Giving feedback that changes something

Feedback is treated as self-evidently useful. Reviews of the research find effects that are positive on average, highly variable, and negative in a substantial minority of cases.

Which means feedback is an intervention that can make performance worse, and understanding when is more useful than technique.

When it makes things worse

The main finding from the feedback intervention literature is that effects depend heavily on where attention is directed.

Feedback directed at the person rather than the task tends to reduce performance, because it shifts attention to self-evaluation and away from the work.

"You are careless" directs attention to the self. "There were three unchecked figures in section two" directs it at the task.

Praise about ability has the same problem in the opposite direction, which is one of the findings behind work on how children respond to being praised for intelligence rather than effort.

Feedback with no clear path to improvement produces distress without change.

Feedback delivered when the person cannot act on it, which is most feedback delivered in annual reviews about things that happened eight months earlier.

What raises the chance it lands

Proximity in time. Feedback close to the event is more actionable and less disputable, because both people remember it.

Specificity. General feedback cannot be acted on. "Be more strategic" is not a behaviour.

The test: could the person do something differently tomorrow as a result? If not, it is an assessment rather than feedback.

Separating the observation from the interpretation. Discussed elsewhere on this site and it applies here with particular force.

State what happened. Then state what you concluded, as a conclusion rather than as a fact. This invites correction where your interpretation was wrong, which it sometimes is.

Asking before telling. "How do you think that went?" frequently surfaces the same point from the other person, which is far more likely to produce change than being told.

It also reveals whether they already knew, which changes the conversation entirely.

The sandwich problem

Praise, criticism, praise remains widely taught and is widely seen through.

The consequences are that the criticism is diluted or missed entirely, and that praise becomes a signal that criticism is coming — which devalues genuine praise.

The alternative is not brutality. It is separating them: praise when there is something to praise, raise concerns when there is a concern, and do not use one to smuggle the other.

Positive feedback is underused

The asymmetry in most workplaces.

People receive feedback overwhelmingly when something is wrong, which means feedback becomes a negative signal and is dreaded.

Specific positive feedback — naming what was done well and why it mattered — is rare, is remembered, and does the useful work of telling someone which of the things they do are the valuable ones.

General praise does not do this. "Great job" conveys approval and no information.

Receiving it

The harder half.

The instinct is to explain immediately, which prevents the rest of the feedback arriving.

Listening to the whole thing, asking for specifics, and saying you will think about it produces better outcomes than defending in the moment.

Separating the accuracy of the content from the quality of the delivery matters, because badly delivered accurate feedback is common and dismissing it on delivery grounds is expensive.

And asking for feedback specifically — "what would have made that better?" — reliably produces more useful responses than a general invitation, because a specific question is easier to answer honestly.

The organisational condition

Feedback works in a relationship where it is normal and does not work where it is an event.

Somewhere that only gives feedback in formal reviews has made feedback frightening. Somewhere it happens weekly and in both directions has made it ordinary.

That difference is set by managers' behaviour rather than by process, and it determines whether any of the technique above matters.

Asking for it well

The receiving side, which is under half of most advice on this and determines whether any of it arrives.

General requests produce general answers. A specific one — what should I have done differently in that meeting, what would make this draft clearer — produces something usable.

And the response to receiving it determines whether it comes again. Defending once is usually enough to close the channel permanently, which is why so many senior people receive none.