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.
A Gaussian world would expect pay, performance and employee engagement to cluster near the middle, and the goal of talent management is to shift the broad middle to the right. A Paretian world would expect these elements to reflect a large group of low-to-medium performers, with elites who perform far above the median, creating most of the value and receiving most of the rewards.
Should your talent management systems attract, retain and develop the elite few or focus on improving the median performers? The Gaussian versus Paretian assumption can make a big difference in the answer.
Yet, an answer to this question requires considering one more factor — the return on improved performance (ROIP). In Retooling HR I described ROIP as the increased strategic value created by increased employee performance. Similar jobs have a different ROIP under different strategies. What if the performance of aircraft pilots is Paretian — an elite pilot is many times more proficient than a median pilot? The military should invest in attracting, developing and differentially rewarding elite pilots, because the payoff in lives saved and wartime victories is huge.
Yet, commercial airlines or shipping companies should not invest heavily in elite pilots. Commercial pilots must achieve consistency and adherence to a specific standard. An elite pilot in the cockpit of a commercial aircraft is still capable of mind-bending dogfight maneuvers, but they are the last thing you want from a commercial aircraft pilot — same Paretian distribution, different payoff.
Optimizing talent investments increasingly depends on questioning assumptions. Whether Gauss or Pareto best describes your employees’ performance or whether ROIP is steeply sloped or flat, both have important implications to improve employee performance and optimal talent management investments.
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: Using Proven Business Tools to Make Better Decisions About Talent.” He can be reached at firstname.lastname@example.org.