Six Sigma Process Scorer
Turn defect counts into DPMO, sigma level, yield, and the annual cost of poor quality.
Example scenario—not a forecast or verified benchmark. Enable JavaScript to edit inputs, view charts, save locally and export your analysis.
Example inputs
- Process
- Order fulfillment
- Units processed
- 2,400
- Defect opportunities per unit
- 5
- Defects found
- 96
- These counts are
- 12
- Average cost per defect
- 85
- Target sigma level
- 4
Example results
Order fulfillment runs at 3.9 sigma (8,000 DPMO), costing about $98K a year in poor quality; hitting 4 sigma saves $22K.
- Sigma level (short-term)
- 3.9
- DPMO
- 8,000
- Annual cost of poor quality
- $97,920
- Opportunity yield
- 99.2%
- Units with zero defects (est.)
- 96.1%
- Defects per unit
- 0
- Savings at 4 sigma
- $21,914
What stands out
8,000 defects per million opportunities (3.9 sigma)
Order fulfillment gets it right 99.2% of the time at the opportunity level. Only about 96.1% of units come through with zero defects.
Cost of poor quality: $98K a year
1,152 defects a year at $85 each. Hidden costs (lost customers, reputation) are not included.
Reaching 4 sigma removes 258 defects a year
That target allows about 6,210 DPMO and is worth roughly $22K a year in avoided defect cost.
Next actions
- Pareto the 1,152 annual defects by type; the top two types are usually 50%+ of the cost.
- Set a 90-day goal of 4 sigma: about 894 defects a year, down from 1,152.
- Error-proof the most common defect at its source (checklists, required fields, scanning, poka-yoke) rather than adding inspection at the end.
- Track DPMO weekly on a run chart so you can tell real improvement from noise.
Model assumptions & limitations
- Sigma level uses the conventional 1.5-sigma long-term shift: sigma = NORMSINV(1 − DPMO/1,000,000) + 1.5.
- Zero defects is shown as 6.5 sigma (above the 3.4 DPMO six-sigma mark); counting opportunities consistently matters more than precision here.
- Units with zero defects are estimated with a Poisson approximation, e^(−DPU).
- Cost per defect is your estimate; it excludes lost future sales and reputational cost.