The reason you get useless feedback is not that people are conflict-averse, though many are. It is that you asked an impossible question.

"Any feedback on that?" requires the other person to scan everything you did, decide what matters, judge how much candour the relationship can carry, and phrase it kindly — in the four seconds before the silence gets awkward. The rational response is "yeah, that was good". You would do the same.

Everything useful follows from reducing that workload.

Ask a narrow question

Narrow means one dimension, one artefact, and a hypothesis they can confirm or correct. Compare:

  • "How did the presentation go?"
  • "I thought I lost the room in the middle section on pricing. Did it read that way from where you were sitting?"

The second is answerable in one sentence. It also gives permission: you named the weakness, so agreeing is not an accusation.

Other forms that work: "if you had to cut one section, which one?", "what would you have done differently in the third paragraph?", and "what is the first thing that confused you?" The last is best for written work, because confusion is observable and taste is not.

Ask the right person

Three different people give three different kinds of feedback, and mixing them up is why advice contradicts itself.

  1. Someone doing the work now. Best for craft — conventions and failure modes. Worst for judging whether it lands with a senior audience.
  2. Someone one or two levels above. Best for judgement, scope, and whether you framed the problem correctly. Often vague on execution detail.
  3. Someone in the audience the work is aimed at. The only source of truth on whether it worked. A finance stakeholder is the only person who can say whether your analysis answered a finance question.

If your work is not landing and you have only ever asked group one, that is probably the explanation.

The 24-hour rule

Feedback decays fast. Ask within a day and you get specifics — what you said, when the room went quiet, the sentence that did not parse. Ask a week later and you get a summary, which is another word for a generality.

This holds in both directions: a specific moment raised within 24 hours is usable, a pattern described a month later is an argument. So ask before you leave the room, or send one message the same afternoon. Two sentences is enough.

Separate taste from defect

Most disagreement about feedback is a category error. There are three types and they deserve different treatment.

Defects are objectively wrong — a broken calculation, a missing risk, a claim the data does not support. Fix them without debate.

Convention is how things are done here. Not right or wrong, but deviating has a cost you may not see yet. Follow it unless you have a reason.

Taste is preference — sentence length, structure, colour, tone. Reasonable people differ, and it tells you about the reviewer as much as the work.

The skill is asking which you are hearing. "Is that a must-fix or a preference?" is a perfectly acceptable question and almost nobody asks it. Five seconds, and it saves hours of rewriting to satisfy one person's taste.

The 48-hour rule for feedback you disagree with

In the moment, do nothing beyond thanking them and asking clarifying questions. Wait two days, then re-read the work as if someone else wrote it. If you still disagree, the useful question is not "are they right" but "why did they read it that way" — because if one person misread it, others will. The criticism is often wrong and the underlying signal still real.

What to do with feedback you reject

You are allowed to reject feedback. Rejecting it silently and repeatedly from the same person ends the supply.

If you decide not to act, say so and say why, briefly. "I kept the methodology section because the audience will challenge the numbers — but I have moved it to an appendix." That shows you engaged and leaves them room to say the reasoning is wrong.

The alternative — thanking them and changing nothing — reads as dismissal even when it is not, and people quietly stop bothering.

The scarce resource is not feedback. It is people willing to give you a second round after you ignored the first.

When your manager gives you none

This is common and rarely malice. Most managers were never trained to do it, are managing too many people, and default to silence because silence has no immediate cost. Build the loop yourself:

  • Make it small and scheduled. Ask one narrow question in each one-to-one rather than requesting a general review. Over six months that adds up to more than any annual process produces.
  • Ask forward, not back. "What would make the next one better" gets answers that "how was that" does not, because it reframes criticism as advice.
  • Ask about the next level. "What does someone a level above me do that I do not?" is high-yield, and works even with managers who cannot critique directly.
  • Go around them, carefully. Peers and stakeholders see your work. Asking them is normal and needs no permission.
  • Write your own assessment first. Send a short self-review and ask where they disagree. Correcting a draft is far easier than generating one.

Directness varies by market enough to matter. Dutch, German and Israeli workplaces tend to criticise plainly and mean what they say. UK feedback is heavily softened, and "a few small thoughts" can mean serious problems. Much of East and Southeast Asia routes criticism indirectly. If you have changed country, recalibrate before concluding nobody has concerns.