Key Takeaways
Run Six Sigma as a customer-led management system, not a statistics campaign
A system, not a tool. Six Sigma combines customer understanding, disciplined use of evidence, and ongoing management of business processes. Its name refers both to a demanding quality target and to a broader way of improving performance. The target is commonly expressed as 3.4 defects per million opportunities, but the book argues that the management habits matter more than the label or the calculations.
Change the decision habit. At GE Capital, a CEO pressed a team to adopt his preferred fix before their analysis was complete. A Black Belt intervened, defending the team's need to understand the process first. The CEO accepted the correction. That small exchange captures the larger ambition: replace confident guessing with informed problem solving, without losing the drive to act.
What's striking is that the book locates Six Sigma's power in organizational behavior, not in statistical machinery. Data become useful when they change how leaders decide, how teams investigate, and how work crosses departments. That makes the approach accessible beyond engineering. It also raises a practical test: if an initiative produces new terminology and reports but leaves decisions untouched, it has adopted the costume without the operating discipline.
Define what customers value before deciding what your process should fix
Listen in more than one way. The Voice of the Customer, or VOC, is a system for gathering and using information about customer needs, expectations, and changing preferences. It should include more than occasional surveys or the loudest complaints. Current customers, unhappy or lost customers, prospective buyers, and competitors' customers can each reveal different parts of the market.
Separate the deliverable from the experience. Output requirements describe what the customer receives, while Service requirements describe how the organization treats the customer along the way. Kano analysis groups needs into basics customers simply expect, performance factors that affect satisfaction, and unexpected features that can delight. These categories shift: once a novelty becomes normal, customers may treat it as a basic requirement.
This framework resists a common managerial shortcut: treating satisfaction as one score or assuming that the most vocal customer represents everyone. Segmenting feedback helps prevent a company from optimizing for an audience it did not intend to serve. The distinction between the product and the experience is also useful because customers judge the whole transaction, not the organization chart. Requirements still need interpretation, though. Listening is valuable only when evidence is translated into specific standards and revisited as expectations change.
Measure variation, because averages can hide the failures customers actually experience
Averages blur the bad days. Variation means inconsistency in a process or its outputs. A delivery process can average 4.2 days and still miss a six-day promise for more than 15 percent of orders. The average looks reassuring; the late deliveries are what customers experience. Six Sigma therefore asks teams to define customer requirements, count failures against them, and examine how consistently the process performs.
Make the denominator honest. Defects per million opportunities, or DPMO, adjusts a defect count for the number of chances a process has to fail. The book presents Six Sigma performance as about 3.4 defects per million opportunities, equivalent to 99.9997 percent perfect under its convention. The measure can compare unlike processes, but only if opportunities and defects are defined consistently.
The emphasis on variation changes the question from 'How good is the average?' to 'How predictable is the experience?' That is a meaningful shift for services, where customers often care as much about reliable timing as about average speed. The caution is that a polished Sigma score can conceal weak assumptions. Teams can inflate opportunity counts or measure what is easy rather than what matters. The metric earns trust only when customer requirements, data collection, and definitions are defensible.
Map the business before launching projects, or improvement targets become guesses
Start with the whole flow. The book's five-step roadmap begins by identifying core processes and customers, then defining requirements, measuring current performance, improving priority processes, and integrating the new practices into daily management. A core process is a connected set of activities that delivers value to external customers. Support processes provide resources that enable this work.
Make the work visible. SIPOC is a high-level map of Suppliers, Inputs, Process, Outputs, and Customers. It shows where work begins, what transforms, and who receives the result. FieldFresh, a long-established food company facing leadership turnover, used process mapping to see how its departments connected. The exercise revealed that knowledge of the entire business sat largely with senior managers, making succession and adaptation risky.
A process map is valuable less as a perfect diagram than as a shared object for asking better questions. It exposes handoffs, hidden dependencies, and assumptions that department-based views can miss. That makes mapping an organizational learning practice as much as an analysis technique. The limitation is that a map can freeze a changing operation into a tidy picture. It should be treated as a testable description of current work, not as an official truth that employees must follow regardless of what happens.
