Search for advice on brand colours and you will find a lot of confident claims. Blue means trust. Red creates urgency. Green suggests growth.
There is a grain of truth in this and far less than the confidence suggests. Context does most of the work. Red means urgency in a sale banner and danger on a warning sign and celebration at a wedding, and the colour did not change.
What genuinely matters when choosing is more practical, and almost nobody writes about it.
Does colour psychology actually work?
Weakly, and it is heavily outweighed by what your industry already looks like.
The useful version is not “blue means trust.” It is that people arrive with expectations set by everything else they have seen in your category. Finance is mostly blue, so blue in finance reads as normal rather than as trustworthy. Organic food is mostly green, so green reads as expected.
That gives you a genuine strategic choice. Match the category and be immediately legible as what you are. Or break from it deliberately and be memorable, at the cost of having to work harder to be understood.
Both are valid. What does not work is choosing a colour because an article said it means something, without looking at what your competitors are doing.
How many colours should a brand have?
One primary, one dark, one light, and a small number of supports.
Most small brands are better served by fewer colours than they think. One accent, used consistently, becomes recognisable. Five accents, used interchangeably, become noise.
The practical set is: a primary colour that appears on buttons and highlights, a near-black for text, a near-white for backgrounds, and one or two neutrals in between for borders and secondary text. Anything beyond that should earn its place by having a specific job.
What do most people get wrong?
They choose colours by looking at swatches instead of at usage.
A colour that looks beautiful in a large square can be unusable as a button, too pale for text, or invisible against your photographs. The swatch is not the test. The test is the thing you will actually make.
Two constraints matter more than aesthetics and get ignored constantly.
Contrast. Text must be readable against its background. There is a measurable standard for this, and pale grey on white fails it, however elegant it looks. Anyone over forty will thank you, and a meaningful share of your buyers are over forty.
Behaviour on screens. Very saturated colours can appear to vibrate against each other, and pure black on pure white is harsher than most people expect. Off values are usually kinder.
How do I pick a colour that will still work in five years?
Avoid whatever is currently everywhere.
Design has fashions like anything else. Right now certain gradients and certain purples are extremely common, and in four years they will date a brand precisely because so many businesses adopted them at once.
A useful test: does this colour appear in the current crop of app icons and startup websites? If it does, it will read as “made around 2026” for the rest of its life.
Sturdier choices tend to be slightly unusual versions of ordinary colours. Not the default blue, but a specific blue. Not green, but a particular one you can name and own.
What about accessibility?
It is not a separate consideration. It is part of choosing well.
Around one in twelve men has some degree of colour blindness, most commonly difficulty separating red from green. If your only signal for something important is a colour difference, a meaningful slice of your audience misses it.
Two rules cover most of it. Never rely on colour alone to convey meaning, add a word or a shape. And check text contrast against its background before committing, using any of the free online checkers.
This is not a constraint that makes design worse. It rules out a small number of combinations that were going to cause problems anyway.
How do I actually decide?
A sequence that works.
Look at fifteen competitors and note their colours. You now know what normal looks like in your category, and can decide whether to belong or stand apart.
Pick one primary colour and commit. Not a palette, one colour, the one that will become yours.
Test it where it will be hardest. As a button with white text on it. As small text. Against a photograph. On a phone screen outdoors. Most candidates fail here and it is much cheaper to find out now.
Then build the rest around it. Dark, light, and only the neutrals you genuinely need.
Then write the values down, in both screen and print formats, and stop revisiting the decision. A colour becomes yours through repetition, and repetition requires you to stop changing your mind.
We choose colours by testing them where they will actually be used. Tell us what you are working on.
