The Commencement Address

Before submitting this column, I had just completed my human resources classes for MBA students at USC. Among the flurry of end-of-semester emails, many were about performance management.

Many of my students aspire to be general managers and to do performance management well.  Yet recent articles on the subject often are more confusing than enlightening. What these articles often depict are debates about performance management techniques rather than principles.

Some argue, for instance, that forced-distribution — or “rank and yank” — is always essential, while others contend it’s always harmful. Many articles state that performance management is the foundation of talent management; others say it should be abandoned. My MBAs see their own organizations devoting efforts to find the right performance rating form or platform.

In honor of graduation, I offer a sort of commencement address to highlight a few fundamental principles. Shall we “commence” to have more performance management conversations that get beyond the hype and treat performance management like any other management process?

Any map works if you don’t know the destination: The debate about performance management focuses on techniques before stating a goal. The goal is often some combination of development and evaluation. Evaluation goals include differentiating rewards and getting rid of low performers. Learning goals include identifying capability gaps, improvement opportunities and development.

It’s tempting to say “maximize both,” but logically learning is often best done in situations where there’s some performance risk. If you can do something perfectly, what’s to learn? There’s an important balance between return on improved performance and return on improved mastery, and you can’t always have it both ways. Ask, “Where would improving the quality or learning of our talent make the biggest difference to our success?”

Question the normal: Performance management often makes an assumption that performance levels follow a “normal” or bell-shaped curve — which is often the rationale for encouraging or requiring leaders to divide their employees into the top 20 percent, middle 70 percent and bottom 10 percent.

But this may not be as normal as we think. In arenas such as entertainment, academic publishing and athletics, there is a vast majority of individuals at the low end and a group of superstars who are far above the average. The top 20 percent may account for 80 percent of the group’s total performance. Some have suggested an overhaul of performance management to reflect a new reality of superstars, but research suggests it’s premature to abandon the normal curve. The important point is to question the normal assumptions.

If you weed out poor performers early — such as through selection or probationary periods — then you may not need to focus on low performance, providing a window to focus on superstars. But in a job that is a testing ground for new hires, finding the low-performers may be essential. Ask, “What are our assumptions about performance differences?”

Not everything that counts can be counted: Performance management is more about the outcomes than the numbers. An HR executive colleague of mine is getting a lot of attention for “abandoning” numerical ratings in performance management. He told me that he’s a bit embarrassed by the idea that this is a path-breaking decision because, as he put it, when they had numerical ratings managers typically gave everyone a rating at the middle of the scale. As a result, the company wasn’t losing much by simply ditching the numbers.

Abandoning the numbers encouraged a focus on deeper conversations about performance improvement, with little loss of information. A fixation on measurement can miss that. On the other hand, when numbers carry a lot of information — such as in call centers or sales, where performance numbers are very closely tied to system performance — it may be wrongheaded to try to get rid of them. Ask, “Are we counting what counts and not counting what doesn’t count?”

Let’s commence to have performance management conversations based on valid principles, not only about today’s popular techniques.

John Boudreau is professor and research director at the University of Southern California’s Marshall School of Business and Center for Effective Organizations, and author of Retooling HR. He can be reached at