One of the best engineers I ever managed came to me in a one-on-one and said
"I want to try something different."
She wasn't unhappy. She wasn't burned out. She wasn't looking to leave. She'd been on the same team for two years, owned a critical piece of the platform, and had gotten so good at her domain that the problems stopped being problems. They were just patterns she'd already solved.
She wanted to move to a different team. Different problem space. Different stack. Not because she was running from something. Because she'd stopped growing and she was honest enough to say it.
I knew what the easy answer was. I'd heard other leaders give it dozens of times.
"You're too critical. We need you right here."
The Compliment That's Actually a Cage
That sentence sounds like praise. It sounds like recognition. An engineer hears it and thinks "they value my work. They see how important I am."
But listen to what it actually says. Your growth matters less than our convenience. We've built a dependency on you and we'd rather keep that dependency than invest in removing it. The thing that makes you valuable to us right now is more important than the thing that would make you valuable to yourself long term.
I've heard this sentence deployed in planning meetings, skip levels, and hallway conversations. Always with the same tone. Warm. Appreciative. Final.
And it works. For about six months. The engineer feels seen. They recommit. They tell themselves the move can wait.
Then the flatness settles back in. The same patterns. The same problems. The same domain they've already mastered. I wrote about what that flatness feels like from the inside. Except now there's something new underneath it. Resentment. Because they asked for growth and got a compliment instead.
The 18-Month Clock
Here's what I've observed across every team I've led and every org I've been part of. When a strong engineer asks to move and gets blocked, an invisible clock starts ticking.
They don't leave immediately. That's what makes it so easy for leadership to think the block worked. The engineer stays. They keep delivering. For a while, it even looks like the conversation never happened.
But they've started looking. Not aggressively. Not with a recruiter on speed dial. Just ... noticing. Noticing when a friend at another company describes an interesting problem. Noticing when a recruiter's message lands differently than it used to. Noticing that the curiosity they wanted to feed internally is now looking for food somewhere else.
Within 18 months, they're gone. I've watched this pattern enough times that I'd bet my own credibility on the timeline. The engineer who gets blocked doesn't become more loyal. They become more certain that their growth isn't a priority here.
And when they leave, they don't just take their skills. They take every piece of undocumented context, every relationship, every shortcut they built over years that nobody else knows about. The "too critical to move" engineer becomes the "too critical to lose" engineer. And leadership acts surprised, as if the resignation came out of nowhere.
It didn't come out of nowhere. It came out of a one-on-one where someone asked to grow and was told to stay.
Why Leaders Block Moves
Before I get self-righteous about this, I need to be honest. I understand the pull.
When she asked, my first reaction wasn't strategic. It was a gut punch. Not because I didn't know it was the right call. Because change was about to happen to a team I'd built and grown close to. You build something with people. You invest in the relationships. You watch the team become a unit. And then someone says "I'm ready for something else" and your body reacts before your brain does.
That's the oof moment. Every leader who's built a real team knows it.
But here's what I've learned about that moment. If you let the oof make the decision, you're not leading the person. You're leading them toward you. And the worst version of that is when you convince someone to stay and the golden handcuffs go right back on. You can try to give them a new vision. You can try to ramp up the challenge. But you can't manufacture the growth someone needs when the ceiling is the role itself. And if you try, you're just delaying the same conversation by six months while the resentment builds underneath.
The leaders who block moves aren't making a strategic call. They're flinching. And the engineer pays for it.
I wrote about this in my post on bus factor. The engineer who's too critical to move is the same engineer whose absence would paralyze the team. That's not a retention problem. That's a structural problem you're using a person to paper over.
What I Actually Do
When that engineer came to me and asked to move, here's what went through my head. In order.
First. If she's asking, she's already halfway out the door mentally. Not out of the company. Out of the role. The curiosity has already shifted. The question isn't whether she'll move on from this work. The question is whether she moves on inside the org or outside it.
Second. I have maybe a two-month window to make this work internally before she starts solving the problem herself. Engineers are problem solvers. If the problem is "I'm not growing here," they will find a solution. You either help them find it inside the company or they find it at a different one.
So here's what I did. I told her yes. Not "let me think about it." Not "let's revisit next quarter." Yes.
And within the month, she was on a new squad.
Here's the part that matters. I didn't have to scramble. There was no emergency documentation sprint. No panicked knowledge transfer. No six weeks of catching up on work I should have done earlier. Because we'd been doing it along the way. The runbooks existed. The context was distributed. The team was operating as a team, not as a collection of individuals dependent on one person.
What Happened
She Moved
She took on a problem she'd never touched before. Struggled for the first month in exactly the way that produces growth.
Three months later she was contributing at a level that surprised everyone, including her. Six months later she was leading a workstream. The curiosity that had gone flat on my team was fully alive on the new one.
And my team? It kept shipping. The two engineers who'd been sharing her context stepped up in ways they never would have if she'd stayed. The space she left didn't stay empty. Someone grew into it. That's the same pattern I described in my post on stepping back from being essential.
She stayed at the company for another three years. Three years of high engagement, increasing scope, and compounding impact. All because someone said yes when she asked to grow.
If I'd said "you're too critical," she'd have been gone in 18 months. I'm certain of it.
When There's Nowhere Internal to Go
That story had a clean ending because there was a team to move to. It doesn't always work that way.
I've had engineers come to me when I didn't have a place for them. No adjacent team that fit. No open role that matched what they needed. The growth they were looking for didn't exist inside my org.
So I became their biggest advocate for leaving.
Not passively. Not "I understand if you need to explore other options." Actively. Helping them think through what they wanted. Making introductions where I could. Being honest that the next chapter for them wasn't here and that staying would cost them more than leaving.
That's the version of this conversation that actually tests whether you mean it. Saying yes to an internal move is relatively easy. You keep the person in the company. You can tell yourself it's still a win.
Helping someone leave entirely is the version that proves your leadership isn't conditional on them staying useful to you. I wrote about this in my post on mentoring people, not engineers. Your job was never to retain engineers. Your job is to develop people. Some of them will stay. Some will leave. The ones who leave because you helped them find their path will send better people your way than the ones who stayed because they felt trapped.
The Question to Ask Yourself
If you're a leader and someone on your team asks to move, ask yourself one question before you respond.
Am I about to say no because it's the right thing for them, or because it's the easier thing for me?
If the answer is the second one, say yes.
Or say no. Call it a compliment. And start the 18-month clock.
I write daily about engineering leadership at jonoherrington.com.
Top comments (0)