How to Give Feedback That Someone Can Use
By Dara Halloran · Community & Practice
Feedback after a talk usually arrives in one of two failure modes.
Too kind. "That was great, really enjoyed it." The speaker learns nothing and knows they learned nothing, which also quietly teaches them not to trust praise.
Too vague. "Maybe be a bit more confident?" This names a symptom as though it were an action. Nobody has ever gone home and successfully practised being more confident.
Both come from the same place: the person giving feedback wants to help and does not want to hurt, so they reach for generality. Generality is where both problems live.
Feedback can make things worse
It is easy to assume that any feedback is better than none. The research says otherwise.
In 1996 Avraham Kluger and Angelo DeNisi pulled together decades of studies on feedback and performance in a large meta-analysis.1 On average, feedback helped. But in more than a third of the comparisons, people who got feedback went on to perform worse than people who got none. Their explanation was about where the feedback points attention. Feedback that keeps attention on the task, on what to do differently, tends to help. Feedback that pulls attention towards the self, towards how good or bad the person is, tends to help less or to backfire.
John Hattie and Helen Timperley reached a compatible conclusion in a 2007 review.2 Useful feedback, they argued, answers three questions: where am I going, how am I going, and where to next. Feedback about the person, such as praise for being talented, is the kind least likely to change what someone does.
Read with that in mind, "you were great" and "be more confident" are the same mistake. Both are about the speaker. Neither is about the talk.
Specificity is what makes honesty survivable
The counter-intuitive part: specific criticism is often easier to hear than vague criticism, not harder.
"You lost me in the middle" is a judgement on the talk as a whole. The speaker has nothing to do with it except feel bad.
"You lost me at the point where you moved from the hospital story to the statistics; the transition happened in about four words" is a much sharper observation, and it lands better, because it is obviously about a fixable thing rather than about them.
Vagueness feels gentle to the giver and generalises to the whole person in the receiver. This is exactly backwards, and it is why well-meaning rooms produce feedback nobody can use.
The shape
Three items, in this order, no more:
1. One moment that worked. A specific beat. "The pause after 'she left in March', when the room went completely quiet." Not "good energy." The speaker should be able to identify the exact second.
2. One moment you disengaged. Where your attention went. Report it as an observation about yourself rather than a verdict on them: "I drifted during the middle section" is fair and hard to argue with, and it points at the same place a harsher framing would.
3. One thing to try. An action, small enough for next week. "Try cutting the second example entirely and see if the section still works." Not "improve your structure."
These map neatly onto Hattie and Timperley's questions. The moment that worked shows where the speaker is going right. The moment you drifted shows how it is going. The thing to try is where to next.
A worked example
Here is the same feedback written twice.
Version one: "Really good talk, you obviously know your stuff. Maybe a bit fast in places and could be more engaging. Overall great though!"
Version two: "The line about the six-day document got a real reaction, people laughed and then went quiet. I lost the thread when the second chart came up, because I could not tell what it was meant to show. Next time, try saying the conclusion of the chart before you show it."
The first version is longer on praise and shorter on information. The second contains one compliment the speaker can repeat on purpose, one location to fix, and one experiment. It would also take less than a minute to say.
Language that does the work
- Describe, do not diagnose. "You looked at your notes for most of the second half" beats "you seemed underprepared."
- Own your reaction. "I lost the thread" beats "it was confusing."
- Prefer subtraction. "What if you cut this" is usually easier to act on than "what if you added something."
- One thing. A long list rarely produces a long list of changes; one clear suggestion is far more likely to be tried.
Write it to someone who heard it
Feedback is only as good as the listening behind it. On Exprea, a written note to another member has two parts, what was strong and what to work on, and can only be sent to someone you were actually in a session with. The reason is simple: an evaluation from somebody who did not hear you speak is worth nothing to either of you.
The same rule works anywhere. If you did not hear the talk properly, say so, and say less.
Receiving it
The counterpart skill, and the harder one: when you receive feedback, say thank you and write it down. Do not explain.
Explaining ("right, but the reason I did that was...") is completely natural, and it reliably discourages the feedback. The person concludes that giving it costs an argument, and next week they say "really enjoyed it."
You are allowed to disagree. Disagree later, privately, when you decide what to change. In the moment, the only useful response is to collect the data.
What to do next time
The next time someone asks what you thought of their talk, give them exactly three sentences: one moment that worked, one moment you drifted, one thing to try. Then stop. If they want more, they will ask.
References
- Kluger, A. N., & DeNisi, A. (1996). The effects of feedback interventions on performance: A historical review, a meta-analysis, and a preliminary feedback intervention theory. Psychological Bulletin, 119(2), 254–284. https://doi.org/10.1037/0033-2909.119.2.254
- Hattie, J., & Timperley, H. (2007). The Power of Feedback. Review of Educational Research, 77(1), 81–112. https://doi.org/10.3102/003465430298487