I thnk the first place to start with is the criticality of your product. If you're making heart pacemakers you cannot afford even 1 defective component. Thus the extent and type of inspection is going to be proportional to the risk involved.
I'm up to my ears with a customer which has started a program where they are now requiring corrective actions for what used to be line accums - the onesie twosies. It was getting to the point of being rediculous. What it boiled down to is we had a meeting and told them we would group like defectives (over 2000 shipping part numbers) and we would not agree to 100% defect free product as that was not how the job was quoted. It was quoted as involving hand operation(s) (metal stamping) and we told them for 100% defect free we would have to go to optical inspection or a poke yoke. The bottom line was: "You want a requote, we'll tell you what it will cost for 100% defect free product". So far we've heard nothing back. We would be looking at new dies in many cases, an increased die PM frequency, a poka yoke inspection system, possibly some robotics, etc., etc.
The folks above are right: 100% inspection by a human will not stop all defectives all the time from getting through.
And I do agree that preventing the problem is the ideal. We all want to do that. But in many cases the reality is that there are many processes out there that simply in real life are not capable of producing 100% defect free parts at a price that is acceptable.
Can you imagine a 100% defect free car? Hmmmmmm? I wonder how much it would cost...
I recently had a 2 month old hard drive die on me...
Pacemakers are expensive for a reason.