Why Your Best Performer Becomes Your Worst Manager

You have a star. They out-produce everyone, they never miss, and a management seat just opened.

You already know what you’re going to do. Everyone does. It feels less like a decision than like arithmetic.

Nine months later that person is miserable, their team is quietly updating résumés, and you have lost your best individual contributor and acquired a struggling manager. One promotion, two losses.

This is so common it has a name, and the name is wrong.

The Peter Principle was a joke. Literally.

You’ve heard that people “rise to their level of incompetence.” It gets cited in business writing constantly, usually as though it were a research finding.

It was a comedy book. 1969. Laurence Peter and Raymond Hull were being satirical.

The actual research arrived fifty years later, and it’s more interesting than the joke. Alan Benson, Danielle Li and Kelly Shue, in the Quarterly Journal of Economics in 2019, used personnel data on roughly 53,000 sales workers across 214 firms. What they found: companies promoted disproportionately on current sales performance rather than on any prediction of management ability — and measurably produced worse managers than a different promotion policy would have.

So the joke was right. It just wasn’t evidence until someone did the work.

Worth scoping honestly: that’s sales-force data, not a universal law about every industry. But if the pattern holds anywhere, it holds in the place where individual output is easiest to measure — which is exactly where the temptation is strongest.

Why we do it anyway

Three reasons, none of them stupid.

Promotion is the only reward most organizations have. If the sole way to say “you’re excellent” is to change someone’s job title, you will eventually change the job title of someone who was excellent at the job they had.

Performance is measurable and readiness isn’t. You have numbers for the first. For the second you have a feeling, and feelings lose arguments with spreadsheets.

It feels like fairness. Not promoting your top producer feels like punishing them. It reads as a rebuke rather than a design decision.

None of that makes it work.

The three questions you’re asking as one

Here’s the fix, and it’s less complicated than it sounds. Three genuinely different questions get collapsed into a single conversation.

Performance — how well are they doing the job they have? Timeframe: past and present. Evidence: results, consistency, quality. Anyone with the data can judge it.

Potential — what more might they be capable of? Timeframe: long-term. Evidence: how fast they learn, curiosity, what they do in genuine ambiguity. This requires a trained eye, not a dashboard.

Readiness — can they do the next thing now? Timeframe: this quarter. Evidence: demonstrated behavior in a smaller version of the future role.

That last line is the whole thing. Readiness isn’t a prediction. It’s an observation — and you can only observe it if you’ve built somewhere to observe it.

The danger zone

Put those together and you get a grid.

	Ready now	Not yet ready
High performer	Promote — and build the smaller stage first anyway	The danger zone. This is where good employees get turned into struggling managers as a reward.
Struggling performer	Rare, and worth investigating — usually the wrong seat, not a bad employee	Develop in place

Almost every failed promotion you’ve watched lives in that top-right box.

The smaller stage

The move is not to withhold the promotion. It’s to stop guessing.

Before you make someone a manager, give them a bounded version of the job and watch:

• Lead a project with real stakes and no authority to compel anyone. Influence without a title is the actual skill.
• Give them one person to develop — a new hire to onboard, a junior to mentor. Then ask the question that matters: does that person come out of the interaction more capable, or smaller?
• Watch them be wrong in public. How someone handles their own error in front of others tells you more about their management future than any performance metric.
• Give them something confidential. Not a test, just an ordinary piece of sensitive context. Notice what happens to it.

None of this requires a title, a budget, or an org-chart change. All of it produces evidence.

And build somewhere else to go

Some of your best people should never manage anyone, and telling them so should not be a demotion.

Senior, principal, or master-craftsperson titles. Project leadership without people-management. Mentor-of-record roles with real recognition attached. If your organization’s only path upward runs through supervising humans, you will keep converting excellent specialists into unhappy managers, and you’ll keep calling it a reward.

The point isn’t to promote fewer people. It’s to stop making one decision when you’re actually facing three.