ColorSense Blog
We Asked 106 People What's Hardest About Color
Cohesion and mood were 81.6% of the answers. Accessibility was 5.1%. This article examines what the gap may mean.
We asked 106 people who signed up for our color palette guide a simple question: what do you find hardest about color? Ninety-eight answered. Finding palettes that go together and getting the right mood were the two hardest problems, totaling 80 responses, or 81.6%.
Accessibility was near the bottom. Five people out of ninety-eight, 5.1%, chose it. Only the catch-all "Other" category ranked lower, with four responses.
That is an uncomfortable result to publish, because accessibility is the thing this company is built around. We check contrast. We score palettes. It is most of what we talk about. And the people who answered our question ranked it below every specific color problem on the list.
We think that result is worth publishing anyway, and we think it is misleading. Not wrong, misleading. This article is about the difference.
What we asked, and who answered
Before the numbers, the limits, because a number without its method is not evidence.
106 people joined the waiting list for the ColorSense Palette Playbook, a guide of contrast-tested color palettes. At signup they were asked three optional questions. Between 97 and 98 answered each one, a response rate of roughly 92%.
These are not designers. They are not a random sample of anyone. They are people who chose to sign up for a free color resource, which means they already care about color enough to give an email address for it. That self-selection matters and we will come back to it.
Every percentage below is calculated against the number who answered that specific question, not against 106. Where we say 46.9%, we mean 46 of 98.
This is a small dataset. It is also a real one, collected from real people rather than assembled from other people's articles, and as far as we know nobody else has published it.
What people said was hardest
| What is hardest about color | People | Share of 98 |
|---|
| Finding palettes that go together | 46 | 46.9% |
| Getting the right mood or feel | 34 | 34.7% |
| Brand consistency | 9 | 9.2% |
| Accessibility | 5 | 5.1% |
| Other | 4 | 4.1% |
Two answers took almost everything. Cohesion and mood together account for 80 of 98 responses, 81.6%.
Accessibility took 5.
That is a ratio of 16 to 1. For every person who said accessibility was their hardest color problem, sixteen said their problem was that colors would not sit together properly, or would not produce the feeling they wanted.
Why the survey does not show that people dismiss accessibility
One possible reading is that people do not care about accessibility. We do not think that is what this survey shows.
Nobody was asked whether accessibility matters. They were asked what they find hardest. Those are different questions, and the difference is the entire finding.
Accessibility is not hard in the way cohesion is hard. Cohesion is hard because you can look at five colors, feel that something is wrong, and have no idea which one is causing it. There is no obvious next step. You move things around until it stops bothering you.
One possible explanation is familiarity. When you already know what the text says, you can overlook weak contrast that slows down or blocks a first-time reader. That makes an accessibility problem easier to miss during your own review, so the design can appear finished before it has been tested.
In this interpretation, accessibility may not show up as a hard problem. It may appear to be no problem until someone says they could not read the design.
A problem you do not perceive may not feel difficult. It may feel absent.
How cohesion, mood, and accessibility may overlap
Here is the part we did not expect when we looked at the results.
We see a possible overlap between the 46 people who said their palettes would not go together, the 34 who could not get the mood right, and the 5 who named accessibility. These answers may describe different symptoms of palettes whose colors do not yet have clear roles or enough separation. That is our interpretation of the pattern, not something the survey directly tested.
Think about what "these colors do not go together" can mean in practice. Two of them may be fighting for the same job. Or everything may have the same weight, so nothing leads. Or there may be no color dark enough to sit type on and no color light enough to give the eye a rest.
Each of those can indicate a structural problem. Fixing the structure can create clearer light and dark roles, which makes readable combinations easier to choose. It does not prove that a pair passes WCAG contrast. The intended text and background still need to be measured.
One reason the problems can feel different is that disharmony may be noticeable immediately, while weak contrast can be overlooked during a familiar design review.
That can lead people to solve the visible issue, publish the work, and discover the readability problem later.
Why the mood answer deserves more attention
Thirty-four people, 34.7%, said the hardest part was getting the right mood or feel. That is the second largest answer and it is the one most likely to be dismissed as vague, so it is worth taking seriously.
