Act as an experienced Lean Six Sigma Green Belt coach. Build a DMAIC project plan for [PROCESS] at [ORGANIZATION OR TEAM]. The business problem is [PROBLEM], the affected customers or stakeholders are [CUSTOMERS], and the available baseline data is [BASELINE DATA]. Constraints, deadline, and available project roles are [CONSTRAINTS AND ROLES]. Do not invent baseline metrics, root causes, savings, defect rates, or control results. Distinguish observed facts from hypotheses that need validation. Keep the scope narrow enough for a Green Belt project: one process, a defined start and end point, measurable CTQs, and an accountable process owner. Produce: 1. A project charter with problem statement, goal statement, scope, business case, CTQs, sponsor, process owner, team, and out-of-scope items. 2. A DMAIC plan in a table with phase, objective, activities, deliverables, owner, dependencies, and week-by-week timing. 3. A measurement plan defining operational definitions, numerator, denominator, data source, collection frequency, sample logic, and likely data-quality risks for each metric. 4. A cause-validation plan using SIPOC, process map, Pareto analysis, 5 Whys, fishbone categories, and tests appropriate to the data. Explain which tools are exploratory versus evidence of causation. 5. An improve-and-control plan with pilot design, success criteria, failure triggers, standard work, control chart or monitoring approach where appropriate, escalation owner, and 30/60/90-day review cadence. Flag where a Black Belt, statistician, safety expert, or process owner must approve the approach. Before answering, check that the goal is measurable from the supplied baseline and that no proposed solution is presented as a confirmed root cause. Ask up to 3 clarifying questions only if a required input is missing.
Fill in
| Placeholder | What to enter | Example |
|---|---|---|
| [PROCESS] | Name the process and its clear start and end points. | New-customer support ticket triage from ticket creation to assignment to the correct specialist queue. |
| [ORGANIZATION OR TEAM] | Enter the business unit or team that owns the process. | Customer Support Operations |
| [PROBLEM] | Describe the measurable performance problem and operational impact. | First-response time for enterprise tickets exceeds the four-hour SLA, causing escalations and delayed onboarding. |
| [CUSTOMERS] | Identify internal or external customers affected and what matters to them. | Enterprise administrators and implementation managers; they need accurate routing and a response within four business hours. |
| [BASELINE DATA] | Provide current volume, defect, time, cost, quality, or service-level data and its source. | April-June: 2,400 enterprise tickets; median assignment time 5.8 hours; 31% miss SLA; categories and timestamps exist in Zendesk. |
| [CONSTRAINTS AND ROLES] | List timing, budget, systems, compliance constraints, sponsor, process owner, and team members. | Twelve-week project; no new software budget; sponsor is VP Support; process owner is Support Ops manager; analyst can export Zendesk data; changes must preserve privacy controls. |
How to use
- Define the process boundary in operational terms, then paste only baseline data you can trace to a system or report.
- Check that the goal uses a metric with a clear numerator, denominator, data source, and measurement period.
- Have the process owner validate the current-state map before treating any cause as real.
- Follow up with: “Create a data-collection sheet and operational definitions for the measurement plan.”
Variations
DMAIC charter
Use this when you need approval before beginning detailed analysis.
Create a Lean Six Sigma DMAIC charter for [PROCESS] with the problem [PROBLEM], baseline [BASELINE], customer CTQs [CTQS], and sponsor and owner [ROLES]. Produce a one-page charter with problem statement, SMART goal, business case, scope, exclusions, milestones, risks, and team responsibilities. Do not calculate savings or improvement targets that the baseline cannot support. Flag missing data needed to approve the charter. Check that the goal includes a baseline, target, deadline, and measurable unit. Ask up to 3 questions only if required inputs are missing.
Root-cause workshop
Use this when a team needs a structured session before choosing corrective actions.
Design a [LENGTH]-minute root-cause analysis workshop for [PROCESS PROBLEM]. Participants are [PARTICIPANTS], known data is [DATA], and the process boundary is [BOUNDARY]. Provide a facilitator agenda, SIPOC prompt, fishbone categories tailored to the process, 5 Whys rules, evidence log, and criteria for selecting causes to validate. Do not let the team label opinions as root causes. Include a post-workshop validation plan with owners and dates. Check that proposed tests can distinguish correlation from a plausible cause. Ask up to 3 questions only if needed.
Control plan
Use this after a pilot or improvement is ready to be sustained.
Create a Lean Six Sigma control plan for [IMPROVED PROCESS]. The implemented change is [CHANGE], critical metrics and current results are [METRICS], process owner is [OWNER], and monitoring tools available are [TOOLS]. Produce a table with CTQ, operational definition, target and limits, data source, sampling frequency, owner, reaction plan, escalation path, and review cadence. Include standard-work update and training checks. Do not claim the change is sustained without post-implementation data. Check that every out-of-limit condition has a named response owner. Ask up to 3 questions only if inputs are missing.
Tips
- Start with a precise operational problem statement, such as a rate, elapsed time, or error count over a defined period, rather than “the process is inefficient.”
- Validate the process map with frontline workers; documented workflows often omit rework loops, handoffs, and exception paths that create most delay.
- Keep 5 Whys and fishbone outputs labeled as hypotheses until data, observation, or a controlled pilot supports them.
- A control plan needs a reaction plan, not just a dashboard; specify who investigates, what threshold triggers action, and how standard work is restored.
FAQ
Can AI complete a Lean Six Sigma Green Belt project for me?
AI can help structure documents, calculations, and questions, but the project requires real process data, stakeholder validation, and accountable ownership.
What makes a goal suitable for DMAIC?
It addresses an existing measurable process problem with a stable enough baseline to improve. New product or process design is often better handled with DMADV or another planning method.
When should I involve a Black Belt or statistician?
Involve one for complex experimental design, advanced statistical inference, high-risk processes, or decisions that could affect safety, compliance, or major investment.