In the sixth chapter of their book Six Sigma: The Breakthrough Strategy Revolutionizing the World’s Top Corporations, by Ikel Harry, Ph.D., and Richard Schroeder, called “Unmasking the Hidden Factory,” the authors discuss the metric called “yield”. The basic concept of yield is that it takes the number of units that leave an inspection point after passing inspection, called the “output”, and divides it by the number of units that enter that particular inspection point, known as the “input”. So if 100% of the units pass, the yield is 100%, but if only 50% of them pass, the yield is 50%. This is sometimes referred to as “first-time yield” if the yield is measured at the first point of inspection. The “final yield” would be, in contrast, the yield measured after the final point of inspection.
What is the problem with this metric? Well, the problem with is that not all yields are created alike. From a quality perspective, the total defects per unit is a better measure of the quality of a unit. If 50% of the units of product A and product B are found to have at least one defect, then the yields would be considered the same. But a better metric comparing product A and product B would be the total number of defects per unit (TDPU). So if product A has 2 defects per unit and product B only has on average 1 defect per unit, then product B has higher quality.
First-time yield therefore does not give an indication of quality as reliable as the TDPU. Another way that it is not a reliable indicator is the factor that, if an inspection shows that 5 out of 10 units contain defects, and the inspector throws out the 5 defect-containing units, then someone down the line may think, “wow, all of the units coming out of the last inspection point are defect-free; we must be doing high quality work!” Unfortunately, this is only if the cost of the 5 scrapped units are not taken into consideration.
So that is why it is important to have a good metric of “yield”, because not all yields are created alike.
The next post goes into more detail about the difference between first-time yield, the total number of units with one or more defects, and throughput yield, the total number of defects per defect opportunity.