What should I do when someone disagrees with my feedback?

Perhaps disagreement is information, not resistance

Few moments in a difficult feedback conversation feel as uncomfortable as the one that follows “I don’t think that’s fair.” I have seen capable leaders respond to it in very different ways. Some back away quickly, softening the message until little of it remains. Some dig in and produce more examples. Some quietly conclude that the person is resistant to feedback and stop listening. All three responses are understandable, and none of them is particularly useful.

When an employee disagrees with feedback, at least two quite different things might be happening. The disagreement may be telling me something I need to know: that I have misread what happened, that the person had context I lacked, or that the standard I am applying has not been applied evenly. Or it may be a first reaction to a message that is hard to hear, one that softens or shifts once the person has had time to think. Those call for different responses, and I cannot tell which I am dealing with until I have found out what the disagreement actually is.

I would keep the phrase feedback resistance for something more specific: a pattern in which someone repeatedly declines to engage with well evidenced feedback over time. A single piece of pushback on feedback does not amount to that. The risk of reaching for the label too early is one I explored in my article on giving feedback without making someone defensive. Once I have decided that someone is resistant, everything they say next becomes further proof.

Work out what they are actually disagreeing with

When an employee says feedback is unfair, that is rarely a single claim. It usually bundles together several different objections, and responding to the wrong one is a reliable way to have a long and unproductive conversation. I find it helpful to ask myself which of these they might mean.

The last two rows deserve particular care. Sometimes the claim of unfairness is well founded. If the expectation was never made clear, or if the same behaviour is tolerated in colleagues, the employee is raising a legitimate point about the process rather than resisting the message. I would rather discover that in the conversation than have them carry it out of the room.

The simplest way to find out which objection I am facing is to ask, and to mean it. Which part of this feels unfair? is more useful than Why are you being defensive?, and considerably more useful than a second example. It also tells me whether the disagreement is about the facts, the meaning, the consequences or something else.

One pattern to watch for is the conversation moving sideways. The employee raises a concern about something I did, or about how the team is managed, and the original feedback disappears from view. Some of what they raise may matter a good deal. I try to acknowledge it, agree to come back to it, and then return to the original point so that neither topic gets lost.

Test the evidence, including your own

If someone disputes the facts, the most useful question is also the most uncomfortable one: how good is my evidence? Before I defend my position, I want to know how much of what I said comes from something I saw directly, how much from something I was told, and how much from a general impression built up over time. The BOOST feedback model asks whether feedback is observed, objective and specific, and those questions work well as a stress test here. If my evidence turns out to be one occasion, a secondhand report or a feeling, I would rather recognise that early and say so.

Testing evidence is not the same as arguing about it. It means putting what I have on the table and inviting the other person to add to it, challenge it or correct it. The Examples stage of the CEDAR feedback model is designed for exactly this: building a shared picture of what happened before anyone argues about what it means. When someone says “I don’t think I did that,” I do not repeat the claim more firmly. I set out specifically what I noticed and ask how it looks from where they sit. If I find that much of my evidence is really interpretation, my article on giving feedback about behaviour rather than personality looks at that shift in more detail.

Imagine I tell someone that their handover notes have been missing key information. They tell me that is not true and that they always send them. Rather than repeat myself, I ask them to show me the most recent one. We look together and find that the notes were sent, but to a distribution list that does not include the next shift. We were both partly right. The facts were more complicated than either of us had described, and the conversation moves from whether the notes exist to how the information actually reaches people.

Testing the evidence tends to lead to one of three places. The evidence holds, and I can say so and continue. The evidence partly holds, and I can acknowledge what I had wrong or incomplete while staying clear about what remains. Or the evidence does not hold, and I need to say so plainly. That last outcome is uncomfortable, but it often does more for credibility than holding a position I can no longer support. A leader who can say “I had that wrong, thank you for showing me” is usually easier to believe the next time they say “I think this is a problem.”

Separate a different perspective from a different set of facts

Some disagreements are not really about facts at all. Two people can agree completely on what was said in a meeting and still disagree about whether it was dismissive. Social psychology has long suggested that people tend to explain their own behaviour by pointing to the circumstances they were in, while explaining other people’s behaviour by pointing to the kind of people they are. The strength of that tendency varies, but it is a useful possibility to hold in mind. If it applies here, I am inclined to read a missed deadline as a pattern in the person, while the person reads it as the consequence of a situation I may not have seen. Neither of us needs to be lying.

Consider “You were dismissive of the client’s concerns” alongside “I was trying to keep us on time because the client only had thirty minutes.” Both can be sincere. The first is an interpretation of behaviour, and the second is an explanation of intent. I can accept that the intent was reasonable and still say that the effect on the client was what it was. Intent and impact are different things, and it is possible to take the first seriously without letting it cancel the second.

