMAIC.

A real DMAIC example

Define: a higher-education institution struggled with the utilization rate of its internal photography and videography service, which supports the school’s communication, marketing and visibility. Low utilization pushed work out to an outside agency, adding unnecessary cost. Here is how the team worked the phases (the full write-up is in this case study).

Measure: the team tracked how often deadlines were missed and where requests went wrong. Only 16% of internal media requests met their original deadline, mostly because of incomplete information. Analyze: a fishbone diagram traced the delays to inconsistent request channels (phone, email, verbal) and the lack of a standard request form, so even repeat customers left out key details.

Improve: the team introduced a standardized request form inside the existing TOPdesk system, with mandatory fields for shoot date, deadline and visual preferences, plus automated notifications so the media team saw new requests immediately. On-time delivery went from 16% to 100%, and the team handled 12% more projects per month with no extra resources. Control: the form was built on Poka-Yoke (mistake-proofing) principles so it could not be submitted incomplete, backed by regular audits and customer feedback loops. The result was an estimated $331,500 to $582,000 in annual savings from reduced reliance on outside agencies.

Putting DMAIC to work

DMAIC is powerful because it forces discipline: define the problem, prove its size with data, find the real cause, fix that cause, and lock in the gain. The organizations that get the most from it build the skill into their own people rather than outsourcing every project. It also sits alongside the wider Six Sigma principles and starts, like most projects, with a SIPOC to scope the work.

The Lean Six Sigma Company runs in-company group training and full deployment support across the United States. If you want your team to run DMAIC projects that actually stick, those are the right starting points.

Frequently Asked Questions