Mood is not only a property of one color. It also emerges from relationships within the finished design.
A single color can carry associations, but the feeling of a finished design also depends on relationships: how much of each color appears, how far apart the colors sit in lightness, and which one dominates the composition.
You can see this effect quickly. Take a palette described as warm and calm. Now use the brightest color for eighty percent of the surface instead of five percent. The colors have not changed, but the result may no longer feel calm. Nothing about the hex values moved.
This is why mood can feel slippery. People may try to solve it by swapping colors when proportion, hierarchy, or contrast is part of the problem. Too little separation can make a palette feel flat or murky. Too much can make it feel loud even when the individual swatches look soft.
Our interpretation is that cohesion, mood, and accessibility can sometimes share a structural cause: colors without distinct roles or enough separation to perform those roles. The survey did not test that mechanism directly, so the raw counts and our interpretation should be read separately.
That suggests one possible structural cause with three different symptoms, some easier to notice without measuring than others.
Where people actually get their palettes
The second question was where palettes come from. Ninety-seven people answered.
| Where palettes come from | People | Share of 97 |
|---|
| Pinterest | 31 | 32.0% |
| I wing it | 24 | 24.7% |
| Google | 14 | 14.4% |
| A tool like Coolors or ColorSense | 13 | 13.4% |
| Other, including Threads | 15 | 15.5% |
Pinterest is the single largest source, at nearly a third.
This may explain part of the result. A Pinterest palette is not a finished palette system. It is often a photograph, or a screenshot of a photograph, from which five colors have been pulled. It usually has no assigned roles. Nothing in it is designated as the text color, the background, or a color to use sparingly. Looking good in a thumbnail grid is a different use case from working in a layout at reading size.
You can absolutely start from a Pinterest board. A useful palette can begin with something you liked. The gap is that a board may hand you five colors and no instructions, leaving the colors without assigned roles when they move into a layout.
A quarter of people say they wing it
Twenty-four of 97, 24.7%, answered "I wing it".
We want to be careful here, because that answer is easy to read as an admission of incompetence and it is not one. Picking colors by feel is a normal way to start, and good work can come out of it.
But winging it is not necessarily a skill gap. It can be a lack of repeatable structure. Someone who wings it may be making five simultaneous decisions with nothing to anchor them, starting from zero each time. That can make a successful result harder to reproduce and a failed result harder to diagnose.
That is why the first two answers dominate. Without a repeatable way to make decisions, cohesion is genuinely hard.
If you want one of these checks a week, the kind that takes a minute and saves a rebuild, that is what ColorSense Weekly is. One email a week, Fridays.
Almost nobody uses a color tool
Add it up: 84 of 97 people, 86.6%, do not primarily use a palette tool.
They use Pinterest, or Google, or nothing. Only 13 named a tool, and that group includes people who named ours while signing up for our own guide.
For a company that makes color tools, this is the most useful line in the dataset. Our interpretation of this sample is that the main alternative is not another palette tool. It is a Pinterest board, a Google search, or choosing by feel.
Half of this is personal work
The third question was what people use palettes for. Ninety-eight answered.
| What palettes get used for | People | Share of 98 |
|---|
| Personal projects | 49 | 50.0% |
| Branding | 19 | 19.4% |
| Social or content | 12 | 12.2% |
| Web or UI | 12 | 12.2% |
| Other | 6 | 6.1% |
Exactly half of respondents named personal projects.
Not client work. Not deliverables. A room, a wedding, a zine, a game, a shop that has not opened yet. This work may receive less formal review than paid work, so the person making it may be the main quality check.
That reframes the accessibility answer again. Personal projects can still reach other people, but a readability problem may remain hidden when nobody else reviews the work before it is shared.
What this finding changes
This survey does not point to one universal formula for choosing colors. It points to a better first question: is the problem cohesion, mood, readability, or an undefined role? Naming the failure keeps you from replacing every color when only one relationship needs attention.
If the whole set feels unstable, Palette Health checks the relationships across a palette. If a specific text and background pair is in question, use the ColorSense WCAG Contrast Checker and compare the result with the thresholds in our complete WCAG color contrast guide. Our guide to accessible color combinations for buttons shows how the same test applies to common interface states.
