Let's try this
Dear Dawn:
To follow up on the responses you have received already, let me try this. I still remember a chart on the wall when I visited the folks in the purchasing department at GM. It was something like this. Let's say you get a quote from three different suppliers.
If you save $0.01 on the part
At 100 pieces, you are saving just $ 1.
At 1000 pieces, you are saving $10.
At 1000,000 pieces, you are saving $10,000.
At 10,000,000 pieces (remember there are four tires per car), you are saving $100,000.
Take all the millions of parts that are being produced everyday and go into a typical automobile; pretty soon, the $0.01 becomes a big chunk of money - several billions of dollars really.
The same goes in many other areas. Engineers are usually thinking about manufacturing processes and number of "defects" or "defectives" (one part, or one defective, can have more one or more defects). If you are a Six Sigma company, you are operating at 3.4 defects per million opportunities, or shall we say, 1 defect per million opportunities - or 1 in a million.
Mathematically, 1 PPM is 1 times 10^(-6), just as 1% is 1 times 10^(-2).
The post office delivers millions of pieces of mail everyday. The UPS, FedEx, etc. deliver millions of packages everyday. Guess what would happen if just 1% of these hundreds of millions of items were to be delivered, every single day, to the wrong address? There would be total chaos.
I remember reading about how the former Chairman and CEO of GE, Jack Welch, got sold on the idea of implementing the Six Sigma philosophy or methodology (originally developed at Motorola), after he grasped the idea of "variability". Statistically speaking, this is measured by what is called the standard deviation, or sigma, for N pieces of data. If you cut delivery times from 16 days to 8 days, you have not really achieved a 50% improvement. A process may be operating at 16 PPM one day and at 8 PPM the other day. Again, this is not 50% improvement. This is all just a part of the natural "variation" that we see every day in complex processes - described mathematically by the "normal" or the Gaussian distribution model.
The challenge is to reduce variability. Why? Ultimately, all of this (PPM talk, I mean) has to do with what we call customer satisfaction, or dissatisfaction. If you have to wait 6 minutes in the drive through lane at McDonald's instead of 2 minutes, you may never come back. Hope this helps.
Charmed