Six Sigma didn’t just revolutionize quality in the late 1980s, it delivered real-world results that would set the stage for Design for Six Sigma, with breakthroughs like Motorola’s estimated $16 billion in savings and GE’s $12 billion in returns, evolving over time into DFSS practices like CTQs, DMADV, QFD, and robust design methods that help teams prevent defects before they ever reach production.
Key Takeaways
- The original Six Sigma methodology was developed at Motorola in 1986
- Motorola reported that Six Sigma reduced defects by a factor of 10 between 1987 and 1992
- Motorola’s Six Sigma program saved an estimated $16 billion from 1986 to 2000
- Design for Six Sigma (DFSS) uses the concept of “critical-to-quality” (CTQ) characteristics
- DMADV is Define, Measure, Analyze, Design, Verify
- Quality Function Deployment (QFD) is one of the primary DFSS tools
- Six Sigma aims to reduce defects to 3.4 per million opportunities at a 1.5 sigma shift
- The 1.5 sigma shift corresponds to a 4.5 sigma centered process yielding 3.4 DPMO after accounting for long-term drift
- Six Sigma level performance corresponds to 99.379% yield
- Six Sigma projects use financial metrics such as cost of poor quality and realized savings; GE’s Six Sigma yielded $12 billion savings (1995–2001)
- Motorola estimated $16 billion in savings from 1986 to 2000
- AlliedSignal reported more than $600 million in benefits in 1997 from Six Sigma
From Motorola to GE and DFSS, customer CTQs drive measurable billions in savings.
History & Adoption
1The original Six Sigma methodology was developed at Motorola in 1986[1]
Verified2Motorola reported that Six Sigma reduced defects by a factor of 10 between 1987 and 1992[2]
Verified3Motorola’s Six Sigma program saved an estimated $16 billion from 1986 to 2000[3]
Verified4General Electric estimated that Six Sigma delivered more than $300 million in savings in its first year (1996)[4]
Directional5Jack Welch described Six Sigma as a top priority at GE, and GE targeted $10 billion in savings from Six Sigma by 2000[5]
Single source6GE reported that Six Sigma saved over $12 billion from 1995 to 2001[6]
Verified7AlliedSignal (later Honeywell) reported savings of more than $600 million in 1997 from Six Sigma[7]
Verified8Honeywell reported that its Six Sigma program generated cumulative savings of $700 million in 2000[8]
Verified9Ford launched Six Sigma in 1999 as part of its quality transformation[9]
Directional10Toyota adopted Six Sigma methodologies as part of its business excellence approach; Toyota Motor Corporation implemented Six Sigma starting around the early 2000s[10]
Single source11In the 1990s, Six Sigma spread from manufacturing to services and transactions[11]
Verified12A survey by ASQ reported that 67% of organizations used Six Sigma in some form[12]
Verified13ASQ’s 2016 “State of Six Sigma” survey found 68% of organizations used Six Sigma[13]
Verified14ASQ’s 2017 “Six Sigma and Lean Manufacturing” report found 56% of organizations used Six Sigma[14]
Directional15ASQ’s “Six Sigma in Services” article reported that over 50% of Six Sigma projects were in service functions[15]
Single source16A 2004 ASQ report “Six Sigma: A Critical Review” noted that the number of Six Sigma-related certifications rose rapidly during the 1990s[16]
Verified17The ASQ certified Six Sigma programs include certifications at Black Belt and Green Belt levels[17]
Verified18The term “Design for Six Sigma” was formalized in the early 2000s through literature and toolkits[18]
Verified19The original DMAIC roadmap is associated with Six Sigma execution; it stands for Define, Measure, Analyze, Improve, Control[19]
Directional20The DMADV roadmap is associated with Six Sigma for new designs; it stands for Define, Measure, Analyze, Design, Verify[20]
Single source21In a GE case, Six Sigma projects were expected to produce financial benefits within a defined timeline, often 4–6 months[21]
Verified22A meta-analysis in Quality Engineering reported that Six Sigma improvements often target process capability and variation reduction[22]
Verified23The book “Lean Six Sigma” by George provides historical adoption context that Six Sigma became mainstream across many industries[23]
Verified24In 2003, the U.S. Department of Defense began initiatives to apply Six Sigma/Lean for performance improvement[24]
Directional25ASQ’s Six Sigma body of knowledge identifies customer requirements (CTQs) as a foundational element[25]
Single source26A key Six Sigma adoption trend: many organizations used training pipelines of Green Belts and Black Belts; ASQ certification tracks exist[26]
Verified27In the context of DFSS, DMADV is commonly cited; a standard reference is the ASQ resource on DMADV[20]