If your starting point is a photograph or mood board, the ColorSense color picker gives you exact values so you can test the real colors instead of estimating them from a screenshot.
The distinction matters because a palette can fail in more than one way. Two colors can look pleasant beside each other but offer too little contrast for text. A high-contrast pair can remain visually disconnected from the mood of the rest of the set. A palette can also contain strong individual colors but lack a clear dominant color, leaving every element to compete at the same level.
Those failures call for different tests. Cohesion is judged across the whole set and in the proportions used in a layout. Readability belongs to a specific foreground and background pair under WCAG 2.2. Mood is evaluated in context, where the same swatches can feel calm or intense depending on scale and emphasis. Treating all three as a vague problem of choosing better colors makes revision slower because it hides which relationship needs to change.
This is also why the low accessibility response should not be used to dismiss accessibility. The survey measured what people spontaneously named as difficult. It did not measure whether their work contained contrast failures, whether their audience could read it, or whether they considered accessibility important. The result is evidence about perceived difficulty, not evidence about the real frequency or importance of inaccessible color choices.
What we are changing because of this
Two things, stated plainly because we would rather be held to them.
We talk about accessibility too early. Leading with contrast ratios to an audience whose stated problem is cohesion is answering a question nobody asked. The contrast argument is more persuasive when it arrives as the reason your palette keeps falling apart, not as a compliance obligation.
And we have been writing for professionals when half of respondents in this sample named personal work. Those are different readers with different stakes and the same underlying need.
The honest limits of this data
We would rather state these than have them pointed out.
This is 106 people who signed up for a color guide. It is not a survey of designers, not a sample of any industry, and not representative of anything beyond itself. Self-selected and self-reported.
Single choice. People were asked to pick one hardest thing. Somebody who finds three things equally hard is represented once. The real distribution is almost certainly flatter than the table looks.
The options shaped the answers. "Accessibility" was one word on a list. Presented differently, as "people telling you they cannot read it", it would very likely have scored higher. That is partly our point, and it is also a limitation.
We are not neutral. We make color tools, and we published a result that flatters our thesis. Judge it accordingly. The raw counts are all here so you can disagree with our interpretation while using our numbers.
Keep learning with ColorSense Weekly
Get one practical color lesson each week, with useful palettes, design examples, and ideas you can apply to your next project. One email a week, Fridays.
FAQ
How many people were surveyed?
106 people joined the ColorSense Palette Playbook waiting list and were asked three optional questions at signup. Between 97 and 98 answered each question, roughly a 92% response rate. All percentages are calculated against the number who answered that specific question.
What do people find hardest about color?
Finding palettes that go together, named by 46 of 98 people, 46.9%. Getting the right mood or feel was second at 34 of 98, 34.7%. Together those two accounted for 81.6% of all answers.
Why did accessibility score so low?
Accessibility was named by 5 of 98 people, 5.1%. We do not read this as evidence that people do not care. The question asked what is hardest, not what matters. One possible explanation is that familiarity with a design can make weak contrast easier to overlook during a creator's own review.
Where do most people get their color palettes?
Pinterest, named by 31 of 97 people, 32.0%. A quarter, 24 of 97, said they wing it. Only 13 of 97, 13.4%, said they primarily use a palette tool.
Is a Pinterest palette good enough to use?
Treat a Pinterest palette as source material, not a finished system. Extract the exact values, decide where the palette will be used, and test the relationships that matter in that context, especially text against its intended background.
What should I do when a palette feels wrong?
Name the problem before replacing colors. Check whether the palette lacks clear roles, enough light-to-dark separation, the intended mood, or a readable text-and-background pair. Then test that specific relationship with a palette or contrast tool.
Can I cite this data?
Yes. Please attribute it to ColorSense and describe it accurately as 106 people who signed up for a color palette guide, with the relevant denominator, rather than as a survey of designers.
Try it now
Check any color pair against WCAG — free, no signup
Pick a foreground and background, see the exact contrast ratio plus AA/AAA verdicts for normal and large text. One click auto-fixes a failing pair while keeping your hue.
Open WCAG Contrast Checker →