DPMO Calculator — Defects Per Million Opportunities

DPMO — defects per million opportunities — is the standard Six Sigma measure of process performance for discrete (attribute) data. Instead of just counting defective units, DPMO accounts for how many chances each unit has to be defective, which makes it possible to compare a simple process (one opportunity per unit) with a complex one (dozens of opportunities per unit) on the same scale.

DPMO is the bridge between everyday defect counts and the sigma level scale: once you know your DPMO, the standard conversion table tells you whether your process performs at 2σ, 3σ, 4σ, or better.

DPMO = (defects ÷ (units × opportunities per unit)) × 1,000,000

Three inputs: the number of defects observed, the number of units inspected, and the number of opportunities per unit — the independent ways each unit could be defective. Yield follows as 1 − DPMO/1,000,000, and DPU (defects per unit) is simply defects ÷ units.

A loan-processing team reviews 400 applications in a month. Each application has 5 defect opportunities (missing signature, wrong amount, incomplete ID, unverified income, mis-keyed account). Across the month they log 38 defects.

DPMO = (38 ÷ (400 × 5)) × 1,000,000 = 19,000
Sigma level ≈ 3.6σ (short-term)   ·   Yield = 98.1%

19,000 DPMO sits between the 3σ (66,807) and 4σ (6,210) benchmarks — a typical, stable-but-improvable business process. Getting to 4σ would mean cutting defects roughly two-thirds, from 38 to about 12 per month.

These are the industry-standard values (including the 1.5σ shift) used in Green Belt and Black Belt training. The calculator interpolates between table rows so its output matches what you would look up by hand.

Treat DPMO as a baseline, not a verdict. Measure it over a representative period, then use it to size the improvement opportunity: each sigma level roughly means an order-of-magnitude fewer defects. In a DMAIC project, DPMO measured in the Measure phase becomes the yardstick the Improve phase is judged against — the same calculation, before and after, with the opportunity definition held constant.

Watch the two classic traps: changing the opportunity count mid-project (which moves the goalposts), and computing sigma from a sample too small to contain any defects — zero observed defects in 50 units does not mean a six-sigma process.