The Case For 20-70-10

Critics call it rank and yank. But this is the kindest, fairest evaluation system out there

By Jack and Suzy Welch

I really support differentiation—ranking employees into performance categories of the top 20%, middle 70%, and bottom 10%, and then managing them "up or out" accordingly. But don't companies face all sorts of resistance when they try to implement this system?

— Yoshihisa Tanizawa, Kuala Lumpur, Malaysia

Resistance may be too soft a word. Yes, differentiation has its advocates, and even its hard core devotees, but no other management practice that we talk or write about ignites the same firestorm of controversy.

So thanks for asking about it. Your question gives us a chance to sort differentiation myth from reality. Done right, differentiation is not, for instance, "rank and yank," with its purported public firings of stunned victims once a year. Nor is it cruel, corrupted by favoritism, and culturally inappropriate.

What is it, then? Very simply, differentiation is based on the principle that the team with the best players wins. If you don't agree with that, then differentiation will never make sense to you. But if you do, then differentiation provides a methodology for making that principle spring to life. How? By rewarding stars in an outsized way that is both soul-satisfying and financially satisfying; by developing "the middle 70" with training and coaching; and by moving out bottom-tier performers so better talent can be brought in.

Basically, differentiation is a way to build meritocracies and continually raise the performance bar, increasing a company's competitiveness with every upward notch.

So why does it spark "all sorts of resistance," as you so accurately suggest? By far the most common reason given is that differentiation is cruel.

We'd make the exact opposite case. Look at it this way. Because of differentiation's performance appraisals, people know where they stand. Maybe the news is not always good, but it does allow them to control their own destinies.

COMPARE THAT WITH COMPANIES where managers, in the name of kindness, allow people, and particularly underperformers, to plod along for years. Then a downturn occurs. Middle-aged underperformers are always the first to get the ax. One by one, their manager calls them in for a conversation that usually goes like this:

"Joe, I'm afraid you have to leave."

"What! Why me?"

" were never very good."

"I've been here 20 years. Why didn't you ever tell me?"

Why not indeed? The employee, years earlier, might have been able to find a job with a future. Now, at age 45 or 50, he must enter a job market more competitive than ever. That's cruel.

The "yank" myth of differentiation says the bottom 10% are summarily fired. In reality, that's rare. More typically, when a person has been in the bottom 10% for a sustained period of time, the manager starts a conversation about moving on. Occasionally, of course, an underperformer doesn't want to go. But confronted with the cold reality of how the organization views them, most people leave of their own accord and very often end up at companies where their skills are a better fit and they are more appreciated.

Moving on to resistance reason No. 2: that differentiation spawns favoritism. The top 20%, it is said, will always be the boss's insiders and pals. To which we say, that's possible. But favoritism is a risk in any evaluation system. At least differentiation's performance reviews, which require quantitative relative assessments of team members, are a countervailing force.

A third common criticism of differentiation is that the continual removal of the bottom 10% eventually forces managers to push out perfectly good employees and thus pits people against one another. But if that thinking is right, why do championship teams replace the bottom of their rosters every year? Because the best organizations, in business as in sports, believe that performance can always improve.

The final reason people resist differentiation is that, even when they see its appeal, they believe it won't work in their company's supposedly unique culture. And what culture is that? Well, we've heard this objection from people at companies small and large, Japanese and Swedish, growing and shrinking, and in every sector under the sun.

In reality, though, we've seen differentiation implemented everywhere. In some situations, it just takes more time, or, because of labor laws, it costs more to let people go. But differentiation should never be rushed anyway. Full-fledged implementation cannot--and should not--occur until an appraisal system has been in place for about three years. People have to get used to candid feedback and the concept of pay for performance.

We're certainly not going to claim that differentiation is perfect. Every management practice has flaws. But we know of no better way than differentiation for companies to build great teams. And neither, we have found, do its critics!

Jack and Suzy Welch look forward to answering your questions about business, company, or career challenges. Please e-mail them at For their podcast discussion of this column, go to

Before it's here, it's on the Bloomberg Terminal.