Your Manager Is Either Amazing or Terrible (And Nobody Knows Why)
The research behind why we can measure great leadership but can't define it
In my 18-year career at Amazon, I had over 20 managers. Some were exceptional, others... not so much.
While I recently wrote and article about how AI might eventually replace front-line and middle management, that future is still years away. Until then, your manager remains one of the most crucial factors in your work life—and by extension, your personal life.
Here's the paradox: Everyone can tell a good manager from a bad one, but no one can explain why. Every positive trait has an equal and opposite interpretation:
"My manager is really engaged and close to the details" → "My manager micromanages everything"
"My manager trusts me to act independently" → "My manager is absent when I need them"
"My manager is really friendly" → "This manager thinks we're friends instead of doing their job"
"My manager gives direct feedback" → "My manager is unnecessarily harsh and critical"
"My manager protects the team from distractions" → "My manager keeps us in the dark about important things"
The Invisible Work of Management
Technical managers and individual contributors (ICs) operate in fundamentally different realms. While ICs focus on direct outputs, managers juggle a complex web of responsibilities:
Hiring and growing their direct reports
Making technical decisions (often in partnership with senior team members)
Setting team vision and direction
Managing up (keeping their bosses informed) and down (cascading information)
Evaluating and guiding IC performance
It's mostly invisible glue work—vital but hard to measure. And there lies the challenge: How do you evaluate something you can't clearly define?
What the Research Says
A fascinating study by Kelly Shue, Alan Benson, and Danielle Li analyzed data from a major tech company with nearly 24,000 workers and 2,000 managers. Their findings were striking:
The Multiplier Effect: Replacing a bottom 10% manager with a top 10% manager has the same impact as adding another worker to a 9-person team—a 10%+ productivity boost.
Natural Selection: The worst managers (bottom 10%) are twice as likely to leave the company as those in the top 90%.
Talent Magnetism: Better managers tend to stay with the company longer.
Amplification: Good managers have an even greater positive effect on high-performing employees than on others.
But here's where it gets interesting: While we can measure these outcomes, we still can't identify what makes these managers better in the first place.
The Survey Problem
Companies try to gauge management effectiveness through employee surveys with questions like:
"Does your boss generate a positive attitude in the team?"
"Is your manager someone you can trust?"
"Does your manager provide guidance on improving your performance?"
During my time at Amazon, we answered these types of questions daily at login. I despised them, and the research suggests I was right to be skeptical.
Steve Tadelis and Mitchell Hoffman's research found no correlation between subjective manager evaluations and concrete outcomes like increased income, promotions, or patent applications. The only reliable indicator? Employee retention.
Making Informed Choices
So what can you do with this information? Whether you're considering becoming a manager or evaluating your current one, focus on two key metrics:
Stability: How long has the manager been with their team? Frequent moves often correlate with lower effectiveness.
Talent Retention: What's the attrition rate among high performers? Great managers create "talent sinks"—teams where the best people stay and thrive.
My own experience validates this. While I had many managers at Amazon, my last one, who I worked with for five years, was exceptional. He’s still leading the same team today, surrounded by high-performing individuals who choose to stay.
The Bottom Line
Management effectiveness might be hard to define, but its impact is undeniable. The best managers create environments where talented people want to stay and grow. While we may never crack the code of what makes a great manager, we can at least identify where they tend to cluster.
What's your experience? What qualities make a manager truly effective in your eyes? Share your thoughts in the comments below.
As a current AWS employee, you're missing one category. There are managers who succeeded and still are terrible managers, the "Yes" Manager. Someone who does little to uplevel others, micromanages, and drives away talented engineers. These managers align themselves with Directors and VPs by creating a mentor/mentee relationship and only say "yes" to senior and executive leadership. They care about only their next promotion and position. I talk to former team members who have dealt and are dealing with this type of manager. Yet, that manager is still around despite significant turnover. Why? Because of the team members who desire promotion or the carrot at the next vesting date are still accomplishing goals and tasks. There is a ton of in between. Not just Amazing or Terrible.
I find it humorous that I can relate to all 5 of your traits and been told from above, not just below, in consulting roles the alter interpretation.
I find the below-average devs and perhaps managers are the ones who often “misinterpret” the traits, it seems especially a group that is good at BSing and suddenly they can’t BS they turn to complaining if you dive into details. Of course, lots of extremely good people also want a manager out of their way to get work done, so by no means am I saying this is majority of people.
I wonder if there’s a degree of confidence or optimism that weighs how people see the traits?