I'm trying to explain the core concept of Six Sigma to a new executive team, and I need a clear way to connect the statistic of DPMO (Defects Per Million Opportunities) to the sigma level ($3\sigma$, $4\sigma$, $6\sigma$, etc.). Can someone provide a concise, powerful explanation that highlights why $6\sigma$ is the gold standard in quality management? Specifically, how much of a difference does moving from $4\sigma$ to $5\sigma$ actually make in terms of process output and customer satisfaction? I need to emphasize the exponential impact of reducing variation.
3 answers
The sigma level measures process variation relative to specs; $6\sigma$ means $3.4$ DPMO. Reducing the sigma level exponentially reduces defects, leading to huge savings and improved quality management.
Sigma Level is a metric that tells you how much variation your process has relative to your customer specification limits; it's the number of standard deviations ($\sigma$) that fit between the process mean and the nearest specification limit, assuming a $1.5\sigma$ shift. DPMO is the resulting defect rate. The $6\sigma$ standard translates to $3.4$ Defects Per Million Opportunities. The difference between levels is huge and often misunderstood. For example, moving from $4\sigma$ ($6,210$ DPMO) to $5\sigma$ ($233$ DPMO) is a $96\%$ reduction in defects. It demonstrates the non-linear, exponential benefit of reducing process variation to achieve world-class quality management and customer satisfaction.
Linda's explanation on the exponential drop in DPMO is great for the executive pitch! But I have a follow-up question regarding the $1.5\sigma$ shift assumption. Why is this shift incorporated into the calculation of the sigma level? Does the $1.5\sigma$ shift represent a typical, real-world process drift, and is it always necessary to include it when calculating DPMO for a given sigma level in a Six Sigma project, or is it sometimes excluded for specific types of process excellence reporting?
Christopher, the $1.5\sigma$ shift is a standard convention in Six Sigma to account for the fact that a process mean will naturally drift over the long term (the difference between short-term and long-term process capability). It's a way to provide a more realistic, conservative estimate of the defect rate (DPMO) that will be sustained in a real-world operating environment. While a project team may analyze the process with and without the shift, the standard published sigma levels (like $6\sigma = 3.4$ DPMO) always incorporate this shift to represent long-term quality management capability.
And remember, Steven, the formula for DPMO is: (Number of Defects / (Number of Opportunities $\times$ Number of Units)) $\times$ 1,000,000. It ties the measure of performance directly to the sigma level calculation.