Join · $1,199
Leadership Jul 1, 2026 · 5 min read

The Four Mistakes New Engineering Managers Make (And What They Cost)

You got promoted because you were the best engineer. That is the worst reason there is to make someone a manager, and the most common. Here are the four that follow.

Ryan Murphy
Ryan Murphy / Founder, EM Accelerator
Newsletter

New essays for engineering managers, delivered when they drop. No spam, unsubscribe in one click.

You got promoted because you were the best engineer on the team. That is the worst reason there is to make someone a manager. It is also the most common.

The job you were good at and the job you just got handed have almost nothing in common. You find that out live, on real people, one mistake at a time.

Each one looks obvious written down, which is exactly why you make it anyway. In the moment they do not arrive labelled as mistakes. They arrive as reasonable choices under time pressure, and you pick the wrong one for good reasons.

These are the four calls I built the New Manager Simulator around. It drops you into a compressed first 90 days and makes you make each one, without telling you which is the trap. Most people do not get the good ending first time. Read the four below and find out which one is yours.

You try to stay everyone’s mate

On Friday you were on the team. On Monday you run it. The friendships are real, so you soften the hard conversations and let a few things slide. Or you overcorrect, go cold, and start acting like someone you are not to prove you are serious now.

You are not their mate any more. That ended the day you got the title, whether you have admitted it to yourself or not. Pretend otherwise and you find out the hard way, the first time you have to give one of them a review they did not want to hear.

The friend who takes the piss out of your first call in standup is not being malicious, they just have not updated either. Laugh it off to keep the peace and you have taught the whole room that your decisions are optional.

Say it out loud early, before a slipped deadline or a promotion they did not get says it for you.

You fix it yourself

The release you inherited is slipping. You can ship the hard part faster than anyone on the team, because until recently you were the person who did. So you jump in. It feels like helping.

It is the most natural thing in the world for someone who got the job by being good at the work, and it is how good engineers burn themselves out as managers. Do it, and you make yourself the bottleneck while the team never has to grow.

Shipping the release stopped being your job the day you took the title. Every deadline you save by doing it yourself teaches the team it cannot ship without you. You are not helping them, you are making yourself impossible to promote.

So you sit on your hands and watch someone take three days over something you could have done in an afternoon. That discomfort is the job now. Swallow it and the team gets a notch better at working without you. Give in to it and you stay the bottleneck while they stay dependent on you.

You miss the person going quiet

Your best engineer’s output is still fine. But they have stopped pushing back in code reviews, and they have turned down two pieces of work they would normally have jumped at. Nothing trips an alarm, because nothing is technically wrong.

You think you would notice if your best person was on their way out. You would not. You will find out the day they hand you the resignation, you will call it a surprise, and it will not have been one. They went quiet weeks ago. You were reading their tickets, not their face.

When your best engineer stops arguing with you, that is not agreement. It is the sound of someone deciding you are not worth the disagreement. The one who used to fight you in reviews and now approves everything without a word has not come round to your thinking. They have checked out, and the silent approvals are the proof.

By the time the output drops, they have already gone. The decision got made in the silence you did not clock, back when everything still looked fine.

You get the underperformance call wrong

Your manager wants you to put someone on a formal plan. You think coaching would turn them around. So you wait, and you tell yourself you are giving them a fair shot.

You are not giving them a chance, you are hoping they quit so you never have to be the one to do it. Waiting is not kindness. It is you handing the hard decision to time and praying it gets made for you, while the person gets no honest signal and your manager watches you flinch.

While you wait, the rest of the team is watching. They knew who was coasting long before you did, and your delay tells them you either cannot see it or will not act. Both make your good people start updating their own reasons to stay. You think you are protecting one person. Everyone else is drawing conclusions about you.

Decide what you actually believe. If they can turn it around, coach them and mean it. If they cannot, stop calling the delay compassion.

Final Thoughts

None of these four mean you would be a bad manager. They are the normal mistakes of a hard job, made by people who were good at a different one. Seeing them coming is what lets you make the call on purpose instead of by accident.

Dr Claire Knight (Director of Engineering at n8n, previously GitHub and Netlify) walked me through what actually goes wrong in a new manager’s first 90 days, if you want the same ground from someone who has run it for real.

Play the simulator and find out which one you make when the pressure is on.

Ryan Murphy

Ryan Murphy

Nearly 20 years of building and leading engineering teams. Founder of EM Accelerator - the premium training platform for software engineering managers.

Keep reading

Stay sharp

Get the latest issues
in your inbox

Scripts and frameworks for engineering managers who want to lead well. Not just survive.

We will never share your email. Unsubscribe in one click.