I audit consumer brands. That means reading a company's actual decisions rather than its explanations of them, and colour comes up in almost every audit I write. What I have never once found is a brand whose growth can be traced to the emotional meaning of its hue. What I find constantly is a brand that became easy to identify, and stayed identical for long enough that recognition became automatic.
Those are two different claims, and the marketing internet has spent twenty years merging them.
Colour psychology is real. The version sold to marketers is mostly not. The gap between the two is where the useful work is.
The claim that does not survive scrutiny
The standard model goes: a colour produces an emotion, the emotion produces a preference, the preference produces a purchase. Each step is offered as settled. None of them is.
Start with the last step. If a hue reliably moved purchase behaviour, the effect would be enormous and trivially exploitable, and every category would have converged on the winning colour decades ago. Categories have not converged. They have fragmented, and the brands that grew fastest are frequently the ones that took the colour their category was not using.
Then take the first step, which is where the whole model breaks. In my audit of BYD, red is the primary identity because in Chinese culture red is the most auspicious colour there is, tied to prosperity, fortune and celebration. In the markets that company is now exporting into, the same red is the colour of a warning label and a clearance sticker. One hue, two incompatible readings, both correct in context.
A rule that reverses itself at a border is not a rule about colour. It is a rule about culture wearing a colour as a costume.
The context problem is just as severe within a single culture. Red means sale in a shop window, error in a piece of software, danger on a road sign, luxury on a sports car and celebration at a wedding. Asked what red means, most people answer with whichever context they were last standing in.
What actually holds up: distinctiveness, not emotion
There is a body of work that does survive contact with evidence, and it is not about feelings. Byron Sharp and Jenni Romaniuk at the Ehrenberg-Bass Institute argue that distinctive brand assets — colours, shapes, characters, sounds — drive far more real-world growth than marketers credit them for. Not because they persuade, but because they let a brand be identified instantly, with almost no attention spent.
That reframes the entire question. The job of a brand colour is not to make somebody feel something. It is to make somebody know who you are before they have finished looking.
Spotify is the cleanest demonstration I have audited. Its green is not green because green means growth or freshness. It is that particular green because nothing else on a phone screen is, and a user finds the icon in milliseconds without reading a single label. The value is in the milliseconds.
Nvidia is the same logic in a harder category. Technology branding runs on a famously narrow palette: blue for trust and enterprise, black for premium, grey for serious. Green was unoccupied. Owning it required no theory about what green means — only the observation that competitors had left it empty.
A brand colour is worth what your competitors are not doing with it. That value is measured against the category, never against a chart of emotions.
The red and yellow question
No palette is explained more confidently than McDonald's red and yellow. You will be told red stimulates appetite and yellow triggers happiness, and that the combination was engineered by psychologists to make people eat faster.
Being careful matters more than being interesting here. The appetite claim has thin support and is usually traced to studies that measured something else. What is well established is duller and more useful: red and yellow together have high visual salience, they are noticed faster than muted palettes at distance and at speed, and sixty years of unbroken pairing has built an associative link so strong that the colours alone identify the brand with the arches removed.
That is a real and enormous asset. It just is not the asset the folklore describes. It was built by repetition, not by neuroscience, and repetition is available to anyone patient enough.
Four tests a brand colour has to pass
This is what I actually run when a client asks me to choose or defend a colour. None of the four involves looking up what a hue means.
The category contrast test
Put your colour beside your six nearest competitors at thumbnail size. If a stranger cannot separate you from the group in under a second, the colour has failed, however well it tests in isolation. This is the test that most rebrands skip, and it is the only one that measures the thing distinctiveness is for.
The name-removed test
Show the colour and the shape with the wordmark deleted. If nobody identifies you, you do not own a colour yet — you are simply using one. Ownership takes consistency measured in years, which is precisely why frequent rebrands quietly destroy value that took a decade to build.
The reproduction test
The colour must survive one-colour print, embroidery on a uniform, a laser-etched product, a dark interface and a compressed video thumbnail. A gradient that only exists on a designer's monitor is not an identity, it is a mockup. The palettes that last are the ones that degrade well.
The contrast test
Text on your brand colour needs a contrast ratio of at least 4.5:1 to be readable by people with ordinary age-related vision loss, and to satisfy accessibility law in the markets that enforce it. Brands routinely pick a colour and then discover no text can sit on it, which is how you end up with a logo colour that appears nowhere in the actual product.
