How to Give Feedback

How to Give Feedback

31 December 2021

Giving Feedback is one of the most important tools available to a manager. It’s also one of the most under-used. The majority of employees say they don’t get enough feedback, and the reason for this is that the majority of managers say they’re uncomfortable giving it.

Unfortunately, on the flipside of this, the managers that are comfortable giving it often give it way too often and at the wrong times. There are few things more infuriating than a manager who constantly tells you things you already know, so striking the balance is important.

In this guide we’ll help you do that by breaking down why, when, and how to give feedback.

Why do you give Feedback?

This might seem obvious, but not understanding this is one of the most common problems when it comes to giving Feedback.

There are only two reasons to give feedback:

  • You want to change a behaviour
  • You want to encourage a behaviour

A lot of managers fall into the trap of thinking things that are actually just criticism or complaining are feedback. The truth is that it’s only feedback if it’s intended to change a behaviour.

The moment you consider giving someone feedback, stop and consider two things:

  • Do they know what I’m going to tell them already?
  • Will telling them make anything better?

Let’s look at a quick example. You have asked one of your people, Mark, to create and send a newsletter to all your key clients. He did this promptly, but after sending it he notices that it contains quite a few typos, and he’s come to you to tell you about it. He knows he made a mistake, and he’s annoyed with himself. He normally gets them proofread before he sends them, it just slipped his mind this time.

So, should we give Mark feedback?

Let’s look at it logically.

Mark is already aware he made a mistake. He’s upset with himself about it. He already said he usually proofreads them so there’s nothing new you can tell him, He obviously knows that it’s bad to send an email with typos in it to clients, or he wouldn’t be there telling you about it. And feedback you give him here will not give him any information he doesn’t already have.

Will telling him anything better?

In this case, giving Mark feedback wouldn’t just not make things better, it would make things worse. Mark came to you and told you that he made a mistake, if you explain to him why the mistake he made is bad, it’s not going to make him any less likely to repeat it, and it’s almost certainly going to irritate him. More likely than not when the two of you separate he’s going to be muttering unpleasant things about you. He’ll be annoyed with himself for making a mistake, and when you tell him what he already knew, he’ll direct that anger at you.

Now, let’s imagine a slightly different version of this scenario. The same thing has happened, but this time Mark didn’t notice, someone else did and they told him about it. You overheard the conversation, and Mark didn’t seem to recognise that it was a big deal. However, it is a big deal to you. You don’t want typos in your newsletters.

  • Do they know what I’m going to tell them already?
  • Will telling them make anything better?

In this scenario, there is absolutely a need to give Mark feedback. You don’t know that he realises it’s important not to repeat the mistake, and therefore giving him feedback might mean it’s less likely that he will.

When do you give Feedback?

So, we know why we give feedback now. The next question is when.

Do you do it right away?

What I’m going to say sounds a little messed up, but stick with me. There’s a rule when you’re training a dog about when to scold or praise. Very simply, you have to do it the moment the dog does the thing. If you wait even a few seconds, they won’t make the connection.

Now, obviously (hopefully?) your staff are smarter than dogs, but it’s still a good rule to follow. The less time there is between the event and the feedback, the more effective it’s going to be.

In my experience, if you observe a behaviour and you don’t say anything at the time, saying something later will probably illicit a bad response. Most people will think “why didn’t you say something!” or some variation of that.

That’s not to say you can’t give feedback on something that happened in the past that you have only just found out about. In those scenarios, just avoid seeing something and then waiting to talk about it another time. See something, say something, as the airport tells us all the time.

Do we wait to do it in private?

Giving feedback in public should never be a problem. In fact, if you feel like you can’t give the feedback in public, it’s a good indication that you’re not delivering it well. Feedback should never be a telling off. It’s not a negative experience, so it shouldn’t be something you need to do in private. In fact, if you make a big deal of doing it in private, you’re tacitly sending the message that they’ve done something they should be embarrassed about. There’s no need to give feedback in private.