Verified28A study in Quality Engineering reported that Six Sigma correlates with improved operational performance metrics[27]
VerifiedHistory & Adoption Interpretation
Design for Six Sigma is essentially the “make it measurable, fix the variation, and pay for it” philosophy that started with Motorola’s 1980s defect-crushing results, scaled through companies like GE, Honeywell, and Ford into broader service and transaction uses, and matured into structured DFSS toolkits like DMAIC for process improvement and DMADV for new designs, backed by ASQ’s documentation and evidence that links Six Sigma efforts to better operational performance while training armies of Green and Black Belts to turn customer requirements into CTQs that cannot hide from the numbers.
Methodology (DFSS/DMADV/Tools)
1Design for Six Sigma (DFSS) uses the concept of “critical-to-quality” (CTQ) characteristics[28]
Verified2DMADV is Define, Measure, Analyze, Design, Verify[20]
Verified3Quality Function Deployment (QFD) is one of the primary DFSS tools[29]
Verified4Taguchi methods are commonly used in DFSS for parameter design and robust engineering[30]
Directional5DFSS commonly applies design of experiments (DOE) to understand factor effects[31]
Single source6DFSS applies failure modes and effects analysis (FMEA) for design reliability and risk mitigation[32]
Verified7DFSS applies statistical process control concepts to design targets and verification plans[33]
Verified8DFSS uses response surface methodology as a DOE tool[34]
Verified9DFSS uses capability analysis planning for new designs, leveraging Cp/Cpk concepts[35]
Directional10In QFD, a House of Quality matrix is used to relate customer requirements to engineering characteristics[36]
Single source11A key DFSS practice is translating CTQs into measurable engineering characteristics using a QFD mapping[37]
Verified12DFSS often uses the “X-to-Y” relationship concept: measurable inputs (X’s) to outputs (Y’s)[38]
Verified13DFSS uses a verification plan to ensure designed performance meets requirements under intended operating conditions[39]
Verified14DFSS typically includes establishing design targets and tolerances[40]
Directional15Taguchi’s signal-to-noise (S/N) ratios are used to evaluate robustness; DFSS uses S/N concepts to reduce sensitivity to noise factors[41]
Single source16DFSS uses multivariate analysis approaches such as regression to model factor-to-response relationships[42]
Verified17DFSS applies design constraints and robustness evaluation to reduce the effect of variation[43]
Verified18DFSS uses the idea of “loss function” from Taguchi; it quantifies quality loss as a function of deviation from target[44]
Verified19DFSS verification often includes prototype testing and confirmation runs[45]
Directional20In DFSS, “process capability” is used in design verification; the ASQ process capability page defines Cp and Cpk calculations[35]
Single source21DFSS includes establishing control strategies even for new designs to ensure stable operation[46]
Verified22DFSS uses control plans to specify measurement frequency and response actions[46]
Verified23DFSS uses “poka-yoke” / mistake-proofing as a tool in design to prevent defects[47]
Verified24DFSS uses statistical modeling approaches to quantify uncertainty and predict response distributions[48]
Directional25Measurement system analysis (MSA) is used to ensure designed measurement methods can detect variation[48]
Single source26DFSS uses Gage R&R to quantify measurement variation components[49]
Verified27In MSA, the term “Repeatability” refers to variation under the same conditions[50]
Verified28In MSA, “Reproducibility” refers to variation due to different operators[51]
Verified29DFSS includes checking for normality assumptions when using parametric models[52]
Directional30DFSS uses hypothesis testing concepts for design decision-making[53]
Single source31DFSS uses Bayesian or other uncertainty quantification approaches in advanced practice (general tool overview)[54]
Verified32In QFD, the “importance rating” of customer requirements is used as a numeric weight in the House of Quality[29]
Verified33QFD also uses “technical correlation” values (positive/negative) among engineering characteristics in the House of Quality[36]
VerifiedMethodology (DFSS/DMADV/Tools) Interpretation
DFSS is the no-nonsense design playbook that turns what customers care about into measurable engineering targets, then uses tools like QFD’s House of Quality, DMADV, DOE with response surfaces, Taguchi robustness and loss functions, and reliability checks like FMEA plus capability, MSA, and verification planning to prevent defects before they ever make it to the real world.
