“To me, one man is worth ten thousand if he is first-rate.” – Heraclitus (Fragments, c. 500 B.C.)

In a memo to one of my clients, I recommended that he create a policy for promoting a standard of employee excellence that would include his eliminating the worst 15% of his employees every year. He didn’t like the idea. Thought it would be bad for morale.

Announcing the policy might make people nervous — so this is not the kind of policy I’d recommend publicizing — but putting it into effect would improve the long-term esprit de corps.

In my view, what’s bad for morale is being surrounded by deadbeats who get paid and sometimes even promoted despite their shiftlessness. What’s also bad for morale is the idea that something other than the quality of work determines success in a given corporate culture.

If you believe, as I do, that culling weak employees is essential to success, you may sometimes be worried about which employees to keep and nurture, which to warn and sanction, and which to dismiss peremptorily.

This is a particularly difficult question if you are a manager of managers, since you will be making decisions about nothing but capable and relatively hardworking people.

One way to help make these decisions is to use a tool developed at General Electric by legendary CEO Jack Welch.

Welch divided his employees into four groups:

A. those who deliver on their commitments and share corporate values

B. those who NEITHER meet their commitments NOR share corporate values

C. those who don’t meet their commitments but do share corporate values

D. those who deliver on their commitments but do NOT share corporate values

In Welch’s view, A employees were to be kept and nurtured and B employees were to be dismissed right away. But knowing what to do with C and D employees could determine how successful a company would eventually be.

According to earlier reports on this subject, Welch was said to be in favor of working on the C’s and dismissing the D’s, reasoning that workers with the right attitude can learn to be productive but that those with the wrong attitude can’t be changed.

I’ve found that the opposite is often true. I’ve been able to tame employee-eating performers by counseling them in moderation, but I’ve never been able to turn a loyal lagger into a star performer.

And there is a danger in surrounding yourself with people who buy into your program (or tell you they do) but can’t seem to get the job done. Because they are such strong believers, you tend to put the blame for their failures where they do — somewhere else. Do that enough, and your business will one day stop working and you and your loyal non-performers will wonder what went wrong.

Welch himself seems to have come to that same conclusion later on in his career. In a relatively recent speech to GE managers, he said, “Too many of you work too hard to make C’s into B’s. It’s a wheel-spinning exercise. Push C’s on to B companies or C companies, and they’ll do just fine. We’re an A-plus company. We want only A-plus players.”

My suggestion is this: With C and D employees, be frank in telling them where they fall short and insist that they measure up. Give them a deadline for improvement (serious, measurable improvement) and stick to it.

Most of the time, your efforts will be fruitless — but so long as you keep them to a limited number of clear and precise communications, the time you invest will be amply compensated for by the few transformations you do achieve.