Do we do it over Slack or email?

No. Simple as that. Never. They need to hear your voice, and you need to be able to discuss things. Give feedback in person or on a call or hangout. But never over written messages.

How do you give feedback?

So, remember the two reasons you would give someone feedback?

  • You want to change a behaviour
  • You want to encourage a behaviour

These two situations require slightly different approaches to giving feedback.

Adjusting Feedback

When we want to change a behaviour, we use adjusting feedback. The model for adjusting feedback has the following three steps:

  1. Ask if you can give the feedback
  2. Explain the situation, behaviour, and impact
  3. Agree a different future behaviour


This step is so often missed off, but it’s the most important one. Taking this step will give you a far better chance of achieving the thing you want, which is changing the behaviour.

Let’s go back to our situation with Mark. You just overheard him being told about his mistake, and you noticed that he didn’t seem that bothered by it.

Imagine you just go over and say “Mark, sending out an email full of typos makes us look bad”. There’s nothing too bad about that, but you are coming a bit hot. There’s definitely scope for him to feel a bit put upon. It’s not that there’s anything particularly wrong with doing it this way, but there’s still a much better approach.

“Hey Mark, can I give you some feedback about that email?”

You’ve not just jumped on him, you’ve asked if you can talk about it. It gives him time to switch gears, and by the time he answers, you know he’s ready to listen.

One of the reasons I’ve heard some people use for not including this step is that you shouldn’t give people a choice. What if they say no - won’t that undermine you?

It’s actually the complete opposite. When we make mistakes we’re often embarrassed and frustrated, and that makes us more likely to get angry with other people. Especially people who are pointing out that mistake. If you just go to mark and immediately start giving him the feedback before asking him, and he’s in the kind of mood where if you had asked him he’d have said no, he might react badly. He might argue with you, or just try to brush you off. That actually is undermining. However, if you ask him if you can give him some feedback and he says no, that’s fine. Just say “No problem, we can talk about it another time.”. You instigated the exchange by giving a choice if he wanted to be in it or not, so if he takes the option not to it doesn’t undermine you at all. You’d planned for it.

And here’s the thing. You probably won’t have to even talk to him later. He would have known why you wanted to give him the feedback, and he’d have filled in the blanks without you even talking. Just the fact you wanted to talk about it was enough. Once he’s calmed down he’ll probably come to you, and if he doesn’t you can decide if it’s something you need to follow up on and talk to him once he’s in a better place.

Remember the purpose of the feedback is to change his behaviour. If he’s not emotionally ready to listen to you it won’t be successful. So asking if you can give the feedback  is always the right choice.

Situation, Behaviour & Impact

You might see this model broken down to consider these things as three different steps. I don’t think that’s helpful, as the chances are you’re going to do this in one short statement. The shorter you can make this, the better. Otherwise it can feel like a lecture.

Ok, so let’s go back to Mark. You’ve asked him if you can give him some feedback, and he’s agreed. What do you do next?

There’s a golden rule here that you absolutely must not break: don’t say anything that you haven’t observed and don’t know to be true.

For example, you can’t say to Mark:

“You sent the email out to our key client without proof reading it, and I think that makes us look unprofessional”

Whilst structurally this is good feedback, there’s a problem.

  • Do you know Mark didn’t proofread it?
  • Do you know Mark was even supposed to proofread it?
  • Do you know he even sent the email himself? He could have asked someone else to proofread it and sent it afterwards.

The only thing you know is that the email was sent, and it had typos in it. That’s what you know, so that’s what you can say.

“That email I asked you to send went out with typos in it, and I think that makes us look unprofessional”

If you stick to what’s true, there is very little chance Mark can get angry. If you assume anything, you’re going to get things wrong, and you’re going to derail the entire process.