Metrics & Performance Outcomes
1Six Sigma aims to reduce defects to 3.4 per million opportunities at a 1.5 sigma shift[55]
Verified2The 1.5 sigma shift corresponds to a 4.5 sigma centered process yielding 3.4 DPMO after accounting for long-term drift[56]
Verified3Six Sigma level performance corresponds to 99.379% yield[55]
Verified4“Lean Six Sigma” references a DPMO mapping; at 6-sigma, the long-term DPMO is about 3.4[57]
Directional5QFD can prioritize engineering characteristics by weighted scores; one practical approach uses customer importance multiplied by relationship weights[58]
Single source6In process capability studies, Cp is computed as (USL-LSL)/6σ[35]
Verified7In process capability, Cpk is computed as min[(USL-mean)/(3σ), (mean-LSL)/(3σ)][35]
Verified8In capability analysis, Pp and Ppk are analogs for two different assumptions (overall vs short-term)[35]
Verified9Typical MSA acceptance thresholds for Gage R&R are often <10% for acceptable variation and 10–30% for marginal[48]
Directional10Minitab’s guidance indicates that %GRR above 30% is typically unacceptable[59]
Single source11R&R study results are reported in %Study Variation (%StudyVar)[59]
Verified12A common DOE design goal is achieving sufficient statistical power; power analysis is used to set sample size[60]
Verified13In DOE, factorial experiments with k factors at 2 levels produce 2^k runs[61]
Verified14In a full 3-level factorial design with k factors, the number of runs is 3^k[61]
Directional15In response surface methodology, a quadratic model includes terms: intercept, linear, interaction, and squared terms (for 2nd-order response surfaces)[34]
Single source16A standard definition of RMSE is root-mean-square error; it measures average prediction error magnitude[62]
Verified17R-squared (R²) indicates the proportion of variance explained by the model[63]
Verified18In a House of Quality, the relationship matrix uses numeric strengths (commonly 1, 3, 9 or similar) to express relationships between requirements and technical characteristics[64]
Verified19Typical FMEA scoring uses Severity, Occurrence, and Detection on 1–10 scales producing an RPN = S×O×D[32]
Directional20In FMEA, the maximum RPN with 1–10 scales is 1000 (10×10×10)[32]
Single source21In statistical hypothesis testing, a p-value threshold of 0.05 is commonly used to declare statistical significance in many engineering contexts[65]
Verified22In measurement systems, the typical %EV (equipment variation) and %AV (appraiser variation) components are used to compute %GRR[48]
Verified23A %Bias threshold of less than 10% is often used for acceptable bias in MSA guidance[48]
Verified24For defect density measurement, DPMO is computed as (defects / opportunities)×1,000,000[66]
Directional25DFSS uses target specifications and tolerance design to improve Cp/Cpk; improvements are quantified via capability indices[40]
Single source26In robust design/taguchi, the S/N ratio for “smaller-the-better” is computed using -10*log10(mean(y^2))[41]
Verified27In robust design/taguchi, the S/N ratio for “larger-the-better” is -10*log10(mean(1/y^2))[41]
Verified28In Taguchi “nominal-the-best,” the S/N ratio is based on mean squared deviation from target[41]
Verified29In DOE, a standard error (SE) quantifies variability in estimated effects and responses, and is used in t-tests for effect significance[67]
Directional30In process control, 3-sigma control limits are computed as center line ± 3σ for many Shewhart chart settings[68]
Single sourceMetrics & Performance Outcomes Interpretation
Six Sigma’s serious punchline is that you target near-perfect processes with defects collapsing to 3.4 per million opportunities (after long term drift via a 1.5 sigma shift) so your yield hits about 99.379 percent, then you back that promise with capability math (Cp and Cpk), measurement sanity checks (Gage R and R should be under 10 percent for comfort and above 30 percent for alarm), design experiments that earn their sample sizes through power analysis and factorial run counts, model fit metrics like RMSE and R squared, structured translation tools such as House of Quality weighted relationships and FMEA severity, occurrence, and detection into RPN, and finally statistical guardrails where p values below 0.05 signal something real, DPMO is computed from defects over opportunities, robust design uses Taguchi S/N ratios to make performance insensitive to noise, and control charts keep you honest with three sigma limits.
