Overview
Define is where a Green Belt turns a business pain into a bounded, measurable project. The section covers project identification (selection criteria, SIPOC and stakeholders), the voice of the customer (identifying customers, collecting data, translating needs into critical-to-quality requirements with tools such as the Kano model and QFD), project management basics (charter, scope, goals, communication plan, schedule and phase reviews), the seven management and planning tools, business results metrics such as DPU, DPMO, RTY and cost of poor quality, and team dynamics.
Candidates must be able to write and critique a charter, compute the core defect and yield metrics by hand, and know which planning tool fits a described situation. Define connects backward to Domain 1's organizational goals and forward to Measure, where the CTQs and process boundaries chosen here decide what gets measured.
How it's tested
Define questions mix short calculations with charter and team judgment calls. Expect to compute DPU, DPMO, RTY or cost of poor quality from a few numbers, to choose the correct tool for organizing ideas or planning, and to spot the flaw in a problem statement or scope. BEST and FIRST wording is common in team and stakeholder items.
-
Compute the business metric
A process with units, opportunities and defects, or several step yields, is described and you calculate DPU, DPMO, RTY or a sigma level from the numbers given.
-
Critique the charter
A draft problem statement, goal or scope is shown and you identify what is wrong, such as an embedded solution, an unmeasurable goal, or a scope that covers processes the team cannot influence.
-
Match the tool or team stage
A situation such as grouping many brainstormed ideas, sequencing tasks, or a team arguing over roles asks which planning tool fits or which team development stage is occurring and what the leader should do.
A stem says: 'Problem statement: customer complaints are rising because we lack an automated order system.' Asked what is wrong, test it against the rules: a problem statement describes the gap, where and when it occurs, and its size, without naming a cause or a solution. This one asserts a cause and a solution and gives no measure. The best answer rewrites it with the complaint rate, time period and impact only.
Topics 6
Official topics from the ASQ outline, in outline order.
2.1 Project identificationSelecting projects, defining process elements with SIPOC, benchmarking and identifying owners and stakeholders.
- What it means
- Project identification picks problems worth solving and frames the process around them. Selection weighs benefit, alignment, data availability and feasibility; SIPOC (suppliers, inputs, process, outputs, customers) sets the high-level boundaries; benchmarking compares performance with internal, competitive or best-in-class references; stakeholder analysis identifies who is affected and who owns the process.
- What to understand
- A good Green Belt project has a chronic, measurable problem with an unknown root cause, a scope achievable in a few months, and a process owner who will sustain the result. SIPOC is drawn at about five to seven high-level steps and is used to agree on start and end points before detailed mapping.
- How it shows up
- A scenario asks what tool to use first to clarify project boundaries with the sponsor, expecting SIPOC, or which candidate project is suitable, expecting the one with a measurable problem and unknown cause.
- Common confusion
- Candidates treat 'just do it' fixes with known solutions as good Six Sigma projects; if the cause and solution are already known, a DMAIC project adds little.
2.2 Voice of the customerIdentifying customers, collecting their needs and converting them into measurable requirements.
- What it means
- Voice of the customer (VOC) is the stated and unstated needs of internal and external customers, gathered through surveys, interviews, focus groups, complaint data and observation, then translated into critical-to-quality (CTQ) characteristics with measurable specifications.
- What to understand
- Use a CTQ tree to go from a vague need ('fast service') to a driver and then to a measurable requirement ('order confirmed within 2 hours'). Know the Kano categories: basic (must-be) needs cause dissatisfaction when absent but no delight when present; performance needs scale with fulfillment; delighters excite when present but are not missed when absent. Quality function deployment (QFD) relates customer needs to technical requirements in the house of quality.
- How it shows up
- A stem describes a feature customers never mention in surveys but complain loudly about when it fails, and you classify it as a basic Kano need, or you select the step that turns a verbatim comment into a CTQ.
- Common confusion
- Customers rarely mention basic needs because they assume them, so candidates who rely on survey frequency underrate the most important requirements.
2.3 Project management (PM) basicsCharter, scope, goals, communication plan, schedule, tools and phase reviews.
- What it means
- The project charter documents the business case, problem statement, goal, scope, team, and timeline and is signed by the sponsor. Supporting PM elements include a communication plan, a work breakdown structure and Gantt chart, risk tracking and tollgate (phase) reviews at the end of each DMAIC phase.
