Skip to main content
Paula Farrugia
  • Leading
  • Feedback
  • Communication

Early Enough to Use: How I Give Feedback as an Engineering Manager

7 min read


What I've learnt about giving feedback as an engineering manager, and what I do now.

A junior developer on my team wasn't meeting expectations. I wasn't direct enough in our 1:1s, and I never made clear how serious it was. Then came his review. I was honest in writing, and the low score that came with it didn't match what he'd heard from me in our 1:1s.

He was really upset. Eventually he was let go because he couldn't keep up with the performance expectations. I accept some of the blame for that, because my unclear feedback is what led to that point.

This was in my first role as a lead. Everything I practise today comes from it. Here is what I do now.

Closeness isn't safety

Psychological safety is what makes feedback possible, but it isn't the same as being close. Safety means a person can hear something hard from me and trust that I mean well. Comfort means nobody says the hard thing.

With the junior developer, I mistook one for the other. Now I make sure I'm never so close to someone that I avoid giving them feedback, or that they take it as a suggestion and don't understand how serious it is. I assume best intent, even when someone has made a mistake, and I treat it as if they were doing their best.

I also look after the moment itself. Before critical feedback, I think about whether the person is in the right headspace to hear it. If they're already upset before the meeting, I ask whether we should meet another time. I still have the conversation, but I give them space first. I never give growth or critical feedback in front of a group. Praise is different: I'm happy to give it in front of the team, as long as I'm aware of how the person will react. I also don't give critical feedback in chat, because tone gets lost and the message doesn't land. Growth feedback can go in a direct message, but if there's a lot of back and forth, I switch to a conversation.

Start early, even on small things

Feedback only works if the person can hear it when you give it, and still has time to act on it. It should be a habit before it becomes a conversation people dread. I start early, even on small things, so people get used to hearing it. If I notice someone could be better at prioritising their work or at estimates, I say it then. If I wait, it slowly turns into not meeting expectations, and the person has never heard it from me. Giving feedback in the moment keeps that from happening, and it keeps each piece of feedback small.

Small feedback means I can give one thing at a time. If I pile on too much at once, the feedback gets muddy and people only hear parts of it. Sometimes it also means there are simply too many things for the person to change, and that tells me something about whether the team is the right fit for them.

Early feedback covers what's going well too. I give positive reinforcement constantly, because people want to know what's going well and where they stand. "No news is good news" doesn't work for most people. When someone does a good job, I send them a message right away, along the lines of: "This is really great, I love to see more of this, keep doing it. I can see how it's showing up and helping the team." I also show appreciation publicly when people do things right. It recognises them in front of the team.

I encourage my team to do the same among themselves, and it's nice to see. People send each other kudos at random, because my team and I built a culture of praise. They also aren't shy about giving each other constructive criticism, because the team feels safe enough and hears each other out.

Be clear, and keep it about the work

Early feedback only helps if it's clear. Being clear is kinder than being comfortable. "You're not meeting expectations on this" and "I think you could improve on that" are two very different messages, and the person has to know which one I mean.

I practise before the conversation. I write down what I want to say. I don't read it out as a script, but writing it helps me think about how the message will be interpreted and how to make it clear. I focus on the goal: that we set the expectations, and then the two of us work on them together.

In the conversation, I start from what I've noticed, and I put my own feelings on the table. For example: "I've noticed you're missing some timelines lately and I'm worried about it. Is that something you're noticing too?" A question like that shows them this is a conversation, and that they're being heard. It also opens up things I might not know, such as someone struggling with something outside work. I can still say they're not meeting my expectations, but I'll change how I approach it. If someone is already dealing with something big, I don't pile on.

I keep it about the work and the outcome, not the person. I focus on being direct without being harsh. Instead of saying someone isn't good at something, I say that what they're doing isn't helpful, or that I'm seeing they're not able to deliver something. I use concrete examples, recent ones where possible. People can get hung up on the example, though, so I keep coming back to the common theme.

Back them up, then follow up

Feedback isn't finished when the conversation ends, and the person shouldn't be left to act on it alone. Whether it's something they want to work on or something they don't, they need to know I'm in it with them.

The clearest example of the first is a developer who wanted to become a senior but wasn't showing up the way a senior would. I told him he needed to speak up more and take more ownership of architecture designs. Then I worked on it with him every month until he got the role.

What made the difference was that I didn't only give him the feedback. He was shy about speaking up, so I told him that every time he did, I would back him up, or step in and facilitate the conversation if he needed it. He could try what I was asking, knowing I had his back.

The second is a developer who wasn't very good at communicating. Instead of calling him out, I made an agreement with him: after every call, we would check in, and I would give him my feedback on how he could have done better. It gave us a "we're in it together" mentality, and he improved.

Backing someone up also means following up after a critical conversation. I give the person time to digest it, because they might be defensive in the moment, and I do a follow-up conversation a week after. I also send the feedback in writing, usually the next day, to emphasise the expectations and make sure we're both using the same language about what needs to be fixed. I ask them to outline a plan for meeting the expectations, and I outline everything I'll do to help as their manager, apart from the expectations themselves.

Once they've had a week or two to work on it, I check in. If I think they're improving, I tell them, with some positive reinforcement or more feedback on their approach. It gives people a signal that they're moving in the right direction.

When it doesn't improve

Sometimes feedback alone isn't enough, and I have to be ready for that. If things aren't improving, I ask myself whether I should keep giving feedback, or whether this team isn't the right fit for them. As a manager, I think about a timeline. It doesn't have to be "I need to see this in a week", but I keep it in mind.

With the junior developer, it came to that. I can't say clearer feedback would have changed it.

Early enough to use

What I wrote in the junior developer's review was honest. What was missing was the same message, said clearly and said early, while he still had time to act on it. That's what I try to do now.

I still care about the people I manage. I just don't let that stop me from saying the hard thing. I want my team to feel safe enough to hear it, and to hear it early enough to use it.

Struggling with giving feedback?