Stop Trying to Promote My Best Engineers

There has always been a disconnect between the hiring process and finding the best engineers. But when we somehow find them, the career ladder ensures that they don't remain in that position of strength. An incompetent company might create the conditions for engineers to leave for better jobs. A generous company will apply the Peter Principle and promote engineers to their level of incompetence. Either way, the best engineers never remain in that position of strength.

The Interview Theater

How do you recognize a great engineer? Is it someone who aces all the leetcode during the interview process? Is it someone who is a great communicator? Or is it someone who went to an elite university? The processes we currently have in place can only determine so much. Candidates have limited time to audition for the role they're applying for. Over the span of a few interviews, they're supposed to convey the experience from all their past work, show that they know how to do the job, and also talk about their greatest weakness. It's a performance that some people know how to game.

AI-powered hiring tools haven't changed this problem. They don't magically give you better candidates. You're still sifting through the same pool, just with fancier filters. The disconnect between interview performance and actual job performance remains.

A few years back, I interviewed someone I'll call the greatest communicator I've ever seen. It was for a web engineer position on another team. He seemed to understand the front end, the backend, and the jargon of the job. But what impressed me most was how he broke down each problem I posed into small parts and thoroughly resolved each one. It was as if he was creating Jira tickets in real time and writing documentation along the way before the task was even completed. I gave the thumbs up and he was hired.

A couple of months later, I remembered him. I searched for his name in the directory and learned that he was let go. "Why?" I asked around. The answer was "he was pretty bad, couldn't complete a single task."

Yet he was able to pass the job interview. The inverse also happens. You take a chance on someone who seemed merely adequate during interviews, and somehow they turn into one of your best engineers. I've often found myself in teams where I have zero doubts about the ability of my teammates. But then, as the end of the year approaches, the inevitable discussion turns to promotion.

The Promotion Trap

It's actually much easier to identify a great engineer on the job than in an interview. You see their work, their growth, their impact. And when you finally have that clarity, when you know without a doubt that this person excels at what they do, the system insists you move them away from it.

When you are good at your job, the logical step for a manager is to reward you with a promotion, moving you away from the job you are actually good at. That's the Peter Principle in action. Managers believe their only tool for compensation is moving you up and down the ladder. A great developer gets promoted to senior developer, then to tech lead, then to manager. At each step, we strip away more of what made them valuable in the first place.

The underlying assumption is that a great engineer will nurture a team into great engineers. But teaching and applying a skill are two distinct occupations. You may be great at one, but terrible at the other. My instinct is to help great engineers continue to grow in their expertise, not switch them to a role where they're no longer competent. It's important not to throw away all their knowledge and put them in a position of authority where they can't exercise their skill.

Yet many employees themselves don't know what the next step up should be. They see "senior" or "lead" or "manager" as the only path forward, not because they want those responsibilities, but because that's the only way to get recognition and compensation.

Rethinking the Ladder

What if we stopped thinking about career advancement as climbing a ladder? What if the goal wasn't always upward, but deeper?

The traditional career ladder assumes that everyone wants to eventually stop doing technical work. It assumes that the best reward for mastering a craft is to stop practicing it. But some of the best engineers I've worked with have no interest in management. They want to write code, solve hard problems, and mentor others without taking on hiring, performance reviews, and budget planning.

We need to normalize horizontal growth. This means creating paths where engineers can gain expertise, take on more complex challenges, and yes, earn more money, without leaving their position of strength. It means recognizing that a senior engineer who has been writing excellent code for ten years is not "stuck" or "lacking ambition." They're mastering their craft.

It also means changing how we structure compensation. If the only way to give someone a significant raise is to promote them, then we've built a system that punishes expertise. Companies should be able to pay top-tier compensation for top-tier individual contributors, not just managers.

The irony is that we struggle to identify great engineers in interviews, yet when we finally find them on the job, we immediately try to change what they do. We should be asking ourselves, if this person is exceptional at their current role, why is our first instinct to move them? Maybe the answer isn't to promote them out of their position of strength, but to let them get even better at what they already do exceptionally well.

After all, if interviews can't reliably identify great engineers, shouldn't we do everything possible to keep them exactly where they are when we finally find them?

---