- What to understand
- Problem statements state what, where, when and how big, with no cause or solution. Goals should be SMART (specific, measurable, achievable, relevant, time-bound). Scope lists what is in and out, and scope creep is controlled by returning to the sponsor. Phase reviews confirm deliverables before the team moves on.
- How it shows up
- Questions show a charter element and ask what is missing or wrong, or describe a stakeholder requesting extra work mid-project and ask the BEST response, which is to evaluate it against the charter with the sponsor.
- Common confusion
- Candidates think a broad scope shows ambition; on the exam an oversized scope is a classic reason projects fail, and narrowing it is the correct move.
2.4 Management and planning toolsAffinity diagrams, interrelationship digraphs, tree diagrams, prioritization matrices, matrix diagrams, PDPC and activity network diagrams.
- What it means
- These seven tools organize qualitative information and plans. Affinity diagrams group many ideas into themes; interrelationship digraphs show cause-and-effect links between issues to find key drivers; tree diagrams break a goal into actions; prioritization matrices weight options against criteria; matrix diagrams show relationships between two or more sets of items; the process decision program chart (PDPC) anticipates what can go wrong in a plan; activity network diagrams show task sequence and the critical path.
- What to understand
- Choose by purpose: organizing ideas (affinity), finding root drivers among issues (interrelationship digraph, where the item with the most outgoing arrows is a key driver), detailing a plan (tree), choosing among options (prioritization matrix), contingency planning (PDPC) and scheduling (activity network).
- How it shows up
- A stem describes the team's need, such as 'sort 80 brainstormed comments into themes' or 'identify what could go wrong in the rollout plan', and asks which tool fits.
- Common confusion
- In an interrelationship digraph, the item with the most outgoing arrows is a driver (cause) and the one with the most incoming arrows is an outcome; candidates swap them.
2.5 Business results for projectsProcess performance metrics such as DPU, DPMO, RTY, sigma level and cost of poor quality, and communicating results.
- What it means
- Business results express how a process performs and what a project is worth: defects per unit (DPU), defects per million opportunities (DPMO), first-pass yield, rolled throughput yield (RTY), process sigma, and cost of poor quality (COPQ) split into prevention, appraisal, internal failure and external failure costs.
- What to understand
- DPU = defects / units. DPMO = defects / (units x opportunities per unit) x 1,000,000. RTY is the product of each step's first-pass yield, and RTY is approximately e^(-DPU) for a process. Sigma level is found from the long-term defect rate with the conventional 1.5 sigma shift added. COPQ is mostly hidden internal and external failure cost.
- How it shows up
- Expect calculations: a three-step process with yields of 95%, 90% and 98% has RTY of 0.95 x 0.90 x 0.98 = 0.838, or a batch with known defects and opportunities requires DPMO.
- Common confusion
- Final yield after inspection and rework hides the hidden factory; RTY counts only units that pass every step first time, so it is always lower than or equal to final yield.
2.6 Team dynamics and performanceTeam stages, roles and responsibilities, and team tools for generating and narrowing ideas.
- What it means
- Teams move through forming, storming, norming, performing and adjourning (Tuckman). Six Sigma roles include the champion or sponsor (owns the business case, removes barriers), process owner (sustains the change), Black Belt (leads complex projects and coaches), Green Belt (leads smaller projects or supports Black Belts) and team members. Team tools include brainstorming, nominal group technique and multivoting.
- What to understand
- Match leader behavior to stage: clarify purpose in forming, surface and resolve conflict in storming, reinforce agreed norms in norming, delegate in performing. Nominal group technique gives every member equal voice through silent idea generation and ranking, which counters dominant personalities; multivoting narrows a long list quickly.
- How it shows up
- A scenario describes conflict over roles and approach a few weeks in and asks what stage it is (storming) and what the facilitator should do, or a quiet team dominated by one manager and asks which tool helps (nominal group technique).
- Common confusion
- Candidates treat storming as a failure to be stopped; it is a normal stage, and the right answer usually helps the team work through the conflict rather than replacing members.
