Forced Rankings Are Institutionalized Stupidity at Its Worst

A lot of stupid things go on in the technology industry, but forced ranking and rating, tied negatively to Jack Welch and GE, in my mind, is near the top. This was made clear in a recent Vanity Fair article linked to Microsoft, but I doubt there are many companies in the U.S., thanks largely to Welch, that don’t have a similar policy.

On a national scale, you have to wonder if Welch is largely to blame for why the U.S. as a nation is less competitive in a global market. We’ll leave that for another time as we focus here on forced ranking as an example of institutionalized stupidity.

Forced Ranking: Great Employees Leave, Dead Weight Remains

Generally, when you bring up forced ranking and point to the reasons it is a company killer, the most common defense is that some other successful company does it—but then that would be like pointing to the ex-vice president of the United States and suggesting that his practice of shooting friends in the face with a shotgun should be more widely adopted. In short, forced ranking is an employee evaluation process in which managers are required to distribute ratings for those being evaluated.

What is telling is that Steve Ballmer’s response to Vanity Fair never even mentions Forced Ranking. Like a real-life emperor showing off his new clothes, Ballmer appears absolutely sure that isn’t a problem.

Two causes likely lie underneath this big problem. One is group behavior, which drives people to behave similarly (either in good or bad ways), and the other is the confirmation bias that blinds us to stupid decisions we all make. Forced ranking institutionalizes and drives an unacceptable level of failure—yet even numbers guys like Ballmer are blind to this simple fact and, often like Ballmer, will refuse to even look at it.

If you like forced ranking, I’d like you to step back, look at it mathematically and see what it means to a company. Generally, forced ranking says that about 20 percent of your people are really good, about 40 percent of your people are good, 20 percent are OK, about 13 percent of your people should only barely keep their jobs and about 7 percent of your folks should be fired, according to Microsoft’s breakdown.

Sounds reasonable, right? Except why the hell wouldn’t you fix a process in which only 20 percent of your people were really doing a great job? Think about it. If you had a product with a 7 percent failure rate, you’d be out of business—yet depending on where you make that lower cut, forced ranking institutionalizes a high failure rate.

You see, you can’t predict quality. A good quality process measures actual quality and has remedial action based on the actual results. If you have a run in which 90 percent of the products are low quality, then you throw out 90 percent of the products or remanufacture them. If you have run with no defects, you might double-check the results, but you wouldn’t throw out 7 percent at random because “it’s policy.” I say you wouldn’t, but I’ll bet if we looked, we could find a line that did.

Commentary: Microsoft Turns 35: Best, Worst, Most Notable Moments

If you’re in quality control and you know you have to have a 7 percent failure rate, then you design a process to reduce it. You see, when you talk to experienced managers, they intentionally hire the number of underperforming employees they need in order to be able to reward the people they want to keep. In effect, part of the cost of having great people is employing folks who will be, at best, dead weight and, at worst, walking disasters and being perfectly OK with it. (Or you just eliminate a high percentage of the recently hired.) When I first ran into this practice, my manager told me I got a low review because all first-year employees are reviewed this way because we clearly couldn’t know our jobs yet. Let’s just say that review didn’t stick.

What is interesting is that executives such as Ballmer support the program having never been measured under it themselves— and forced ranking looks good on paper. You can see how the industry got there.

As you can imagine, there are a couple of problems with this. One is discrimination and the need to put in place some kind of objective metric to assure everyone was treated equally, and the other is a tendency, particularly during the growth phases of a company, to hand out raises indiscriminately. Forced ranking appears to be race and sex independent (though it really isn’t) and allows executive to be far better able to predict future compensation. Now, you could do that with fixed compensation pools, but managers tend to divvy them up equally, which meant the high performers leave the company and the low performers aren’t motivated to improve.

How Forced Ranking Asserts Confirmation Bias

There are two problems with forced ranking, both identified in the Vanity Fair article. One is that you are generally held to a maximum of 20 percent overachievers. This is OK in a market that dramatically needs people, but it’s not so wise in times of high unemployment. The other, far bigger problem is that it puts employees in heavy competition.

When employees compete, they steal credit, excessively criticize or torpedo peer projects and engage in other behaviors you’d rather not see in an entity called a “company” where folks are expected to cooperative. You thus have a process that institutionalizes performers in a way that the majority of employees are at or below average, while those who effectively cheat are rewarded. I bet we could go through every major Microsoft failure over the last decade and find a group of people at the core of it that got into that position by gaming this flawed system.

Blog: Microsoft’s Future: What’s the Worst That Could Happen?

No sane person could sustain the argument for forced ranking once it’s applied it to products instead of people. Apply it to automobiles and make 20 percent or even 10 percent of any run unsatisfactory by policy, regardless of actual quality, and you’d immediately see that you were institutionalizing bad quality. With people, though, folks remain blind to the fact that forced ranking is walking example of confirmation bias.

Executives simply can’t see that, like the aforementioned naked emperor, they will at some point feel really foolish for not abolishing forced ranking in favor of practices that encourage the creation of a firm made up of overachievers—kind of like how Microsoft started out.

My lasting lesson isn’t on Forced Ranking at all, though. Rather, it’s how confirmation bias can make otherwise smart people do really stupid things. If someone as smart as Ballmer or Bill Gates can’t see this, then how many things that you take for granted deserve another look? Maybe that employee who raised his hand to say what the firm was doing was stupid had a point.

If you want to learn more about forced ranking, this thoughtful piece recommends that, if you have a choice, you should never be an employee at a company that supports forced ranking. Confirmation bias is the far bigger problem, but reading up on argumentative theory will help explain it. Another interesting and critical piece of work is one I return to often, Five Tips to Avoid Confirmation Bias, since like you I am equally hardwired to do stupid things I’ll later regret. I use the fifth tip, “Google (or Bing) better,” a lot—I often find that my initial position results from confirmation bias, and following that rule has saved my butt more than once.

Rob Enderle is president and principal analyst of the Enderle Group. Previously, he was the Senior Research Fellow for Forrester Research and the Giga Information Group. Prior to that he worked for IBM and held positions in Internal Audit, Competitive Analysis, Marketing, Finance and Security. Currently, Rob writes on emerging technology, security and Linux for a variety of publications and appears on national news TV shows that include CNBC, FOX, Bloomberg and NPR.

Follow everything from CIO.com on Twitter @CIOonline, on Facebook, and on Google +.