How to Choose a Color Palette for a Brand

Updated September 2026 · Written and maintained by the Progression Agency strategy team

A brand palette is a working system, not a mood board. The constraints — who it must appeal to, what it must be legible against, and where it will physically be produced — eliminate most options before taste gets a vote. This is the process: what to decide first, how many colors you actually need, how to check contrast against the real background at the real size, why colors behave differently in print, and how to document the palette so it survives contact with everybody who has to apply it.

How do you choose a color palette for a brand?

Start from the job the palette has to do — who it must appeal to, what it must be legible against, and where it will be used — then pick one lead color and build everything else around it.

The mistake almost everybody makes is starting from colors they like. A palette is a working system, and the constraints decide most of it before taste gets a vote.

What has to be decided before picking any color?

The audience, the category conventions, the surfaces it will appear on, and the accessibility requirements — because each of these eliminates options.

A palette chosen without these constraints will be revised the first time somebody tries to set small text in it or print it on packaging.

Who it has to appeal to

Not what you like. Color associations differ by audience, by category and by culture, and the audience is who has to respond to it.

What the category already looks like

Conventions exist because they help people recognize what something is. Breaking them buys distinctiveness at the cost of immediate comprehension.

Where it will actually be used

Screen, print, packaging, signage, embroidery, vehicle livery. Each has different constraints and some colors fail badly in one of them.

What contrast it must achieve

If text will be set in it, the contrast requirement is a hard constraint rather than a preference.

How many colors the work genuinely needs

A small brand producing a few things needs fewer than a product with dozens of interface states.

Who will apply it

A palette with subtle distinctions falls apart in the hands of people without design training.

How many colors should a palette have?

One lead color, one or two supporting colors, a neutral range, and whatever functional colors the work actually requires — usually fewer than people expect.

Large palettes are harder to apply consistently. The discipline of a small one is what makes work recognizable.

The last row is worth doing deliberately. Keeping one color out of the system leaves room for a sub-brand or a campaign later without breaking what already exists.

Where should the lead color come from?

From something true about the business or the category rather than from a trend — and it should be a color the competition is not already using.

A palette that is distinctive within its own category is doing its main job. A palette that is beautiful and identical to three competitors is not.

How do you check contrast properly?

Against the actual background it will be used on, at the actual size, using a contrast ratio check rather than judgment by eye.

Eyes adapt. A color pairing that looks fine on a bright monitor in a dark room can be unreadable on a phone outdoors, and the ratio is what settles it.

Check every text and background pairing

Not just the obvious one. Buttons, links, captions and form labels all need checking separately.

Check at the sizes you will actually use

Contrast requirements differ between large text and body text, which matters for captions and labels.

Do not rely on color alone to convey meaning

Color vision differences are common enough that any state signalled only by color will be missed by some people.

Check the hover and focus states too

These are frequently designed last and frequently fail contrast.

Check it on a phone in daylight

The most common real viewing condition and the harshest.

Document the approved pairings

So people applying the palette do not have to work it out.

Why do colors look different in print?

Because screens emit light and print reflects it, so the same specification does not produce the same result — and some screen colors cannot be printed at all.

Brand colors should be specified separately for screen and for print, and checked on the actual stock where packaging or signage is involved.

The pattern here is that every physical medium narrows what is achievable, which is why a palette built only on screen tends to disappoint the first time it is produced.

What are the common mistakes?

Too many colors, no neutrals, a lead color the category already owns, and no contrast checking until after launch.

All four are cheap to avoid at the start and expensive to correct once the palette is applied across everything a business produces.

Should you follow color trends?

Not for a lead brand color. Trends date, and an identity is meant to work for years rather than for a season.

Trends are more usefully applied to campaign material, where a dated look is a small and correctable problem.

How do you document a palette so it survives?

With exact values for every context, the approved text pairings, what each color is for, and what not to do — in a document people can actually find.

An undocumented palette drifts within months, because everybody who applies it makes small reasonable decisions that add up to inconsistency.

What should the documentation include?

Screen and print values, neutral range, approved contrast pairings, proportions, and examples of correct and incorrect use.

The proportion guidance matters more than people expect: the same palette used in different proportions produces work that does not look related.

The proportion row is the one that most often explains why a business’s material does not look consistent even though everybody used the right colors.

How do you test a palette before committing to it?

Apply it to real material — a page, a document, a piece of packaging — rather than judging it as swatches.

Swatches look good in almost any combination. The test is whether the palette still works when one color occupies ninety per cent of a layout and another occupies two.

What do specific colors actually signal?

Associations are real but weak, culturally variable and heavily category-dependent — which means they are a tiebreaker rather than a decision rule.

Anybody presenting color psychology as settled science is overstating it. What is reliable is that a color reads differently depending on the category it appears in.

The consistent point is the right-hand column: every association is contingent on category and execution, which is why competitor context matters more than the association itself.

Frequently asked questions