A design critique is meant to test a design against what it was made to do. Held without the goals written down in front of everyone, it tests the design against the room’s taste, and the most senior taste often wins. The brief turns a critique into something other than a vote, so it goes on the screen first.
Without stated goals, a critique turns into a vote on taste
Picture the session. A designer shares a checkout redesign with six colleagues and asks for thoughts. The lead says the summary panel feels heavy. A product manager prefers the old step indicator. Somebody asks whether the green could be warmer. By the end there is a list of reactions, ranked in practice by who said them, and nothing in the room can tell the designer which of them describe a problem.
By the usual standards nobody did anything wrong: the comments were specific and polite. What was missing was a criterion. The Nielsen Norman Group’s article on design critiques (Sarah Gibbons, 2016), which cites Adam Connor and Aaron Irizarry’s Discussing Design, defines the practice as “analyzing a design, and giving feedback on whether it meets its objectives”, says plainly that “it does not mean simply judging a design”, and sets a precondition: “there must be agreement on the problem that needs to be solved.”
The checkout session skipped the precondition. Every remark was therefore a judgement, and with no shared yardstick in that room, the judgements were sorted by rank. The lead’s “heavy” carries more weight than the product manager’s preference for reasons that have nothing to do with buyers.
Critique, feedback and approval ask different questions
Three activities get filed under “review” and booked into the same meeting slot, and a critique often goes wrong when one of them borrows another’s rules. Design feedback is any reaction from anyone with a stake, and giving it well is about making each comment usable. Design approval is a decision taken against a frozen version. Critique sits between them, defined by its yardstick.
| What to compare | Design critique | Design feedback | Design approval |
|---|---|---|---|
| The question it answers | Does this design meet the goals in the brief? | What does this person see, want or worry about? | Is this version the one that ships? |
| What a remark is measured against | A line of the brief, shared by the room | Whichever goal its writer names | The version shown and the round’s scope |
| Who speaks | Anyone who has read the brief | Anyone with a stake | The named approver |
| When it happens | Mid-work, while the design can still move a lot | At any stage, often in writing | At the close of a round |
| What it produces | Remarks sorted against goals, and sometimes an amended brief | Comments the designer weighs | A yes or a no, written against a version |
The second row is where a critique collapses. With the brief absent, the critique column borrows its yardstick from the feedback column, each person measuring against a goal of their own choosing, and the session becomes a feedback round held aloud. The opposite slide happens too: a team treats the critique as the approval, the designer brings unfinished work to a room that believes it is deciding, and the work stops moving at the stage it was supposed to move most.
The brief goes on the screen before the design does
The fix is an order of operations. Before anyone sees a layout, the presenter puts up the part of the brief this critique tests: the objective, the person the design is for, the constraints that are off the table, and the questions the designer wants answered today. For the checkout, that might read: a first-time buyer on a phone should reach payment without leaving the page to check delivery costs; the payment provider’s form cannot be restyled; today’s question is whether the order summary helps that buyer or slows them down.
Goals written as checkable lines, close in spirit to a design’s own acceptance criteria, are far easier to hold a remark against than adjectives like “clean” or “modern”. A remark can be tested against “reach payment without leaving the page”; against “clean” there is nothing to test.
Naming the open questions narrows the session on purpose. A designer three days into a layout has no use for a debate about the brand colour, and a brief that marks the colour as settled lets the facilitator park that remark without arguing it. Written goals also guard against the opposite failure, a room so wary of seeming harsh that it says nothing useful, because a remark tied to a goal is visibly about the work.
Where no brief exists, the first critique should be spent writing one: the team agrees the goals, the brief’s owner writes them down, and the design comes back next time. That costs a session. Running the session without one risks costing more, because each later round can re-argue the same preferences with nothing to settle them against.
Each remark is tied to a goal, filed as taste or read as a gap
The facilitator has one tool, a question asked after every remark: which goal is this about? The answers fall into three groups, and they are handled differently.
A remark that names a goal is critique proper. “On a small phone the summary panel pushes the pay button below the fold” ties straight to the first-time buyer on a phone, and the designer can test a fix against the same line next round.
A remark that names no goal and needs none is taste. The warmer green belongs here. Taste is real information about the people who will live with the result, so it stays on the list, set aside for whoever owns that result. In most teams that is the approver who signs off the next round, and they weigh it after the session. It does not outrank a goal-tied remark, however senior its author.
The third group is the one critiques tend to miss: a remark that names no goal because the brief left one out. Somebody points out that the delivery estimate becomes unreadable when the phone’s text size is turned up. Nothing in the checkout brief mentions accessibility, so by the letter of the brief the remark is out of scope, while by any reasonable standard for the product it clearly matters. The right move is an amendment to the brief, made by whoever owns it, after which the remark is a criterion like the others. A brief that a critique can never amend becomes a polite way of ignoring the people in the room.
A critique ends in a sorted list, and the deciding happens elsewhere
At the close, the designer should hold the remarks in those three groups: goal-tied problems marked against the line each one concerns, taste parked for the approver, and any amendment to the brief written down with the name of whoever owns it. Nothing should be approved in the critique itself. Calls on taste belong to that approver as the owner of the result, and the yes on a version is given in the approval round, against a frozen copy. A critique that ends in a show of hands has quietly become an approval, decided by people who were invited to comment.
When the design under critique is already a built page, “the summary panel” can mean three states at three widths, and a remark written up after the session rarely records which one was on the screen. In Notarus, a member on Pro or Scale can open a review session on the team’s own site, pick the element a remark concerns and file it as a ticket. The ticket keeps the picked element as a DOM selection, so the critique’s list reaches the board already attached to the page. The line of the brief each remark answers is still typed by the person filing it.
Whatever the brief gained in this session becomes part of what the next round is measured against.