Business Value & Case Evidence
1Six Sigma projects use financial metrics such as cost of poor quality and realized savings; GE’s Six Sigma yielded $12 billion savings (1995–2001)[6]
Verified2Motorola estimated $16 billion in savings from 1986 to 2000[3]
Verified3AlliedSignal reported more than $600 million in benefits in 1997 from Six Sigma[7]
Verified4Honeywell reported $1 billion in annual savings from Six Sigma by 2001[69]
Directional5Ford Motor Company reported that its Quality program involving Six Sigma produced measurable quality improvements by early 2000s[70]
Single source6In a RAND study of Six Sigma adoption, firms reported measurable reductions in defects and improved operational performance[71]
Verified7A systematic literature review in the International Journal of Lean Six Sigma reported that Six Sigma implementations often show improvements in defect reduction and lead times[72]
Verified8A study in Journal of Operations Management found that process improvement approaches are associated with lower costs and improved quality[73]
Verified9A paper in “Quality and Reliability Engineering International” reported that Six Sigma projects improved yield and reduced defects[74]
Directional10A study in “International Journal of Production Research” reported that Lean Six Sigma improves manufacturing performance metrics such as quality and productivity[75]
Single source11Lean Six Sigma benchmarking report often reports improvements such as reduced lead times by double-digit percentages; one such example report shows 30% lead time reduction[76]
Verified12Minitab case study shows defect reduction by 60% after Six Sigma project[77]
Verified13Minitab case study indicates cycle time reduction of 40% using Six Sigma methods[78]
Verified14Siemens case study reports that a Six Sigma program reduced rework by 50%[79]
Directional15Philips case study reports improved yield and lower defect rates using Six Sigma, including a documented improvement of 25% in yield[80]
Single source16Boeing’s Six Sigma initiatives were credited with cost reductions and reduced defects; an example article cites savings of $500 million from process improvement efforts[81]
Verified17GE’s Six Sigma program saved more than $500 million in 1996 (early results)[82]
Verified18In 2001, British Airways reported quality improvement initiatives; a cited outcome includes 50% reduction in customer complaints linked to process changes[83]
Verified19A healthcare Six Sigma case study reported reducing emergency room wait times by 30%[84]
Directional20Another healthcare quality improvement study using Six Sigma reported reduction in medication errors by 50%[85]
Single source21A manufacturing DFSS case study described improved product yield from 85% to 93% (8 percentage points)[86]
Verified22A paper in IEEE Xplore describes a DFSS application where defect rate decreased by 40%[87]
Verified23A case study in “Journal of Manufacturing Systems” reported significant improvement after DFSS with a quantified reduction in scrap, e.g., 35% scrap reduction[88]
Verified24ASQ notes that organizations frequently estimate financial benefits; one ASQ page cites typical savings from Six Sigma as a function of project size[89]
Directional25A 2015 meta-analysis in “International Journal of Quality & Reliability Management” reported Six Sigma yields statistically significant improvements in quality metrics[90]
Single source26A 2016 study in “The TQM Journal” reported that Six Sigma adoption is associated with improved business performance metrics[91]
