How to Get Honest Feedback When You're the Boss
Most leaders have asked for candid feedback, received something vague and generous, and concluded that their team is not comfortable being direct. The diagnosis is usually wrong. Your team’s reticence is not a cultural failing or a courage deficit — it is an accurate reading of the incentives, which are that telling you something unwelcome carries real risk and offers no reward. Once you see it that way, the question changes from how to encourage candour to how to change what candour costs. This page is the practical version.
What’s inside
The silence is rational
Why the standard asks fail
Six techniques that work
The first ten seconds
Sources outside your reporting line
The anonymous-survey trap
Where Blomma fits
The silence is rational
Run the calculation from your report’s side. You control their scope, their compensation, their promotion, and whether they stay. You have a long memory and they have no way of knowing which of your reactions is momentary. The upside of telling you something uncomfortable is that you might improve slightly. The downside is a possible permanent shift in how you see them.
That is a bad bet, and reasonable people do not take bad bets. Which means the silence is not something to be disappointed by — it is information about the structure, and it will not change because you have said you welcome candour.
Two consequences follow. First, any technique that relies on people being brave will fail, because you are asking them to absorb an asymmetric risk for your benefit. Second, the techniques that work are the ones that lower the cost of speaking, narrow what is being asked, or move the request outside the reporting line entirely.
And one hard truth: some categories of feedback will never come from your reports, however good your culture is. For those you need a source with no stake in your opinion of them.
Why the standard asks fail
Four common approaches, and the specific reason each produces nothing.
“My door is always open.”Puts the entire burden — initiating, choosing the moment, framing it — on the person with the least power. Nobody walks through the door with the hard thing.
“Do you have any feedback for me?”Unanswerable. It asks someone to characterise you, on the spot, with no framing. The polite response is “no, all good,” and you have taught them that the ritual exists and means nothing.
Asking in a group.Now they are also managing how they look in front of peers. Group settings produce the safest possible version of any view.
Asking once, seriously, and then not following up.People are testing whether it is real. A single ask followed by no visible change reads as an exercise, and the second ask gets less than the first.
The common thread: each asks for a large, unstructured, unrewarded disclosure. The fixes make the ask small, specific, and visibly safe.
Six techniques that work
One. Ask narrow and immediate.After a specific event:“What did I miss in that meeting?”or“How did that land?”Narrow questions are answerable without characterising you, and immediacy means the answer is about one thing rather than about your general nature.
Two. Ask about the work, not about you.“What would have made that decision easier to execute?”People will tell you about a process problem long before they tell you about you, and the answer is frequently about you — you just got it via a route that carried no risk.
Three. Name a specific hypothesis.“I think I jumped in too early on that — did it land that way?”Confirming your self-criticism is far cheaper than originating criticism. This is the single most effective technique on the list, and it works because you have already paid the social cost.
Four. Ask for one thing, not everything.“If there were one thing I should do differently next quarter, what would you pick?”Bounded, and it gives permission to raise exactly one item rather than opening an assessment.
Five. Ask through a third party.A coach, an HR lead, or a board member conducting a short round of conversations and reporting themes without attribution. This removes the direct risk entirely and reliably surfaces things you would not otherwise hear.
Six. Reward it visibly, the first time.Someone takes the risk; you thank them publicly and act on it visibly. Everyone is watching what happens to that person, and this single event does more than any number of invitations.
The first ten seconds
Whatever technique you use, the whole thing is decided by your reaction, and specifically by the first ten seconds.
The natural response to unwelcome feedback is to explain. Your intent was different, there was context they do not have, the situation was more complicated. Frequently all true. And it teaches the person that the price of telling you something is having to defend it — which means there will not be a next time, and they will tell their peers.
What to do instead: take it, ask for one more example, thank them specifically, and say nothing else in that conversation. The explaining can happen later or not at all. If your intent genuinely matters, raise it a day later — but by then you will usually find it does not.
Then close the loop within a month. Go back and say what you changed and ask whether it landed differently. This is the step almost nobody does and it is what converts one piece of feedback into a supply, because it demonstrates that the risk they took produced something.
One more: do not reward feedback and then act as though nothing was said. Silent non-action is read as disagreement you were too polite to state.
Sources outside your reporting line
Some feedback will not come from people who work for you, no matter what you do. Four alternative sources.
A coach.Someone whose only function is to see your situation clearly, with no stake in the outcome and no career exposure to your reaction. This is the mechanism most senior leaders lack entirely.
Peers outside the company.Other founders at a similar stage, ideally not investors and not friends who are invested in reassuring you. They can recognise a pattern from their own experience and have nothing to lose by naming it.
A board member with the right relationship.Useful and partial — they are assessing you, which limits what you will say to them and colours what they say to you.
Departing employees, asked properly.Exit conversations are the most candid you will get, and they are usually wasted on logistics. Ask specifically: what should I have done differently, and what will you say about this company to a friend? Run these yourself for senior departures.
The anonymous-survey trap
Worth a caution, because it is the standard institutional answer and it has a specific failure mode.
Anonymous surveys do surface things, particularly at scale, and they have two problems. They produce aggregate sentiment without the specifics that would let you act — knowing that thirty percent of people find communication unclear does not tell you what to change. And they can substitute for the direct relationship: a leader who has a survey may feel informed while having no actual channel to anyone.
Use them as a signal detector rather than a feedback mechanism. When a survey flags something, that is the beginning of an inquiry — go and have the narrow, specific conversations that tell you what it actually means.
And never try to identify a respondent, even casually, even to follow up helpfully. One instance of that and the channel closes permanently.
Where Blomma fits
There is a structural limit to everything above: the people with the best information about your effect are the people with the most to lose by sharing it. You can lower that cost considerably and you cannot remove it, which means part of your feedback needs to come from outside the reporting line entirely.
Blomma is an always-on AI career coach with no stake in your good opinion and no career exposure to your reaction. Use it to construct the narrow, specific asks rather than the general ones — particularly technique three, naming your own hypothesis, which most leaders find the hardest to write and the most productive. Use it to work through what a piece of feedback actually means before you react to it, which is what stops you explaining in the first ten seconds. Use it to prepare senior exit conversations, which are the most candid information available and almost always wasted. And use it as the standing source of the read your reports cannot give you.
The second application: everything here applies at every level of your company, and your managers have the same problem with their teams. A company where feedback travels upward is built by the management layer, not by the founder’s example alone..
Your team is not withholding out of timidity. They are making a reasonable calculation about an asymmetric risk — and the fix is to make the ask smaller, the risk lower, and the reward visible.