This is where the Diagnosis stage of CEDAR earns its place. Its purpose is to help both people explore why things are as they are, because until the cause is clear any action plan is guesswork. When an employee disagrees with feedback, Diagnosis shifts the conversation from who is right to what is going on. Anna Wildman, who created the model, suggests asking what the person thinks is behind what happened, and starting with wider circumstances such as time, resources and workload before moving to personal capability. That order can build trust, and it often surfaces context the giver did not have. It can also help me as the giver: if a constraint is affecting one person, it may be affecting others.

Decide what actually needs agreement

This is the question I think leaders most often skip. After some pushback, the natural aim becomes getting the employee to agree that I was right. That is a hard thing to achieve, and it is not always a necessary one. It helps to be deliberate about what actually needs to be agreed and what only needs to be understood.

• Usually worth reaching shared understanding on: the expectation going forward, the effect the behaviour has on the work, the team or the client where that can be evidenced, and what happens next, including when you will check in.
• Worth resolving where they are material: any disputed facts that the conversation depends on. If we cannot agree that the report was late, we cannot usefully discuss why.
• Not necessarily needing agreement: whether the employee thinks the feedback was fair, what their intentions were, how I interpret their motives and whether they like the message.

That last category matters. I do not need someone to agree that they were dismissive in order for them to understand that two colleagues stopped contributing after the exchange and that I need that to change. I need them to understand the effect and the expectation. Whether they accept my reading of why it happened can remain open.

The language of agreement also carries a risk. A nod and “Sure, I’ll do that” can look like agreement and be nothing more than a wish to end the meeting, which is the illusion of agreement I describe in my CEDAR guide. What I am looking for is understanding the person could restate in their own words, not compliance I have obtained through persistence. The BIFF approach, as I use it on the feedback models hub, helps here through its Future step, which asks me to state plainly what I expect. That keeps the expectation separate from the argument about the past.

It might sound like this: “I understand you see this differently, and I’m glad you told me. I don’t need you to agree with how I read it. What I do need is for us to be clear that client notes are in the system by the end of each day, and I’d like to check in on how that is going in two weeks.”

When emotion rises, pause rather than push

Disagreement often arrives with feeling attached. When it does, more evidence rarely helps. BIFF’s Feelings step is a useful prompt here: a genuine check on how the message is landing and whether the person can still take in what I am saying. How is this landing for you? needs to be a real question, not a technique for moving them along.

If someone is angry, I try to stay in Diagnosis rather than match their energy. Tell me why this feels unfair to you invites the same information as any other question, and the anger often points towards what matters most to them. If the conversation is getting heated, I would rather pause and agree a time to resume, perhaps after they have looked at the examples and gathered anything they want to bring. A second conversation that goes well is worth more than a first one that goes on too long.

When you still disagree

Sometimes, after a good conversation, views still differ. That is an acceptable outcome. “I understand why you see it differently, and I still think we need to address what happened” allows both things to be true. Curiosity does not require me to surrender my judgement, just as clarity does not require me to stop listening.

Managers carry authority and responsibility, and some expectations are not open to negotiation. If something is an expectation rather than an observation for discussion, I think it is important to say so plainly, since the same sentence can carry very different weight from a manager than from a peer. It also helps to follow up briefly in writing, setting out what was agreed, where you see things differently and what happens next. That reflects the Review stage of CEDAR, and it keeps an honest record that does not pretend there was more agreement than there was. Where the disagreement relates to a formal performance process, involving HR early is usually sensible.

If the same pushback on feedback returns repeatedly, even after the evidence has been examined, the question shifts from is this feedback right? to what is getting in the way? It may be trust in me or in the process, inconsistent standards, workload, skill, motivation or something I have not yet understood. Sometimes the honest answer is that a more formal conversation is needed. I would want to reach that point having genuinely tried to understand the disagreement, rather than having labelled it feedback resistance at the first sign of pushback.

What does current research suggest?

The research does not offer a script for handling feedback disagreement, but several strands help explain why it happens and what might help. As with much of the feedback literature, the evidence varies in strength, and I would treat it as a way of thinking about the conversation rather than a set of rules.

Less welcome feedback tends to be judged less accurate

Brett and Atwater (2001) studied 360 degree feedback and found that less favourable ratings were associated with beliefs that the feedback was less accurate and with more negative reactions. Both of those, in turn, were associated with seeing the feedback as less useful. The findings are correlational and concern a particular kind of feedback, so I would not generalise too far. They do suggest that a claim of inaccuracy becomes more likely as the news becomes less welcome. That is a reason to test the claim rather than either accepting or dismissing it, and to hold my own confidence a little more loosely than it may feel in the moment.

