Rank and Yank: Would a Rose by Any Other Name Smell as Sweet?

Jack Welch joined the fray about Microsoft’s abandoning “rank and yank” in a Wall Street Journal op-ed titled “Jack Welch: ’Rank and Yank’? That’s Not How It’s Done.” Welch now says that rank and yank was media-created and that it should be more appropriately referred to as “differentiation,” which is a much more caring, humane and employee-centered appraisal process. I was around when Welch was in his heyday at GE and when every word that Jack said in public ended up in lead articles in magazines and newspapers or on TV and I never remember hearing the word “differentiation.” What I remember was that Welch had little confidence in his managers’ ability to give employees bad news.  Consequently, he created a system to force them to do what they were unlikely to do effectively: deal with poor performance.

The problem is that differentiation as Welch describes it is not much better than “rank and yank.”  There are critical flaws in both. Whether it is called rank and yank or differentiation, it is the structure and assumptions about human behavior that have made it the most dreaded activity for managers and employees at thousands of businesses worldwide.  While I appreciate that Welch wants his systems to seem more caring, compassionate, etc., he knows better than anyone that in the end it is not about intention but impact. Lest those who read Mr. Welch’s description of the way differentiation should work actually attempt to implement his ideas, let me take his points one by one and offer a different opinion – one that is based on the science of behavior. First, there are two things in the op-ed that I agree with:  

1. Candor is essential; and
2. Every employee needs to know where he/she stands. 

Below are some of the assumptions that Welch claims make differentiation a better, more effective process than the dreaded rank and yank.  Welch claims:

1. The 20-70-10 distribution is not set in stone – But the ranking or the rating on some curve is set in stone!  It is the heart of the system.  I don’t know where Welch came up with the 20-70-10 distribution, but I doubt if it is data-based.  No company that I know of hires people on a distribution like that.  Would we hire someone who we expect to be at the bottom?  Everyone we hire is expected to be at the top.  Yet no company I know of is happy when all people are rated at the top. That is because we believe performance appraisals of any kind are motivational.  They are not.  Demotivating?  Bet on that.  He says no one thinks that grading in school is cruel.  I don’t want to get started on that because I don’t have enough space to scratch the surface of education. Just let me say that it borders on criminal to promote a student to the next grade when he/she got mostly Cs and Ds.  I believe that the only grade in school should be an A and every student should be coached till he earns it.  (Read my blogs for more on this subject.)
2. Differentiation’s performance appraisals are not about the numbers – If you believe that one, as the saying goes, I have a bridge in Brooklyn I will sell you real cheap.  All traditional appraisals are about the numbers.  Appraisal of behaviors is rolled into the final number that determines the ranking.  How can you have a ranking without numbers?  Of course, it is about the numbers.
3. The middle 70 percent know that they are appreciated – Wrong again.  The 70 percent are not idiots.  They are in the middle, and they have been taught that is mediocre.  Who strives to be mediocre?  What they also know is that those in the top 20percent will almost always be in the top 20percent and they will always be in the 70percent.  Motivational?  I don’t think so.
4. Differentiation management evaluates employees at least once (and preferably twice) a year – The data are clear that annual appraisals do little to change performance.  I do agree that we need to talk to everyone about where they are and where they want to be, but that does not need to be in the form of a performance appraisal.  As I have said often, “The best job you will ever have is one where you know at the end of every workday how well you have done.”  Generally speaking, that job is more likely to be in sports than in business.  In sports, you know how you are doing at every swing of the bat or golf club, after every play – not at the end of the year or every six months.  The annual or semi-annual appraisal is too little, too late.  You don’t make a bad process of any kind better by doing it more often.
5. Differentiation builds great teams  – It does not.  It separates employees.  By its very nature, it is a zero-sum game.  Since it is a ranking, there can be only one person at each position on the distribution.  Although I admit that rewarding employees for teamwork will offset some of the negatives associated with the competition engendered by ranking, the most powerful consequences – promotion, assignments and pay – are associated with the ranking.
6. Differentiation starts with communication – While I agree that communication is important to let people know what numbers and behaviors are important, what changes performance is not about communication but about behavioral consequences – what happens to someone as a result of what he or she does or doesn’t do (i.e., behavior).  These consequences communicate the reality of what is essential and what is not.

Although Welch now says that that the distribution, 20, 70, 10, was not set in stone, it was interpreted as such by many inside GE and in hundreds if not thousands of companies outside GE.

See also: OOPS! 13 Management Practices that Waste Time and Money (and what to do instead).