High-yield concepts
The terms, rules and formulas this domain turns into questions most often.
-
Defects per million opportunities (DPMO)
DPMO = total defects / (units x opportunities per unit) x 1,000,000, which normalizes defect rates across products of different complexity.
- Remember
- Opportunities must be defined consistently and should count only meaningful, customer-relevant ways a unit can fail.
- Tested as
- One of the most frequently calculated Define metrics; the stem gives units, opportunities and defects.
- Trap
- Dividing by units only, which gives DPU, or forgetting to multiply by one million.
-
Defects per unit (DPU)
DPU = total defects / total units, the average number of defects on each unit regardless of how many opportunities it has.
- Remember
- DPU can exceed 1 because one unit can carry several defects; it is not the same as the fraction of defective units.
- Tested as
- Used directly and as an input to the RTY approximation e^(-DPU).
- Trap
- Treating defects and defectives as interchangeable; five defects on one invoice are one defective unit.
-
Rolled throughput yield (RTY)
RTY is the probability that a unit passes through every process step defect-free the first time, calculated by multiplying the first-pass yields of all steps.
- Remember
- Multiply step yields, never average them: 0.95 x 0.90 x 0.98 = 0.838, so only about 84% of units are right first time.
- Tested as
- Multistep yield calculations and questions about the hidden factory of rework.
- Trap
- Reporting final yield after rework as if it were RTY, which hides the cost of fixing defects inside the process.
-
First-pass yield (FPY)
FPY for a single step is units passing that step without rework or scrap divided by units entering the step.
- Remember
- FPY is per step; RTY combines all steps' FPYs.
- Tested as
- Stems give counts entering, scrapped and reworked at a step and ask for its yield before you combine steps.
- Trap
- Counting reworked units that eventually passed as first-pass good units.
-
Sigma level
Process sigma converts a long-term defect rate into a Z value; by convention the reported short-term sigma level is the long-term Z plus a 1.5 sigma shift.
- Remember
- 3.4 DPMO corresponds to 6 sigma, about 6,210 DPMO to 4 sigma and about 66,807 DPMO to 3 sigma under the 1.5 shift convention.
- Tested as
- Questions convert DPMO to an approximate sigma level or compare two processes, usually with a table supplied.
- Trap
- Forgetting whether the table or stem already includes the 1.5 shift and adding it twice.
-
Cost of poor quality (COPQ)
Quality costs fall into prevention (training, mistake-proofing), appraisal (inspection, audits), internal failure (scrap, rework before delivery) and external failure (warranty, returns, complaints after delivery). COPQ usually refers to the failure costs, sometimes with appraisal.
- Remember
- Spending on prevention typically reduces failure costs far more than it costs, and external failure is usually the most expensive category.
- Tested as
- Classification questions ask which category a cost belongs to, and business case questions use COPQ to justify a project.
- Trap
- Classifying inspection as prevention; inspection is appraisal because it finds defects rather than stopping them.
-
Kano model
Customer requirements fall into basic (must-be), performance (one-dimensional) and excitement (delighter) categories, with indifferent and reverse categories also possible.
- Remember
- Delighters become basic needs over time as customers come to expect them.
- Tested as
- A described feature and customer reaction must be classified, or a team must decide which needs to protect first.
- Trap
- Investing in delighters while a basic need is failing; meeting must-be needs comes first.
-
Critical to quality (CTQ) tree
A tree that breaks a broad customer need into quality drivers and then into measurable CTQ requirements with targets and specification limits.
- Remember
- A CTQ is only finished when it is measurable and has an operational definition.
- Tested as
- Questions ask which item in a list is a properly formed CTQ, or what step converts a verbatim comment into a requirement.
- Trap
- Accepting a vague phrase such as 'friendly service' as a CTQ.
-
SIPOC
Suppliers, inputs, process, outputs, customers: a high-level map of about five to seven steps that sets project boundaries.
- Remember
- It is built early in Define to agree on where the process starts and stops; detailed flowcharts come in Measure.
- Tested as
- Tool-choice questions about clarifying scope with stakeholders.
- Trap
- Using a detailed swim-lane flowchart in Define when the team has not yet agreed on boundaries.
