A DPMO calculator turns a count of defects into defects per million opportunities, the standard way to compare the quality of processes of different sizes. Enter the defects, the units checked and the number of ways that each unit can fail. You get the DPMO, the defects per unit, the yield and the sigma level. A second mode goes the other way and gives the DPMO for a sigma level. It runs in your browser.
How to use the DPMO calculator
- Choose from the defects or from a sigma level.
- Enter the defects, the units and the opportunities per unit. Or enter the sigma level.
- Press Calculate.
The DPMO formula
The Wikipedia article on defects per million opportunities defines DPMO as 1,000,000 multiplied by the number of defects, divided by the number of units times the number of defect opportunities per unit.[1] A defect is a nonconformance of a quality characteristic, such as strength, width or response time, to its specification.[1] For example, if you check 1,000 circuit boards, each with 5 places where a fault can occur, and find 25 defects, there are 5,000 opportunities. The DPMO is 1,000,000 × 25 ÷ 5,000 = 5,000. The calculator also gives the defects per unit, 25 ÷ 1,000 = 0.025, and the defects per opportunity, which is 0.005.
Why opportunities matter
The page says that DPMO differs from reporting defective parts per million because it comprehends that one unit may have several defects of the same type, or several types of defects.[1] It adds that identifying the opportunities for defects, and so how to count and categorize defects, is an art. Organizations generally look at their knowledge of the process, industry standards, the relative importance of each type of defect to customer satisfaction, and the cost of counting.[1] The number of opportunities changes the answer a great deal, so use the same definition every time and record it.
Yield
The yield here is the share of opportunities that came out without a defect, which is 1 minus the defects per opportunity. For 5,000 DPMO, it is 99.5 percent. This is a simple way to read the figure, and it is derived from the formula above. It is not the same as the first-pass yield of whole units, which counts a unit as good only if it has no defects at all.
Sigma level and the 1.5 shift
The article on Six Sigma says that processes that operate with six sigma quality over the short term are assumed to produce long-term defect levels below 3.4 defects per million opportunities. That figure is based on a shift of 1.5 sigma explained by Mikel Harry.[2] The shift is there because processes usually do not perform as well in the long term as in the short term, so an empirically based 1.5σ shift is introduced into the calculation.[2] The page gives the formula for the DPMO of a given short-term sigma level as 1,000,000 × (1 − Φ(level − 1.5)), where Φ is the cumulative distribution function of the standard normal distribution.[2]
The calculator uses this formula both ways. For a sigma level of 3, the DPMO is 1,000,000 × (1 − Φ(1.5)), which is about 66,807. For 4 it is about 6,210, for 5 about 233 and for 6 about 3.4. Going back, it finds the level whose DPMO matches yours. For 5,000 DPMO the level is about 4.08. The result is the short-term sigma level with the 1.5 shift, which is the convention of the Six Sigma tables.
Criticism of the shift
The same article notes that the shift is contentious. The statistician Donald Wheeler called it arbitrary, and the figure means that a six sigma process has long-term defect rates equal to 4.5 sigma performance and not to six.[2] The page also states that the sigma value is mainly a comparative figure, and that Six Sigma at 3.4 DPMO is not the goal of all processes.[2] Treat the sigma level as a way to compare processes and to see improvement, and not as an exact probability.
Using the numbers
DPMO lets you compare a process that makes thousands of simple parts with one that makes a few complex products, because each is measured per opportunity. Track it over time to see whether changes help. In a call centre, an opportunity may be a step in a call that must be right. In manufacturing it may be a solder joint, a measurement or a label. In software it may be a test case. Decide what counts, write it down and keep it fixed.
A worked example
A call centre checks 400 calls. Each call has 8 steps that must be right, such as a greeting, an identity check and a closing summary, so there are 3,200 opportunities. The reviewers find 48 defects. The DPMO is 1,000,000 × 48 ÷ 3,200 = 15,000, the defects per unit are 0.12, and the yield is 98.5 percent. The sigma level is about 3.67. If a change cuts the defects to 16, the DPMO falls to 5,000 and the level rises to about 4.08, which shows the improvement in one number. The figures for 15,000 and 5,000 are worked from the formula above.
Common mistakes
- Counting units and not opportunities. Divide by units times opportunities.
- Changing the number of opportunities between reports. This makes the trend useless.
- Reading the sigma level as a probability. It includes the 1.5 shift.
- Counting a defective unit once when it has several defects. DPMO counts each defect.
What next
Use this often? Add Dpmo to your Home screen.
Frequently asked questions
What is a DPMO calculator?
A tool that finds defects per million opportunities, the yield and the sigma level from defects, units and opportunities per unit.
What is the DPMO formula?
DPMO = 1,000,000 × defects ÷ (units × opportunities per unit).
What DPMO is six sigma?
3.4 DPMO, which assumes a 1.5 sigma shift between the short-term and long-term performance.
How do I convert a sigma level to DPMO?
DPMO = 1,000,000 × (1 − Φ(level − 1.5)), where Φ is the standard normal cumulative distribution.
What is the difference between DPMO and PPM defective?
DPMO counts every defect against every opportunity, and PPM defective counts only units that have at least one defect.
Was this tool helpful?
What went wrong?
We only receive the reason you pick, never what you typed into the tool.
Thanks for your feedback.