Make teams test causes with evidence before they prescribe fixes
Use DMAIC to resist premature fixes. DMAIC means Define, Measure, Analyze, Improve, and Control. Teams clarify the problem and goal, establish reliable measures, investigate causes, test solutions, and set up ongoing checks. The sequence is iterative, but its central discipline is clear: plausible explanations are hypotheses, not proof.
AutoRec found the cause in the handoffs. In the book's illustrative case, 40 percent of corporate orders failed to meet requirements, and hardware incompatibilities accounted for most of the defects. Data and process mapping showed that a stale parts forecast left the wrong adapters and connectors on hand. Shipping staff, trying to meet deadlines, helped assemble orders and sometimes included incompatible parts. A just-in-time ordering change and shared performance measures helped cut DPMO from 122,800 to 39,000, moving the process from about 2.7 to 3.3 Sigma.
The AutoRec example shows why cross-functional problems rarely have a single guilty department. Each group acted reasonably within its own incentives, yet the combined process produced failures. This is a systems problem: local efforts to meet delivery targets obscured the upstream inventory issue. DMAIC creates a pause for evidence before blame hardens into policy. Its rigor should not become ritual, however. The appropriate amount of analysis depends on the risk of a wrong fix and the cost of delay.
Choose projects that matter to customers, not just to quarterly cost targets
Balance savings with market value. A good project is both meaningful and manageable. Leaders should weigh customer impact, strategic importance, financial benefit, urgency, available expertise, complexity, and likely organizational learning. Too many projects dilute attention; projects framed too broadly can stall. A project should focus on a real gap whose causes and best solution are not already obvious.
Perfecto Pasta saved while losing relevance. The company chose projects to reduce production waste, streamline orders, and speed invoicing, expecting about $2 million in annual savings. But competitors began tailoring product mixes to individual retailers, while Perfecto kept shipping a standard assortment. Its market share fell from 25 percent to 13 percent and then below 10 percent. Internal efficiency did not compensate for failing to respond to customers' changing needs.
Project selection is a strategy decision disguised as a work-planning task. The Perfecto story makes the opportunity cost visible: resources spent on measurable internal savings may crowd out work that protects customer preference and competitive position. Cost reduction remains legitimate, especially when it removes waste without harming service. The important question is whose definition of value guides the portfolio. A balanced set of projects can address operating costs while also improving the reasons customers choose the business.
Search service work, where invisible waste hides and measurement is weak
The office is not exempt. The book argues that Six Sigma applies to administrative, financial, customer-service, logistics, and other service activities, not only factories. It cites estimates that poor quality can consume as much as 50 percent of a service organization's budget, compared with about 10 to 20 percent in manufacturing. Service processes often perform around 1.5 to 3 Sigma before improvement, and less than 10 percent of total cycle time may involve work customers consider important.
Make invisible work observable. Service work often consists of information, approvals, calls, and decisions, so queues and rework are easy to overlook. Teams can begin by mapping the process, defining observable requirements, and collecting simple measures. The authors caution against forcing advanced statistics too early. Basic data and sharper questions may be the right first step, with more technical methods added when the process needs them.
The case for service improvement is especially persuasive because invisible work can normalize delay. A queue of digital requests attracts less attention than a pile of unfinished parts, although both can obstruct customers. Measurement makes that hidden work discussable. Yet the figures should be treated as broad estimates, not as a universal diagnosis of every service business. Some processes are low-volume, relational, or difficult to reduce to counts. In those cases, careful observation and clearly defined examples can complement quantitative measures.
Redesign processes when small repairs cannot meet changed demands
Improvement and redesign solve different problems. Process improvement targets root causes while keeping the basic structure of work intact. Process design or redesign replaces or substantially changes that structure, often because customer needs, technology, or competition have moved beyond what incremental fixes can handle. Improvement is usually the smaller step; redesign carries greater cost, risk, and organizational disruption.
