I reckon the chances of actually hitting all the low performers is exceptionally low in a corporate context.

Much more likely outcome is you drop a normal distribution of them - some good, some bad, most avg.