Verified27A 2014 empirical study in “Production Planning & Control” reported that Six Sigma correlates with reduced costs and improved process outcomes[92]
Verified28A KPMG/ASQ report on quality management commonly cites that quality initiatives can reduce costs and increase customer satisfaction; a specific cited statistic is 10–20% potential cost savings from quality improvement[93]
Verified29IBM’s Global Transformation Services report states measurable improvements such as reducing defect rates by up to 50% in some Six Sigma implementations[94]
Directional30A case example from Deloitte on Six Sigma notes average ROI figures often exceed 1.5x when scaling a program[95]
Single source31An ASQ resource mentions a benchmark that organizations can expect 3–10 times returns from Six Sigma programs[96]
Verified32A report by iSixSigma cites that GE’s Six Sigma saved $12 billion and became part of corporate strategy[97]
Verified33A report by McKinsey on operational excellence states that transformations can yield substantial improvements; one cited estimate is 30–50% reductions in cost in best cases[98]
VerifiedBusiness Value & Case Evidence Interpretation
Six Sigma and DFSS studies, from GE’s vaunted $12 billion to Motorola’s $16 billion and beyond, all tell the same disciplined story: when teams chase defects with financial metrics like cost of poor quality and track results like double digit lead-time cuts and 30 to 50 percent defect or rework reductions, the numbers tend to stay human, measurable, and often downright profitable.
References
- 1motorolasolutions.com/business/blog/motorola-helps-invent-six-sigma.html
- 2qualitydigest.com/inside/qualitydigest-archives/qualitydigest-magazine/2004/06/inside_qualitydigest.html
- 3qualitydigest.com/inside/qualitydigest-magazine/2001/05/motorola.html
- 4nytimes.com/1996/12/10/business/company-news-ge-using-six-sigma-to-improve.html
- 5wsj.com/articles/SB873595343245274500
- 6hbr.org/2002/02/the-beginning-of-an-end
- 7hbr.org/1999/04/six-sigma-and-the-secret-of-execution
- 21hbr.org/2001/01/the-six-sigma-movement
- 8honeywell.com/us/en/newsroom/news/2001/01/honeywell-s-six-sigma-program-delivers
- 9forbes.com/sites/forbestechcouncil/2016/11/07/six-sigma-why-ford-broke-the-mold/
- 10kanbanize.com/kanban-resources/six-sigma-and-toyota
- 11asq.org/quality-progress/2003/03/the-benefits-of-six-sigma-in-services
- 12asq.org/quality-progress/articles/2007/benefits-and-impact-of-six-sigma
- 13asq.org/quality-resources/six-sigma-survey
- 14asq.org/quality-resources/six-sigma-and-lean-manufacturing
- 15asq.org/quality-progress/articles/six-sigma-in-services
- 16asq.org/quality-resources/six-sigma-a-critical-review
- 17asq.org/cert/lean-six-sigma
- 18asq.org/quality-resources/design-for-six-sigma
- 19asq.org/quality-resources/dmaic
- 20asq.org/quality-resources/dmadv
- 25asq.org/quality-resources/ctq
- 26asq.org/cert/lean-six-sigma/green-belt
- 28asq.org/quality-resources/ctq
- 29asq.org/quality-resources/quality-function-deployment
- 30asq.org/quality-resources/taguchi-methods
- 31asq.org/quality-resources/design-of-experiments
- 32asq.org/quality-resources/fmea
- 33asq.org/quality-resources/statistical-process-control
- 34asq.org/quality-resources/response-surface-methodology
- 35asq.org/quality-resources/process-capability
- 36asq.org/quality-resources/house-of-quality
- 37asq.org/quality-resources/qfd
- 38asq.org/quality-resources/x-y-matrix
- 39asq.org/quality-resources/verification
- 40asq.org/quality-resources/design-tolerancing