People often explain the same behaviour differently

Jones and Nisbett (1971) proposed that actors and observers tend to attribute behaviour differently, with actors pointing to the situation and observers pointing to the person. Later work has qualified how far and when this applies, but it offers a useful explanation for why a manager and an employee can look at the same event and reach quite different conclusions without either being dishonest.

Different triggers sit behind the same sentence

Stone and Heen (2014), writing mainly for people on the receiving end, distinguish between truth triggers, relationship triggers and identity triggers. The same words, “that’s unfair,” might come from any of the three: a belief that the content is wrong, a reaction to who is saying it or how it was said, or a threat to how the person sees themselves. They also describe switchtracking, where the two people in a conversation end up discussing different issues without noticing. For the giver, the practical value is in recognising that an objection may not be about the thing it appears to be about, and in staying curious about why the other person sees it differently before responding.

Credibility and shared responsibility matter

Fulham, Krueger and Cohen (2022) argue that feedback produces learning only when recipients are both receptive to it and able to discern whether it is true and useful, with barriers arising from the feedback, the giver and the recipient. Zyberaj (2026) found in a systematic review that source credibility and feedback quality were associated with how supervisory feedback was perceived and acted upon, although much of the evidence was correlational. Together, these suggest that showing my evidence, explaining my reasoning and being willing to be corrected are not only courteous. They may also influence whether the feedback is taken seriously.

Taken together, the picture points towards treating pushback as something to understand, test and respond to, rather than as a verdict on the recipient or on me.

Five questions to ask yourself when an employee disagrees with feedback

When the pushback comes, I would pause and ask myself:

1. What exactly are they disputing? The facts, the interpretation, the impact, the standard or the way it was raised. Ask rather than assume.
2. How direct is my evidence? Separate what I saw for myself from what I was told and what I have concluded over time.
3. What might I be missing? Consider what context, constraints or history the other person may hold that I do not.
4. What do I need them to understand, and what do I need them to do? Distinguish the expectation and the impact from the parts that can remain open.
5. Where am I willing to change my mind, and where am I not? Be honest with yourself before you are honest with them.

There is then one final practical decision: what do you actually want from this conversation? Shared understanding of what happened, agreement on an expectation, a change in behaviour, or simply to have your view heard? Knowing that before you begin makes it much easier to tell, in the moment, what genuinely needs agreement and what can stay open.

Conclusion

You cannot guarantee that an employee will agree with your feedback, and I would be wary of any approach that treats agreement as the measure of success. What you can do is treat disagreement as information, find out what is actually being disputed, test your evidence honestly, recognise where perspectives differ, and be clear about what needs to be understood and what needs to happen next.

That takes more patience than either backing down or pressing harder, and it asks something of the giver as well as the receiver. But it tends to leave both people with a clearer picture than a conversation won or lost on points.

Three questions to reflect on

1. When someone tells you your feedback is unfair, what is your first internal response: to defend, to doubt yourself or to get curious?
2. Think of feedback you have given that was disputed. Which part was actually being disagreed with: the facts, the interpretation, the impact, the standard or the delivery?
3. At the end of a difficult feedback conversation, what do you genuinely need from the other person, and is agreement the right word for it?

If you would like to explore ways of structuring a difficult feedback conversation, my article on which feedback model to use is a good starting point, and my Feedback Models Hub brings together CEDAR, BOOST, BIFF and several other approaches. I explore giving feedback and responding to feedback in my Leadership Library resources.

Sources

• Brett, J.F. and Atwater, L.E. (2001) ‘360° feedback: Accuracy, reactions, and perceptions of usefulness’, Journal of Applied Psychology, 86(5), pp. 930–942. Available at: https://doi.org/10.1037/0021-9010.86.5.930 (Accessed: 30 September 2026).
• Fulham, N.M., Krueger, K.L. and Cohen, T.R. (2022) ‘Honest feedback: Barriers to receptivity and discerning the truth in feedback’, Current Opinion in Psychology, 46, 101405. Available at: https://doi.org/10.1016/j.copsyc.2022.101405 (Accessed: 13 September 2026).
• Jones, E.E. and Nisbett, R.E. (1971) The actor and the observer: Divergent perceptions of the causes of behavior. Morristown, NJ: General Learning Press.
• Stone, D. and Heen, S. (2014) Thanks for the feedback: The science and art of receiving feedback well. New York: Viking.
• Wildman, A. (n.d.) ‘CEDAR feedback model’, AnnaWildman.com. Available at: https://annawildman.com/cedar-feedback-model/ (Accessed: 30 September 2026).
• Zyberaj, J. (2026) ‘Examining the relationship between supervisory feedback characteristics and employee feedback processing: A systematic review and meta-analysis’, Current Psychology. Available at: https://doi.org/10.1007/s12144-026-09323-y (Accessed: 13 September 2026).