COLA reconsidered its policy process. The insurer's policy completion process averaged 10.4 weeks, despite earlier projects that had shortened it somewhat. A redesign team examined customer needs, process steps, and time spent waiting. It found that only a small share of total time involved value-adding work. The redesigned approach simplified contracts, clarified terms earlier, and assigned coordinators to policies. A pilot averaged 8.5 days, while the team's operating target was a seven-working-day average.
The distinction prevents two opposite errors: trying to repair a structure that no longer fits, and discarding a process that only needs targeted correction. The COLA case also demonstrates that redesign is not synonymous with buying new technology. Changing sequence, decision timing, document complexity, and responsibility can matter just as much. A bold target can focus creativity, but it does not prove feasibility. Pilots, customer feedback, legal constraints, and staff impact are essential checks before a sweeping change becomes permanent.
Make leaders own change, and adapt Six Sigma to local readiness
Leadership must do more than endorse. Executives set priorities, provide resources, review progress, remove obstacles, and remain accountable for results. GE famously tied 40 percent of executive bonus compensation to Six Sigma outcomes. The book also warns that copying another company's rollout is a poor substitute for designing an effort around local needs, capacity, and readiness.
Test the approach before scaling it. A pilot can reveal whether project selection, measurement, training, team roles, or a solution works under ordinary conditions. It should include time to study results and make adjustments. Training should be practical and tied to real work, while roles such as sponsor, coach, team leader, and implementation leader need clear responsibilities. The best first move may be a targeted project, a strategic improvement effort, or a broader transformation.
This advice treats implementation as a managed change process rather than a branding exercise. Piloting gives an organization permission to learn without pretending that the first plan is flawless. Leadership accountability matters because employees notice when executives demand change but keep their own decisions untouched. Still, linking compensation to a metric can create pressure to optimize the number rather than the underlying customer outcome. Incentives work best when paired with transparent measures, qualitative judgment, and room to report setbacks honestly.
Assign process owners and response plans so gains survive the project team
A successful project can still backslide. The Control phase uses documentation, ongoing measures, and response plans to keep an improvement working after a team disbands. A Process Owner is accountable for maintaining process information, monitoring performance, identifying problems, coordinating across groups, and sponsoring further improvement. Measures should balance customer outcomes, process factors, efficiency, and financial results.
Keep feedback connected to action. Up-Home, a retailer, launched a home trial program and appointed an owner to coordinate it across stores. The owner and local coordinators monitored sales, defects, lost items, and customer satisfaction. Their tracking exposed a problem with incomplete customer addresses, allowing a simple correction before it became routine. Process scorecards and clear action thresholds help turn data into timely responses rather than reports that nobody uses.
Sustaining improvement is a governance challenge: someone must own the process after the temporary energy of a team fades. The owner role links measurement to authority and coordination, making it less likely that problems fall between departments. The same structure can become bureaucratic if every process accumulates measures, manuals, and approvals. A useful discipline is to retire indicators that no longer inform decisions and keep responsibility close to the people who understand the work.
Analysis
The book's central contribution is an operating logic for organizational learning. Customer needs define what counts as success, process maps show where value is created or lost, measures distinguish variation from assumption, and improvement methods convert evidence into changes. The sequence is powerful because it links strategy to daily work instead of treating quality as an isolated department or a technical specialty. Stories such as AutoRec and COLA make that logic concrete: failures emerge from interactions among incentives, handoffs, forecasts, rules, and customer expectations, not simply from individual carelessness.
The system's ambition also creates a measurement problem. A target such as 3.4 defects per million opportunities can make quality comparable, but only after people decide what an opportunity and a defect mean. Those definitions are judgments, not neutral facts. If managers reward a score without checking its construction, teams can improve the report while leaving the customer experience unchanged. The book anticipates part of this danger by insisting on valid customer requirements, consistent definitions, and financial validation, yet these safeguards depend on leadership behavior rather than mathematics alone.
