Hi tww - hope you're well this evening. What a great question! And you are, in fact correct. Good for you to question the numbers!
DPMO allows us to examine each unit and determine how many possible ways it can fail. PPM looks at a P/F for each part. They will be the same number if you consider the number of opportunities for failure to be - you guessed it - 1. Here's an example. A company is counting inventory. They scan in barcodes - either the scan works or it doesn't. Now, the numbers are one and the same. If, however, I were scanning for 6 pieces of information, now the numbers are considerable different, yes?
Both numbers can/should be important in most situations - management needs to know the acceptable "pass" rate - that's ppm, right? But DPMO might tell a very different story. If there are 5 opps to fail, and I'm failing 4/5 most of the time - well that's obviously more severe - or requires a different attack plan - vs. if I'm failing on one recurring characteristc, right?
When you calculate DPMO, just make sure you are very clear on what the "opportunities for failure" are, get agreement from management, and remain consistent as you track your improvement efforts.
One more thought/observation: I have seen DPMO manipulated by management to make the number look
better by not counting all the opportunities. I've also seen the number manipulated by Quality Managers in the other direction.
Key point: create a clear
operational definition of the metric and ensure it's communicated and understood. Then you can start to make good decision. Dr. Deming said something like this - Data only has meaning when we know how it was counted/collected. Or something to that effect. If you want more info on
operational definitions, let me know.
Hopes this helps. Cheers!