Tuesday, November 19, 2013

Bell curve or Lake Woebegone? How do you rank your employees?

First Marissa Mayer’s was installing her new bébé in the office next to her own, while declaring a fatwa on working from home. Then she started tooling around with the Yahoo logo. Now, it seems, she’s placing her deft hand on the HR process.

Yahoo employees were up in arms about a new policy that forces managers to rank employees on a bell curve, then fire those at the low end. According to AllThingsD, Marissa Mayer reportedly told Yahoo workers that the rankings weren’t mandatory, but many people disagree. The company hasn’t responded to a request for comment. (Source: Business Week.)

As you may recall, the practice of employee decimation, in which the bottom 10 percent gets regularly whacked out of the organization, was implemented by Neutron Jack Welch when he was steamrolling his way through GE.

Thanks to Master Jack, the practice was picked up by other companies, becoming something of a fad there for a while.

The Institute of Corporate Productivity says the number of companies using either a forced ranking system or some softer facsimile is down significantly from previous years. Companies performing well were less likely to be using forced ranking systems than those that weren’t. Just over 5 percent of high-performing companies used a forced ranking system in 2011, down from almost 20 percent two years earlier.

This isn’t surprising.

When companies are performing poorly, they may grasp at the straw that it’s that malingering bottom of the barrel employee cohort who’s holding them back from Doing Great Things. When, in actuality, the problem is more likely systemic, and/or a result of their having misread the market, built crappy products, failed to react to some business tsunami quickly enough, run into a bad luck maelstrom, or just plain old had terrible leadership.

And guess what? Getting rid of 10% of your employees isn’t going to turn witches brew into gravy and turn a failing company around.

When you are going to have lay-offs, you do, of course, rank your employees, and put those you can most easily live without at the bottom of the list.

This doesn’t necessarily mean that they’re terrible performers.

They may not be as experienced or versatile as the next person in line.

You may just get along with X better than you do with Y, so – all else being equal, so long Y.

You may know that A desperately needs the job, while B has been talking about quitting to become a mandolin maker.

And especially in environments where there are continual lay-off rounds, as has been known to happen in  companies that are circling the drain, after the first round, you’ve probably already shed the poor performers to begin with. MIT agrees:

In a company that is going through layoffs, this gets worse over time, wrote several MIT professors in a study of forced rankings in 2006. “As the company shrinks, the rigid distribution of the bell-curve forces managers to label a high performer as a mediocre. A high performer, unmotivated by such artificial demotion, behaves like a mediocre.”

And if you’re not having lay-offs, and regularly purge the company of those who fall into that dreaded left-hand tail of the bell shaped curve, you’re going to have a colossal morale issue (among managers and worker-bees) on your hands.

Let’s face it, unless someone is in sales or doing piece work, it’s pretty hard to rate employees with the level of objectivity that the bell curve suggests is possible.

What if you have a small group, and everyone’s pretty sharp and hardworking? And they all work on tasks that are as much different as they are the same. E.g., working in different areas that require different types of expertise? How do you pit measure one employee against the other?

And since the curve approach only makes sense over a larger number of employees than anyone could possibly have as their direct reports, who does the measuring? Obviously, it’s the individual managers who know best, no? Not the guy perched atop the pyramid…

I’m sure there are all sorts of forms that “metricize” and “objectify” the evaluation process, but this depends on all managers approaching the evaluation task in the same way.

Ha, I say, ha, ha.

Years ago, I worked (as a worker-bee, not as a manager) in a company where it was decreed that, this year, the vast majority of employees would have to be given a score of three (neutral) on their performance reviews. And that the raises for those who scored a three would be minimal.

My boss, of course, hewed strictly to this policy, and gave us all threes. So we all got crappy raises.

His counterpart in another group of equivalent size ignored the policy. Apparently without much pushback, he gave all his folks a four or a five, and they all got decent raises.

Anyway, Yahoo, which probably didn’t need yet another morale problem to begin with, has one of their very own, didn’t-have to- happen making on their hands. Not a good thing.

This can have a particularly bad impact on innovation, arguably the thing Yahoo most needs now. “When employees worry about being ranked at the bottom of the pile, they take fewer risks, said Cliff Stevenson, who studies workforce issues for the ICP.

…“Inherently the problem in ranking is that, unless it’s based purely on objective data—which you rarely see outside of a call center, it brings in a human element. There’s no way to data-fy that,” says Stevenson. In other words, managers’ prejudices and stray opinions get transformed and codified in what appears to be raw data. This seems to be one of the specific complaints being made by Yahoo employees: The rankings are both high-stakes and completely arbitrary.

High-stakes and completely arbitrary?

Doesn’t get much better than that.

Way to go, Marissa!

No comments: