Organizations and teams can calculate the sigma level of a product or process using the equation below:
Consider a process in a marketing department that distributes letters to customers or prospects. For the purposes of the example, imagine that the process inserts 30,000 letters in preaddressed envelopes each day. In a given business week, the process outputs 150,000 letters.
The marketing department begins receiving complaints that people are receiving letters in envelopes that are addressed to them, but the letters inside are addressed to or relevant to someone else. The marketing department randomly selects 1,000 letters from the next week’s batch and finds that 5 of them have errors. Applying that to the total amount, they estimate that as many as 750 letters could have errors. (Sampling and extrapolation are covered in depth in the advanced chapters on statistics.)
The letter process has 150,000 opportunities for error each week and an estimated 750 defects.
((150,000 – 750) / 150,000) * 100 = a yield of 99.5
Look up a yield of 99.5 in the abridged Sigma table below and you’ll see the process described above is currently between 4 and 4.1 sigma.