The High Performer Who Was Secretly Killing My Company
This story appears in the
May 2026
issue of
Entrepreneur.
Subscribe »
A strange thing happened at my company: Three people, from the same team, all took stress-related sick leave within six months.
Then, a client came to meet with that team. The senior project manager, Mark, snapped at our own production team — undermining their work in front of the client. When Mark’s manager pulled him aside, Mark didn’t flinch. “You should be glad I work here,” he said. “We only have these customers because of me.”
I should have fired Mark that day. But I didn’t. Which made me the problem.
Toxic workers are a huge problem. A report from Harvard Business School found that when a company gets rid of a toxic worker, it’s more than twice as valuable as hiring a superstar. Toxic workers cause 78% of coworkers to decrease their commitment to the company.
So why can’t managers just get rid of them? Here’s my personal experience: It’s a combination of fear and complacency.
Mark was our highest performer; he’d delivered major client engagements and had superior technical knowledge. At first, he was so good that nobody cared about his rough edges. Then he started hoarding client relationships, making himself the only person who understood key accounts. By the time he was openly undermining leadership, I’d already normalized it. I’d heard myself say, “That’s just Mark.”
But for the benefit of my company, I finally realized: Mark had to go.
After I fired him, reactions from our team were split. His mentees, who’d only seen his brilliance, marched into HR to insist we’d fired our best performer. A second group of people came quietly, saying they’d been afraid to speak up for months. (One project manager told me he’d been updating his resume, because he just couldn’t deal with Mark anymore.) That reaction haunts me, because their silence was my fault.
I’d clearly built a system that measured what people produced…but ignored how they treated each other. As a result, we were effectively rewarding bad behavior.
That’s when I realized: Removing Mark wasn’t enough. We needed to fix our system.
First, in every performance review, we started weighing behavior metrics equally alongside performance metrics. Did you share knowledge or hoard it? Did colleagues seek you out or avoid you? These became hard requirements. Now, every manager can use them as a tool to address behavior when it shows up — and not just during a crisis.
Next, we ran quarterly sessions to teach managers and HR what good behavior looks like, and what bad behavior we won’t tolerate.
Once implemented, I started to see true changes. The good workers, who’d watched bad behavior get excused for years, started stepping up.
Here was a great example: In a meeting a few months later, a junior employee questioned a project manager’s decision. In response, the project manager paused and said, “Good catch, let’s rethink this.”
That would have never happened with Mark in the room. Back then, that junior employee would have stayed quiet. The project would have been worse as a result.
Stress-related sick leave vanished. Teams that hadn’t worked together in months started solving problems on their own — no program, no initiative.
Now, whenever I talk with leaders, I tell them this story and ask them: How many of your best people are ready to leave you because of a toxic worker? How many have already left? Because if behavior doesn’t carry the same weight as results in your organization, you’re not measuring performance. You’re measuring half of it and hoping the rest takes care of itself.
A strange thing happened at my company: Three people, from the same team, all took stress-related sick leave within six months.
Then, a client came to meet with that team. The senior project manager, Mark, snapped at our own production team — undermining their work in front of the client. When Mark’s manager pulled him aside, Mark didn’t flinch. “You should be glad I work here,” he said. “We only have these customers because of me.”
I should have fired Mark that day. But I didn’t. Which made me the problem.
Daniel Halvarsson
•
Researcher in economics