> * Because your metrics might not be good enough to actually discriminate low performers from high performers I think this is very true. For many complex role

> * Because your metrics might not be good enough to actually discriminate low performers from high performers I think this is very true. For many complex roles you land-up with all sorts of indirection to create 'metrics' which are often some input with a lot of opinion. I've seen lots of situations where you look through performance review data to see employee ratings bouncing all over the place as they (or whole teams) shift from manager to manager. That's not to imply that managers are negligent, it's just that in many roles it's perfectly possible to have a different view of performance (and everything else). > * Because such a policy encourages rivalry between employees instead of cooperation (not in the benefit of the company) See stack ranking! The other issue is that an organisation can develop a reputation for being too fast to pull the trigger which can impact retaining talent and hiring new people. Given that metrics aren't clear, every 'low performer' situation involves friends and colleagues who don't think that the individual was a low performer. Plus the person leaving is unlikely to be positive about the organisation whatever the compensation is. That may be fine from a management philosophy perspective, some people think it's good as "only poor or average performers will worry".