How to Critique a Logo (Without Guessing)

Most logo feedback is useless, and not because the people giving it lack taste. It is useless because it arrives as a verdict rather than an observation. "I don't love it" tells a designer nothing they can change.

A critique is different from an opinion. It separates what can be verified from what is preference, deals with them in that order, and ends with something specific enough to act on.

Check the objective things first

Start where there is a right answer. This takes five minutes and removes most of the argument before it starts.

• Render it at 32 pixels. Do the strokes survive? Is it still the same shape, or has it become a smudge?
• Remove the colour. Are the parts still distinguishable in greyscale, or does the mark collapse into one silhouette?
• Check the contrast against the backgrounds it will actually sit on, not the one in the presentation.
• Look at the alignment. Does the mark sit level next to type, or does it lean?
• Ask what files exist. A logo with no vector is not finished, whatever it looks like.

Anything that fails here is a craft problem, not a taste problem, and craft problems get fixed rather than debated.

Then separate taste from craft

Once the measurable issues are listed, the rest of your reaction is preference, and it helps enormously to say so out loud. "This is a preference, not a problem" is one of the most useful sentences in a design review.

It is not that preferences do not matter. It is that mixing them with craft issues makes both harder to resolve, because the designer cannot tell which items are negotiable.

A useful test: can you state the change you want without using an adjective? "Make it more premium" fails. "The wordmark is competing with the symbol, try reducing the symbol by a third" passes, and it can be tried in a minute.

Ask what it is being compared against

A logo is never judged in isolation. It is judged next to competitors, in the contexts it will actually appear.

Put it beside three competitor marks at the same size. Most identity problems become obvious immediately, and they are usually one of two things: it looks like everyone else, or it is trying so hard to be different that it stops fitting the category.

Write feedback a designer can act on

Three things make feedback usable.

• Name the problem, not the solution. "It disappears in the app icon" is more useful than "make it bigger", because the designer knows more ways to solve that than you do.
• Say how much it matters. A list of twelve equally weighted notes is unusable. Two blockers and a list of nice-to-haves is a plan.
• Point at the evidence. A screenshot of the mark at favicon size ends an argument that words would sustain for a week.

Know when to stop

Logos are refined into blandness more often than they are improved. Every round of feedback pulls a mark toward the average of everyone's preferences, and the average of enough opinions is always a circle with a gradient in it.

Stop when the objective checks pass and the remaining notes are preferences that contradict each other. That is the signal that the work is done and the process has started eating it.

Or have it scored

If you would rather not run these checks by hand, our logo review does the measurable half for you. Five of the seven dimensions are calculated on the artwork itself, so you get the numbers rather than an impression, and every score comes with the reason behind it.

For the criteria in full, what makes a good logo goes through all seven with what each one catches.

Critiquing your own logo

This is by far the hardest case, because you know what the mark is meant to be, and that knowledge quietly fills in everything a stranger would miss.

Two things help. Put it away for three days and look again cold; almost everyone sees something on the second viewing they were blind to on the first. And view it in a context you have not designed, dropped into a real browser tab or a real phone home screen rather than sitting centred on a presentation slide with generous white space around it.

The centred-on-white-slide view flatters everything. It is the single most misleading way to evaluate a logo, and it is how nearly every logo gets evaluated.

When stakeholders disagree

Conflicting feedback is usually not really about the logo. It is several people optimising for different things without saying which, so the argument never resolves.

The way out is to make the criteria explicit before discussing options. Ask what this mark has to do: does it need to work at 16 pixels, does it need to print in one colour, does it need to read as premium or as approachable. Agreeing on that list takes ten minutes and turns most taste arguments into answerable questions.

Where opinions still conflict after that, someone has to decide. Design by committee is not a process, it is the absence of one, and the output is reliably the least objectionable option rather than the best one.

A critique you can copy

A usable written critique has four parts, in this order.

• What works. Name it specifically, so the designer knows what to protect through revisions. This is not politeness, it is a spec.
• Blockers. Things that must change before launch, with the evidence. Keep this list short and honest; if everything is a blocker, nothing is.
• Worth considering. Real suggestions the designer is free to reject with a reason.
• Preferences. Explicitly labelled, explicitly optional.

That structure does most of the work on its own, because it forces you to sort your own reaction before sending it, which is the step people skip.