You’ve probably encountered managers you admire more for their technical skills than for their actual leadership skills.
Perhaps it’s the familiarity of this experience that lends the Peter Principle its popular appeal. The Peter Principle, laid out in a 1969 book by Dr. Laurence J. Peter, describes the following paradox: if organizations promote the best people at their current jobs, then organizations will inevitably promote people until they’re no longer good at their jobs. In other words, organizations manage careers so that everyone “rises to the level of their incompetence.”
The Peter Principle problem arises when the skills that make someone successful at one job level don’t translate to success in the next level. In these cases, organizations must choose whether to reward the top performer with a promotion or to instead promote the worker that has the best skill match with a managerial position. When organizations reward success in one role with a promotion to another, the usual grumbles ensue; the best engineer doesn’t make the best engineering manager, and the best professor doesn’t make the best dean. The same problem may apply to scientists, physicians, lawyers, or in any other profession where technical aptitude doesn’t necessarily translate into managerial skill.
Investigating the Peter Principle
While the Peter Principle may sound intuitively plausible, it has never been empirically tested using data from many firms. To test whether firms really are passing over the best potential managers by promoting the top performers in their old roles, we examined data on the performance of salespeople and their managers at 214 firms. Sales is an ideal setting to test for the Peter Principle because, unlike other professional settings, it’s easy to identify high performing salespeople and managers—for salespeople, we know their sales records, and for the sales managers, we can measure their managerial ability as the extent to which they help improve the performance of their subordinates. The data, which come from a company that administers sales performance management software over the cloud, allow us to track the sales performance of a large number of salespeople and managers in a large number of firms. Armed with these data, we asked: Do organizations really pass over the best potential managers by promoting the best individual contributors? And if so, how do organizations manage around the Peter Principle?
First, we found that sales performance is highly correlated with promotion to management. For salespeople, each higher sales rank corresponds to about a 15% higher probability of being promoted to sales management.
Second, sales performance is actually negatively correlated with performance as a sales manager: when a salesperson is promoted, each higher sales rank is correlated with a 7.5% decline in the performance of each of the manager’s subordinates following the promotion. We found similar results regardless of whether salespeople were promoted to their own team or to new teams. In other words, firms tend to promote top sales workers into management, even though they become the worst managers.
In our data, among people who were actually promoted, better salespeople ended up being worse managers. But if we could observe the managerial potential of all salespeople, and not just those who were promoted, would we still find a negative correlation between sales performance and managerial performance?
Answering this question is difficult because the promoted managers we observed in the data weren’t promoted at random. For example, if firms promoted by flipping a coin, then poor salespeople could get promoted because they were lucky, rather than being promoted because their employer observed qualities that overcame their deficiencies as salespeople. Although people aren’t getting promoted by coin flips, they are more likely to be promoted if they happen to be in the right place at the right time: using variation in the promotion rates across industry over time to act as our coin flips, we still find that better salespeople tend to be worse managers.
We also found that firms underweight other indicators that a salesperson would be a good manager. In particular, we found that salespeople whose sales credits were shared among a large number of collaborators become very effective managers. Credit sharing for enterprise sales is typically a mark that the salesperson was involved in large, complex deals requiring collaboration. This type of collaboration experience positively predicts managerial quality.
How do firms manage around the Peter Principle?
Firms have long wrestled with the Peter Principle, and our exploration of the data reveals how firms have tried to minimize the costs of promoting top workers who become bad managers.
First, firms can reward top performers with pay rather than promotion. In our data, we found that firms with the strongest pay-for-performance also promoted the best managers. In other words, by rewarding sales performance with greater incentive pay, firms are free to promote the best potential managers. The best salespeople don’t feel they “have to” become managers in order to earn more money.
But promotions aren’t all about pay, they’re also about prestige. Other organizations (such as Microsoft) have avoided tying promotions to changes in responsibility by using dual career ladders, for instance, by promoting excellent programmers up a technical track and excellent leaders up a managerial track, with similar job levels in each equating to similar pay and prestige. These ladders allow people to progress in their career, drawing on their existing passions and talents rather than requiring them to shift job duties.
A second solution is to let managers be managers: promote the best candidates for the managerial job role, let them manage large teams, and isolate their managerial responsibilities from their individual contributor responsibilities. We find that when firms assign managers more responsibility over larger teams, firms are more willing to promote workers who are weaker in terms of sales but more likely to be effective managers. Separating managerial and sales responsibilities also limited conflicts of interests and other issues that arise in “player-coach” arrangements.
Both solutions can be implemented as part of the performance evaluation process. One approach, embedded in evaluation regimes like the ninebox, asks raters to decouple evaluating future career potential from prior job performance. People who score highly on future career potential can be rewarded with promotion to management roles and stock options to retain them until their potential can be realized. People who score highly on prior job performance can be rewarded with bonuses, promotions up an individual contributor track, or recognition. The process should be designed to recognize and reward excellence in one’s role without necessarily changing one’s role.
Incentive pay, dual career ladders, and thoughtful performance evaluations can recognize that people contribute to the success of the organization in different ways. But it seems that, at least in sales, companies nonetheless reward sales talent by promoting top sales workers into management.