Also notice that you don’t have to stick rigidly to the situation, behaviour, impact structure. That’s why I think of it as one step. It’s clear from that statement what happened, what Mark did, and what the result was. The more rigid you are, the less successful you’ll be

“An email was sent with typos in it. You sent the email. The result is that we look unprofessional”.

That sounds weird. The model is there to show you what information you need to convey, and you should do that in as few words as possible.

Future behaviour

This is where you’ll see the benefit of only talking about what you know to be true in the previous step.

Let’s say you’d made the assumption that Mark hadn’t proofread the email, and you’d given him the feedback like this:

“You sent the email out to our key client without proof reading it, and I think that makes us look unprofessional. Can you make sure you proofread it next time please?”

Again, structurally this feedback is acceptable. Situation, behaviour, impact and new behaviour have all been communicated. But what if you’re wrong, and he did proofread it. He’ll just say “Yeah, sure” and nothing will have been improved.

When you don’t assume you know anything, you’ll realise that you might need to ask questions before you can agree on a new behaviour. For example:

“That email I asked you to send went out with typos in it, and I think that makes us look unprofessional. What happened?”

Now, let’s say Mark explains to you that he did proofread the email, he just did it too quickly and mistakes got past him. Now you have some options?

“Are you generally good at proofreading?”

“How do you go about proofreading?”

“How much time did you spend on proofreading?”

If you haven’t assumed you know why the thing happened to begin with, you can have a productive conversation with Mark to get to the bottom of it.

“For important group emails, would it be helpful to get another person to proofread them before they go out?”

“Do you think we need to do something differently to make sure it doesn’t happen again?”

“Could you get me to group approve emails before sending them in future please?”

Agreeing a new behaviour can almost always be done with questions. You can ask Mark if he thinks anything needs to change, and again, it’s fine if he says it doesn’t. If the answer is “Nope, it was just a one off mistake, it will be fine next time” that’s great. He’s promised he’s on top of it. Or you can agree on a new behaviour together. But either way, the process of giving feedback has not been in any way confrontational or negative. You asked for a conversation about a thing that happened, and you agreed what to do to try to stop it happening again in the future.

Affirming Feedback

It’s easy to overlook affirming feedback. but it’s extremely important. Just as much as you want to make sure people change the behaviours that you don’t want, you also want to make sure the ones you do want happen more often.

You can do this with affirming Feedback.

Imagine you have a new starter, and he’s in that typical nervous new starter mindset where he isn’t talking much and is a bit nervous still. Work is carrying on as usual and conversations are kind og going on around him, but he’s not in any of them. Mark works in a different team entirely. He notices that the new guy is nervous, and he goes over to say hello and invites him to lunch.

You like that behaviour. That’s how you’d like people to treat new starters here, and you wish other people had done it too. If you’re a controlling manager, you’ll make a rule that directs people to go and be nice to new starters, but that’s a terrible choice. There’s a much simpler and more effective way to handle it.

Thank Mark.

The next time you’re in a situation where the new guy isn’t around, go over to Mark and say “Hey, I really appreciate you taking the time to welcome the new guy. It’s difficult finding good people and it’s always worrying when they’re not settling in. I felt much better knowing he was being taken care of. Thanks!”

And if other people hear you say it, that’s a bonus. You’ve given them adjusting feedback by proxy!

Notice that the middle of the model is very similar to adjusting feedback:

  • Situation
  • Behaviour
  • Impact

The only difference with affirming feedback is that the first and last steps are just replaced with a thank you. You don’t need to ask, and you don’t need a new behaviour. Just say thank you, and explain why what they did mattered to you.

Remember, every management tool is precisely that - a tool. Nothing more, nothing less. You don't have to use the tool just because you have access to it. Over time, you'll build your relationships to the point where you can instinctively tell how much of the model you need to use when you give Feedback. My CTO and I have reached a point where we can just say "You know that thing you did...well it sucked" and that's all we need to do. We know and trust each other enough that the thinking behind the model is implied.

More Guides