In short
A useful critique starts with the presenter stating the goal, the constraints, and what feedback they actually want. Reviewers then respond to whether the work meets that goal, rather than to personal preference. Without the framing, a critique collects opinions the presenter cannot act on — and conflicting opinions are resolved by returning to the stated goal, not by counting votes.
The framing does most of the work
Before anyone comments, the presenter should state three things.
The goal. What is this trying to achieve, for whom? "Make it easier for a first-time visitor to understand what we do" is a goal. "Redesign the homepage" is not.
The constraints. Brand rules, technical limits, deadline, what has already been decided. Reviewers who do not know the constraints will helpfully suggest things that are impossible.
What feedback is wanted. Early structural exploration needs different input from a final visual pass. "I want to know if the hierarchy works — ignore the colours, they are placeholder" saves everyone twenty minutes.
Without this, a critique defaults to a preference survey, and the presenter receives a set of opinions that cannot be reconciled because they were never answering the same question.
Respond to the goal, not to taste
The difference between useful and useless feedback is whether it references the goal.
"I don't like the blue" is taste. "The blue heading competes with the call to action, so my eye goes there first" is an observation about whether the work achieves its aim.
A useful format: what you observed, then why it matters. "I read the third section before the second" is more valuable than "the layout is confusing," because it is specific and the presenter can work out why.
And where you have a preference rather than a reason, say so — "this is personal taste, ignore if it does not help" is a completely legitimate contribution, and labelling it prevents it being weighted as analysis.
Ask before asserting
Much of what looks wrong in a critique has a reason behind it.
"Why did you put the price above the features?" often gets an answer that resolves the objection entirely. Leading with "the price should be lower down" skips that and puts the presenter into defending rather than explaining.
Questions also surface the decisions worth examining. If the presenter cannot answer why something is as it is, that is the finding.
Structure that works
Presenter frames — goal, constraints, what they want. Two minutes.
Silent review. Everyone looks properly before anyone speaks. Without this, the first comment anchors the discussion and quieter people never introduce their own observation.
Observations first, round the room. What people noticed, before anyone proposes solutions.
Then discussion, focused on the items that recur or that clearly bear on the goal.
Presenter summarises what they are taking away. This is the step most often skipped and the one that makes the session worth holding — it converts a discussion into decisions, and it lets people correct a misunderstanding on the spot.
Handling conflicting feedback
It will conflict. Two people will want opposite things, both plausibly.
Return to the goal. Which suggestion serves it better? That usually resolves it, and where it does not, it reveals that the goal is underspecified — which is a more useful finding than either suggestion.
The presenter decides. A critique is input, not a vote. Design by majority preference produces work nobody chose and nobody defends.
Note unresolved disagreements rather than settling them by fatigue. Some are worth testing rather than arguing.
What undermines a critique
The most senior person speaking first. It anchors everyone. Have them go last, or the session collects their opinion with extra steps.
Solutioning immediately. "Move that to the right" skips the problem, and the presenter has better context for solving it once the problem is named.
Reviewing without context. Someone dropped into a critique with no background gives feedback on the wrong thing.
Too many people. Above six or so, some contribute nothing and the discussion fragments.
No time limit. Critiques expand to fill available time and the last twenty minutes are rarely the valuable ones.
Making it safe enough to be honest
Critique only works when people can say what they actually think, which requires that the presenter is not defending and the reviewers are not scoring points.
Two habits help: critique the work, not the person — "this section is doing two jobs" rather than "you always overcomplicate" — and say what is working, genuinely. Not as a softener, but because knowing which parts to preserve is as useful as knowing what to change.
A team where critique is uncomfortable will quietly stop having them, and the work gets worse without anyone deciding it should.
If you want an outside perspective on work in progress, book a call — an external reviewer has no internal politics to navigate, which is occasionally the whole value.
Common questions
How should a design critique start?
With the presenter stating the goal, the constraints, and what feedback they actually want. Without that framing the session becomes a preference survey, and the presenter leaves with opinions that cannot be reconciled because nobody was answering the same question.
What makes design feedback useful?
Referencing the stated goal rather than personal taste, and describing what you observed before why it matters. 'I read the third section before the second' is more useful than 'the layout is confusing', because it is specific enough for the presenter to work with.
How do you handle conflicting feedback in a critique?
Return to the stated goal and ask which suggestion serves it better — that resolves most conflicts, and where it does not, it reveals the goal is underspecified. The presenter decides; a critique is input rather than a vote, since design by majority preference produces work nobody defends.
Who should speak first in a design critique?
Not the most senior person, whose opinion anchors everyone else's. Start with silent review so people form their own observations, then go round the room with observations before anyone proposes solutions, and have senior voices contribute later.
How do you end a design critique?
With the presenter summarising what they are taking away. This step is most often skipped and is what makes the session worth holding — it converts discussion into decisions and lets people correct a misunderstanding immediately rather than after the work is redone.