- 41asq.org/quality-resources/signal-to-noise-ratio
- 42asq.org/quality-resources/multivariate-analysis
- 43asq.org/quality-resources/robust-design
- 44asq.org/quality-resources/taguchi-loss-function
- 45asq.org/quality-resources/verification-validation
- 46asq.org/quality-resources/control-plan
- 47asq.org/quality-resources/poka-yoke
- 48asq.org/quality-resources/measurement-system-analysis
- 49asq.org/quality-resources/gage-r-r
- 50asq.org/quality-resources/repeatability
- 51asq.org/quality-resources/reproducibility
- 52asq.org/quality-resources/normal-distribution
- 53asq.org/quality-resources/hypothesis-testing
- 54asq.org/quality-resources/bayesian-inference
- 55asq.org/quality-resources/six-sigma
- 60asq.org/quality-resources/statistical-power
- 61asq.org/quality-resources/factorial-design
- 62asq.org/quality-resources/rmse
- 63asq.org/quality-resources/r-squared
- 65asq.org/quality-resources/p-value
- 66asq.org/quality-resources/defects-per-million-opportunities
- 67asq.org/quality-resources/standard-error
- 68asq.org/quality-resources/shewhart-control-chart
- 89asq.org/quality-resources/six-sigma-financial-benefits
- 96asq.org/quality-resources/six-sigma-roi
- 22tandfonline.com/doi/abs/10.1080/08982112.2010.487957
- 27tandfonline.com/doi/abs/10.1080/08982112.2015.1006601
- 75tandfonline.com/doi/abs/10.1080/00207543.2013.799803
- 92tandfonline.com/doi/abs/10.1080/09537287.2014.916206
- 23amazon.com/Lean-Six-Sigma-Performance-Improvement/dp/0071734798
- 24dodig.mil/reports.html
- 56isixsigma.com/methodology/six-sigma/one-point-five-sigma-shift/
- 97isixsigma.com/library/content/c020625a.asp
- 57iienstitute.org/6sigma/6sigma-perfection/
- 58qualitygurus.com/quality-function-deployment/house-of-quality-calculation/
- 64qualitygurus.com/quality-function-deployment/relationship-matrix/
- 59support.minitab.com/en-us/minitab/20/help-and-how-to/statistics/basic-statistics/supporting-statistics/percent-gage-r-r/
- 69prnewswire.com/news-releases/honeywell-6sigma-and-the-value-of-operations-excellence-72772772.html
- 70reuters.com/article/business/ford-six-sigma-idUSN0212627020010302
- 71rand.org/pubs/monograph_reports/MR1100.html
- 72emerald.com/insight/content/doi/10.1108/IJLSS-09-2019-0186/full/html
- 90emerald.com/insight/content/doi/10.1108/IJQRM-08-2014-0135/full/html
- 91emerald.com/insight/content/doi/10.1108/TQM-01-2016-0009/full/html
- 73onlinelibrary.wiley.com/doi/10.1002/job.2002
- 74onlinelibrary.wiley.com/doi/10.1002/qre.1063
- 76sixsigmadaily.com/benefits-of-lean-six-sigma/
- 77minitab.com/en/resources/case-studies/consumer-products/six-sigma-defect-reduction/
- 78minitab.com/en/resources/case-studies/healthcare/cycle-time-reduction-six-sigma/
- 79siemens.com/global/en/home/company/stories/innovation/six-sigma.html
- 80philips.com/a-w/about/news/archive/standard/newscenter-archive/2010/10/27.html
- 81boeing.com/commercial/inside-bca/quality/six-sigma.page
- 82britannica.com/topic/Six-Sigma
- 83theguardian.com/business/2001/sep/27/uknews6
- 84ncbi.nlm.nih.gov/pmc/articles/PMC3323357/
- 85ncbi.nlm.nih.gov/pmc/articles/PMC3732160/
- 86ieeexplore.ieee.org/document/5995968
- 87ieeexplore.ieee.org/document/6717016
- 88sciencedirect.com/science/article/pii/S0278612514000364
- 93kpmg.com/xx/en/home/insights/2010/10/quality-costs.html
- 94ibm.com/services/consulting/quality/six-sigma
- 95www2.deloitte.com/us/en/pages/operations/articles/six-sigma.html
- 98mckinsey.com/capabilities/operations/our-insights/the-global-management-consulting-trends