Real brand colours, and the reason each was chosen
Every entry below comes from a full audit published on this site. Where a company states its hex value, it is given. Where it does not, the column says so rather than offering a guess.
| Brand | Colour | Value | Why it works |
|---|---|---|---|
| Spotify | Spotify green | #1DB954 | Chosen against the category, not for the hue. No other icon on a phone screen is that specific green, so recognition happens before reading does. |
| Nvidia | Nvidia green | not published | The technology palette is famously narrow: blue for trust, black for premium, grey for serious. Green owns a gap that the category left open. |
| Nestle | Deep red | #E3000B | Applied across an enormous portfolio, so the colour does the work of a family resemblance between products that share nothing else. |
| BYD | BYD red | #C8102E | In China red carries prosperity and good fortune. The same hue reads as danger or discount in much of the West. This one pairing disproves universal colour meaning on its own. |
| Nike | Sport red | #C41230 | Used as an accent for urgency, not as the identity. The Swoosh stays constant while the palette shifts per sub-brand, which is how one brand serves Jordan, SB and ACG at once. |
| Careem | Careem green | #00C853 | A moving vehicle you are waiting for has to be identified at distance and at speed. Legibility is the requirement; the emotional reading is secondary. |
| Jazz | Jazz red | #E5002B | A dominance signal in a market where three networks compete for the same shelf, the same billboard and the same shop sign. |
| Grand Theft Auto | Neon green and hot pink | #00FF41 | Borrowed wholesale from a decade the player is meant to feel nostalgic for. The colours are a period reference before they are a psychological trigger. |
| HBL | Green with gold | not published | Green reads as national and as institutional in Pakistan simultaneously. Gold adds age. The combination says established more clearly than any tagline the bank has run. |
| McDonald's | Red and yellow | not published | The most over-explained palette in marketing. What survives scrutiny is salience and sixty years of unbroken pairing, not an appetite effect. |
Colour inside a product is a different discipline
Everything above concerns identity. Colour inside an interface follows different rules, because there it carries meaning that a user has to act on.
The one convention that genuinely transfers across cultures is the traffic light, and it transfers because it was taught rather than felt. My audit of Reliance Jio notes a data gauge running green to yellow to red as an allowance depletes, and users read their remaining balance without processing a number. That works because everybody was trained on it by roads, not because the hues carry innate meaning.
The hard rule in interface work is that colour must never be the only carrier of information. Around one in twelve men has some form of colour vision deficiency. A form that marks errors in red and nothing else is invisible to a substantial share of your users. Pair every colour with a word, an icon or a position.
Then there is the finding that quietly ends most button-colour arguments. Tests showing a red button beating a green one are almost always measuring contrast against the surrounding page, not the hue. Change the background and the winner flips. If your growth plan depends on a button colour, the problem is somewhere else on the page.
What to do instead of chasing colour meaning
- Pick for contrast, then never move. The advantage compounds with time, and it resets to zero the moment you rebrand for the sake of it.
- Name your market before you name your colour. Red for Karachi, red for Shanghai and red for Chicago are three different decisions wearing one word.
- Use one accent, not five. A palette where everything is emphasised has nothing emphasised, and the accent stops functioning as a signal.
- Apply it in the boring places. Invoices, packing tape, hold music screens, the 404 page. Recognition is built by frequency of exposure, and the unglamorous surfaces are where most of the frequency lives.
- Check contrast before you fall in love. It is far cheaper to find the 4.5:1 failure now than after the signage is printed.
Frequently asked questions
Does colour psychology in marketing actually work?
Partly, and not in the way it is usually sold. There is good evidence that colour drives recognition, visual salience and consistency of identity. There is very weak evidence that a specific hue reliably produces a specific emotion or a specific purchase decision across cultures and contexts. Treat colour as an identification tool, not as a persuasion tool.
What colour increases sales the most?
No colour increases sales on its own. Tests that appear to show a winning button colour are almost always measuring contrast: the button that stood out from its background won, and the hue was incidental. Change the background and the winner changes with it.
Why do so many brands use blue?
Because it is safe, it reproduces well, and it carries fewer negative associations across cultures than most alternatives. That is also the argument against it. A colour chosen because it offends nobody tends to belong to nobody, which is why the categories most crowded with blue are the ones where a different colour is worth the most.
Does red really create urgency?
Red attracts attention faster than muted colours, which is a property of contrast and wavelength rather than of urgency. Urgency comes from what the red is attached to: a countdown, a stock count, a price that was higher yesterday. Remove those and the red does nothing.
Is colour meaning the same in every country?
No, and one example settles it. Red in Chinese culture carries prosperity and celebration, which is a reason a Chinese manufacturer builds an identity on it. The same red reads as danger, error or clearance across much of Europe and North America. Any colour rule that does not name its culture is incomplete.
How do I choose a brand colour?
Run four tests: does it contrast with the colours your category already owns, is it recognisable with the name removed, does it survive embroidery, one-colour print and a dark interface, and does text on it reach a 4.5:1 contrast ratio. A colour that passes all four is worth more than a colour with a nice story attached.