The Peter Principle says people rise to their level of incompetence: you get promoted for being good at your job until you land in one you're bad at. For decades it was a joke with a ring of truth. Then economists tested it with real data on tens of thousands of workers and found it holds, in a specific and expensive way.
Key Takeaways
- Benson, Li & Shue analyzed roughly 40,000 sales workers across 131 firms and found firms systematically promote the best performers, who then become worse managers (Promotions and the Peter Principle).
- Other observable traits predicted managerial quality better than sales performance did (the study).
- The costs of these mismatched promotions are substantial (NBER).
- The engineering parallel is direct: promoting your best coder to manager is the same bet.
What the Research Found
"Promotions and the Peter Principle," published in the Quarterly Journal of Economics, used microdata on the performance of sales workers and their managers across 131 firms, roughly 40,000 workers (Benson, Li & Shue). Sales is a useful test case because individual performance is unusually measurable, so you can see exactly who was promoted and how they subsequently did as a manager.
The findings were clear. Firms consistently promoted the highest-performing salespeople. Those people went on to be measurably worse managers than others would have been. And, pointedly, there were observable characteristics that predicted managerial quality better than sales numbers, which the firms had access to and largely ignored. The researchers estimated the cost of these mismatched promotions as substantial (NBER working paper).
Why Firms Do It Anyway
If it's so costly, why is it so common? The researchers offer a real explanation: promotion is also an incentive. Dangling advancement in front of top performers motivates everyone to perform, and that motivational value might partly justify the cost of occasionally promoting the wrong person. Firms also manage the cost intelligently in places, placing less weight on individual sales performance when the management role carries more responsibility (the study).
So it's a genuine tradeoff rather than pure stupidity. But it means many organizations knowingly accept worse managers as the price of a motivational carrot, without ever making that trade explicit.
| What gets rewarded | What predicts management |
|---|---|
| Individual output | Developing others |
| Personal performance | Communication, judgment |
| Being the best doer | Being a multiplier |
The Engineering Version
The parallel to software is exact, and we've written about the symptom before: promoting your best engineer into management is the default path, and it fails a striking share of the time. The Peter Principle research explains why it's structural. Individual engineering excellence is highly visible and easy to reward, while the traits that predict good management, growing people, communicating, making decisions through others, are harder to see and rarely measured. So the promotion goes to the best coder, and the org discovers the mismatch afterward.
A Concrete Version
A team's strongest engineer gets promoted to manage it, because that's the reward available. Six months later the team is struggling: she's still doing the hardest technical work herself because it's what she's good at and what feels urgent, 1:1s keep getting skipped, and two engineers are quietly frustrated at the lack of direction. Meanwhile, a quieter mid-level engineer who instinctively mentors people and communicates clearly, the actual best managerial candidate, was never considered, because his individual output ranked third. The org promoted on the visible metric and ignored the predictive one.
The Honest Counterpoint
This is not an argument that your best engineer can never be a good manager, and plenty are. The research says firms over-weight individual performance, not that top performers are disqualified. Some excellent engineers are also natural multipliers, and promoting them is right. The fix is to stop using individual output as the primary signal and start assessing the traits that actually predict management, then offer a real senior IC track so advancement doesn't require becoming a manager, which is exactly what the engineer/manager pendulum makes possible.
What This Means for Teams
Two practical moves follow. Assess managerial candidates on managerial traits, mentoring, communication, judgment, decision-making through others, rather than on who ships the most code, and validate that assessment the way you'd validate any hiring signal. And build a genuine senior IC ladder, so your best engineers can advance without being pushed into a job the research says they'll likely be worse at. That combination, plus supporting new managers properly, is how you avoid paying the Peter Principle's price. See available engineers.
Frequently Asked Questions
What is the Peter Principle?
The idea that people are promoted based on performance in their current role until they reach one they're not suited for. Research on roughly 40,000 workers found firms do systematically promote top performers who then underperform as managers.
What did the research actually show?
Benson, Li and Shue found firms consistently promoted the best salespeople, those people became measurably worse managers, and other observable traits predicted managerial quality better than sales performance did.
Why do companies keep doing it?
Promotion works as an incentive: the prospect of advancement motivates everyone. Firms may accept some managerial mismatch as the price of that motivation, and they do weight performance less when the role carries more responsibility.
How does this apply to engineering?
Directly. Individual coding excellence is visible and easy to reward, while managerial traits are harder to measure, so the best engineer gets promoted by default. Assess managerial traits instead, and build a senior IC track.
The Bottom Line
The Peter Principle is real and measurable: firms promote their best individual performers, who then make worse managers, while better predictors of managerial quality sit ignored. In engineering that's the default path from best coder to team lead. Promote on the traits that predict management, and give strong engineers a senior IC ladder so advancement doesn't require a job they're likely to be worse at.
Roberto Espinoza is CEO of Ruzora, which helps US startups hire pre-vetted senior LATAM engineers in 72 hours. See available engineers.