-
SMART goals and problem statements
Goals should be specific, measurable, achievable, relevant and time-bound; problem statements describe the gap, its location, timing and magnitude.
- Remember
- Neither the problem statement nor the goal should name the root cause or the solution.
- Tested as
- Charter critique questions.
- Trap
- Picking a problem statement that sounds decisive because it names the fix.
-
Tuckman team stages
Forming, storming, norming, performing and adjourning describe how a team develops from polite uncertainty through conflict to productive work and closure.
- Remember
- Leaders direct in forming, coach through storming, support in norming and delegate in performing.
- Tested as
- A described team behavior must be matched to the stage and the leader's best response.
- Trap
- Assuming a team that disagrees is dysfunctional and needs new members.
Side by side
Concepts candidates mix up, compared on the point that decides the answer.
DPU vs. DPMO vs. RTY
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| DPU | Defects divided by units | Ignores how many opportunities each unit has | Reading DPU as the percentage of defective units |
| DPMO | Defects divided by total opportunities, times one million | Normalizes for complexity so different processes can be compared | Inflating opportunities to make the process look better |
| RTY | Product of first-pass yields across all steps | Measures right-first-time flow through the whole process | Averaging step yields or using final yield after rework |
Prevention vs. appraisal vs. internal failure vs. external failure costs
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| Prevention | Costs to stop defects happening, such as training, mistake-proofing, design reviews | Spent before defects exist | Cutting it to save money, which raises failure costs |
| Appraisal | Costs to find defects, such as inspection, testing, audits | Detects but does not prevent | Classifying inspection as prevention |
| Internal failure | Costs of defects found before the customer receives the product, such as scrap and rework | Caught inside the organization | Ignoring rework time because no material is scrapped |
| External failure | Costs of defects found by the customer, such as warranty, returns, lost business | Usually the most expensive category | Counting only direct warranty cost and missing lost customers |
Affinity diagram vs. interrelationship digraph vs. tree diagram
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| Affinity diagram | Groups many ideas into natural themes | Organizes; shows no cause and effect | Using it to find the root driver |
| Interrelationship digraph | Arrows show which issues influence others | Most outgoing arrows marks a key driver | Choosing the item with most incoming arrows as the root cause |
| Tree diagram | Breaks a goal into progressively detailed actions | Moves from general to specific in a hierarchy | Using it to sequence tasks over time, which is the activity network diagram's job |
Exam traps
Each trap is an answer that looks right. Learn the clue that should change your mind.
-
DPU passed off as DPMO
- What candidates pick
- Defects divided by units, times one million.
- Why it looks right
- It uses the right multiplier and looks like a DPMO-sized number.
- Why it's wrong
- DPMO divides by total opportunities (units x opportunities per unit); skipping that step overstates the defect rate by the number of opportunities.
Clue to watch for The stem states a number of opportunities per unit, which must appear in the denominator.
-
Average of step yields
- What candidates pick
- The arithmetic mean of first-pass yields as the process yield.
- Why it looks right
- Averaging feels like a fair summary of several steps.
- Why it's wrong
- A unit must pass every step, so the probabilities multiply; 95%, 90% and 98% give an RTY of about 83.8%, not 94.3%.
Clue to watch for The question asks for the chance a unit goes through the whole process defect-free.
-
Solution in the problem statement
- What candidates pick
- A problem statement such as 'complaints are high because we need a new ERP system'.
- Why it looks right
- It sounds decisive and action-oriented.
- Why it's wrong
- Naming a cause or solution in Define skips Measure and Analyze and biases the team; the statement should describe what, where, when and how much.
Clue to watch for Words such as because, due to, lack of or need to inside a problem statement.
-
Accepting scope creep
- What candidates pick
- Adding a stakeholder's extra request to keep them happy.
- Why it looks right
- It looks collaborative and responsive.
- Why it's wrong
- Uncontrolled scope growth is a leading cause of failed projects; changes should be assessed against the charter and approved or deferred with the sponsor.
Clue to watch for A request arrives mid-project that is outside the charter's in-scope list.
-
Most incoming arrows as driver
- What candidates pick
- The issue with the most incoming arrows on an interrelationship digraph as the root driver.
