Should your talent management systems attract, retain and develop the elite few or focus on improving median performers?

That is a question to which John Boudreau responds in an article written for Talent Management magazine. Here is a brief excerpt. To check out all the resources and sign up for a free subscription to the TM and/or Chief Learning Officer magazines published by MedfiaTec, please click here.

* * *

In 1906, Italian philosopher Vilfredo Pareto observed that 80 percent of the peas in his garden were contained in only 20 percent of the pods, and proposed the Pareto Principle — roughly 80 percent of effects come from 20 percent of causes. Similarly, in 2002 Microsoft found that fixing the most impactful 20 percent of software bugs would prevent 80 percent of system crashes. Do you think that 80 percent of the value of employee performance might come from the highest-performing 20 percent of your employees?

Carl Friedrich Gauss, who lived a century or so earlier, is regarded as the “Prince of Mathematicians.” He is the namesake of the Gaussian function, or bell curve, one of the most influential ideas in statistics and probability theory. Do you believe that employee performance tends to cluster around the middle and it’s very rare to see performance outliers?

The debate over Gaussian versus Paretian is not as popular as which side you take in the Twilight Saga, but don’t glaze over at the prospect of a statistics lesson. The different views of these two icons of statistical history have important implications for optimizing strategic talent management and where your talent management systems may be making dangerously incorrect assumptions.

Figure 1 contrasts Pareto versus Gauss. The vertical axis is the number of things having a certain value, and the horizontal axis is the different values they take. The Gaussian view is the black line, and the Paretian view is the gray-shaded area. In talent management, the horizontal axis would be employee performance levels, and the vertical axis would be the number of employees at each performance level.

The Gaussian distribution has most employees clustering near the middle, or median, with a symmetrical falloff toward the extremes. The Paretian distribution has employees clustered near the mid- to low level and a longer tail of high performers. The Paretian view has extremely high performers who are much further from the median than the Gaussian view, which is why the elite 20 percent might generate 80 percent of the total performance in a Paretian world, but not in a Gaussian world.

* * *

To read the complete article, please click here.

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 there.