A second tension is between standardization and adaptation. DMAIC provides a shared discipline, while the authors repeatedly urge organizations to tailor tools, training, and scale. That flexibility is a strength, particularly for service work and low-volume processes, but it makes implementation harder to audit. The practical answer is not uniformity for its own sake. It is transparency about why a method fits a process, what evidence supports a decision, and what risks remain.
The most durable insight is therefore not that every company should pursue a numerical perfection target. It is that organizations need a repeatable way to hear customers, detect drift, investigate causes, and preserve improvements. Statistical ambition supplies direction; process ownership and candid learning determine whether the ambition becomes better service or merely a more elaborate dashboard.
Review Summary
The Six Sigma Way receives mixed reviews, with an average rating of 3.63 out of 5. Some readers find it informative and comprehensive, praising its practical approach and integration of Six Sigma concepts. Others criticize it for being dry, textbook-like, and lacking in-depth information. Positive reviews highlight the book's usefulness as a reference and training material, while negative reviews express disappointment in its content and presentation. Many readers appreciate the overview of Six Sigma principles but find the material dated or too general for advanced practitioners.
People Also Read
Glossary
Six Sigma
Customer-led performance systemA flexible approach to business performance built around understanding customer needs, using facts and analysis, and managing processes for improvement. The name also describes a demanding quality target, conventionally expressed as 3.4 defects per million opportunities. An organization can use the system without having reached that level across its processes.
DMAIC
Five-phase improvement cycleA structured method for improving or managing a process. The letters stand for Define, Measure, Analyze, Improve, and Control. Teams clarify the problem and goal, establish performance data, investigate causes, implement solutions, and put measures and response practices in place to sustain the gains. The steps can be revisited as new evidence emerges.
Voice of the Customer
System for customer insightThe information a business gathers and uses to understand customer needs, expectations, complaints, preferences, and emerging changes. An effective system draws on varied sources, not only surveys or vocal customers, and translates findings into observable requirements. It should help the organization adjust as customer priorities and markets change.
SIPOC
High-level process mapA diagram that summarizes Suppliers, Inputs, Process, Outputs, and Customers. It gives teams an at-a-glance view of where work and resources come from, what the process does, what it produces, and who receives the result. SIPOC helps clarify process boundaries and relationships before a team investigates details.
DPMO
Defects per million chancesDefects per Million Opportunities is a measure that scales a process's defect count according to the number of defined opportunities for error. It can support comparisons between processes of different complexity. The comparison is meaningful only when teams define defects and opportunities consistently and in relation to customer requirements.
Critical to Quality
Customer-critical requirementA Critical to Quality characteristic, often abbreviated CTQ, is a product, service, or process feature that matters directly to customer satisfaction. It should be specific enough to observe or measure and should connect to a genuine customer requirement. CTQs help translate broad expectations into usable performance standards.
Kano analysis
Customer-needs categorizationA way to group customer requirements by how they affect satisfaction. Basic requirements are expected, performance requirements tend to improve satisfaction as performance improves, and delighters exceed expectations. The categories are not permanent: a feature that once surprised customers may become an ordinary expectation as markets and habits change.
Process Owner
Accountable process stewardThe person responsible for the ongoing performance and coordination of a process, especially one crossing functional boundaries. Responsibilities can include maintaining process information, monitoring measures, identifying issues, coordinating with suppliers and customers, and sponsoring improvement. The role connects temporary project work to the process's daily operation.
Process Management
Ongoing process oversightA management approach that treats connected work as a process to be understood, measured, coordinated, and improved over time. It assigns responsibility for cross-functional performance and links customer requirements to operating measures and response plans. Its purpose is to make process improvement part of running the business, not an occasional campaign.
Cost of Poor Quality
Financial cost of failuresA measure of the money consumed by defects and inefficient work, including rework, scrap, repeated checks, and other costs associated with failing to meet requirements. It can help leaders compare the financial impact of different problems and justify improvement. It should be interpreted alongside customer effects, which may be harder to price.
Download PDF
Download EPUB
.epub digital book format is ideal for reading ebooks on phones, tablets, and e-readers.