- Why it looks right
- It looks like the most connected and important box.
- Why it's wrong
- Incoming arrows mean the item is affected by many others, so it is an outcome; the item with the most outgoing arrows drives the system.
Clue to watch for Arrow counts are given as in and out for each item.
-
Replacing members during storming
- What candidates pick
- Removing or reassigning the members who disagree.
- Why it looks right
- It ends the conflict fast.
- Why it's wrong
- Storming is a normal stage; the leader should clarify roles, ground rules and decision methods so the team can move to norming.
Clue to watch for The team is a few weeks old and arguing over approach or roles, not over ethics or misconduct.
-
Inspection as prevention cost
- What candidates pick
- Classifying incoming inspection or final testing as a prevention cost.
- Why it looks right
- Inspection prevents bad product reaching the customer.
- Why it's wrong
- Inspection detects defects that already exist, so it is an appraisal cost; prevention costs stop defects from being created.
Clue to watch for The activity checks, tests or audits product or process output.
Answer framework
For Define phase questions, whether about a charter, a metric or a team, work through these steps.
- Identify whether the stem is asking for a calculation, a charter judgment, a tool choice or a team response before reading the options.
- For metrics, write the formula first: DPU is defects over units, DPMO divides by units times opportunities, and RTY multiplies step yields.
- For charters, check that the problem statement has no cause or solution, the goal is SMART, and the scope is something the team controls.
- For tools, match the purpose: organize ideas, find drivers, break down a plan, prioritize, anticipate failures or schedule.
- For teams, name the stage or role, then choose the response that keeps the sponsor engaged and helps the team progress rather than bypassing it.
Practice question
Want more like this? The CSSGB practice test has questions for every domain.
Checklist
0 of 13
Tick each one you can do without notes. Your ticks stay in this browser.
Study strategy
Begin with business results for projects, because the DPU, DPMO, RTY and cost of quality calculations are the most reliable points in the section and they reappear in Measure. Work at least twenty hand calculations until you instinctively write the formula before touching the numbers. Then study the charter and project management basics, practicing on badly written problem statements until you can spot an embedded solution at a glance.
Next cover voice of the customer, with emphasis on CTQ trees and the Kano categories, and the seven management and planning tools. Make a one-line 'use it when' card for each tool, because tool-choice questions are fast points when you know the purpose and slow guesses when you do not.
Candidates commonly underestimate team dynamics and stakeholder questions, which look like common sense but have a specific Six Sigma answer: work through the sponsor, let the team progress through its stages, and use structured tools to equalize voices. Finish by taking a timed Define set, then move to Measure, where the CTQs and SIPOC boundaries you set here determine what data you collect.
FAQ
How many questions does the Define phase carry on the CSSGB?
ASQ lists 23 of the 100 scored questions for this section, tying it with Measure as the largest section of the exam.
What calculations should I master for Define?
DPU, DPMO, first-pass yield, rolled throughput yield, the RTY approximation e^(-DPU), converting DPMO to a sigma level with a table, and classifying cost of quality.
What is the difference between a defect and a defective?
A defect is any single nonconformance; a defective is a unit with one or more defects. One invoice with three wrong fields counts as three defects but one defective.
Why does a problem statement exclude causes and solutions?
Because Define comes before data analysis; naming a cause or fix biases the team and can lead to solving the wrong problem. The statement should describe the gap, where and when it occurs, and its size.
Which management and planning tool finds the key driver among related issues?
The interrelationship digraph. The item with the most outgoing arrows is a likely driver, while the one with the most incoming arrows is a likely outcome.
How long should I study the Define phase?
Plan roughly 25–30 hours within a 120-hour plan, split between metric calculations and the charter, VOC and team topics.
How does Define connect to Measure?
The CTQs, operational definitions and SIPOC boundaries set in Define decide what data Measure collects and where; the baseline DPMO or yield calculated in Define becomes the starting point for capability analysis.
Other Quality Management certifications
Where the same subject shows up in other exams. Cards marked "Same subject" open the matching domain.
Test your Domain 2 knowledge
You have covered chartering, VOC, the planning tools and the DPU, DPMO and RTY calculations. Check whether you can do them accurately against the clock.