Lab notebookThe method behind every rule
Some of this has strong evidence and some is tradition. Each claim carries a seal that says which.
The lightness gap between your eyes, lips, brows and skin has strong support in studies of how faces are seen.
The rest of the answer
The idea that a season suits you comes from color consultants, not from a proven rule. I ask for no photo, I put a seal on each rule, and I say what the system can't do.
Five layers, five levels of evidence
Baseline you
one lever, not five
0 of 4 done
StrongTradition
Why, and how solid
Four of them describe your baseline. The fifth is the impression you choose.
1Colorlight or deep, clear or soft, cool or warm
- What it reads
- How light or deep, how clear or soft and how cool or warm your coloring is, filed into one of 12 seasons plus a neighbor.
- How sure
- Strong as a way to measure color.
Tradition for saying a color suits you. - When
- The free test and the guides open together.
2Contrasthow far your features sit from your skin
- What it reads
- How far your eyes, lips and brows sit from your skin in lightness. Hair counts less. Five levels: low, low to mid, mid, mid to high, high.
- How sure
- Strong as a visible signal.
Tradition for matching clothes to it. - When
- The free test and the guides open together.
3Linesoft or defined edges
- What it reads
- How soft or defined the edge of what you see reads.
- How sure
- Tradition I found no study of it as a style axis.
- When
- Beta, 2027.
4Scaledelicate to bold
- What it reads
- How far each element reads, from delicate to bold.
- How sure
- Tradition I found no direct study of it.
- When
- Beta, 2027.
5Chosen imagethe impression you want to give
- What it reads
- The impression you want to give, picked from eight mood words. You choose it, and I don't diagnose it.
- How sure
- Tradition The definitions come from consultants, and the demand is clear.
- When
- Free test from November 2026.
The evidence seals
StrongStudies agree
It means Peer reviewed studies, with consistent results.
For example Contrast between your features and your skin is a visible signal, and makeup changes it.
ModerateSome studies
It means Some studies, or reasonable indirect evidence.
For example A white base can leave a pale film on deeper skin.
TraditionPractice, not proven
It means Stylist practice without studies. Useful, not proven.
For example The 12 seasons as a frame for shopping.
MythContradicted
It means Contradicted or without basis. I explain it so you can skip it.
For example The wrist vein test. Gold or silver as a test of undertone. Deep skin means Autumn.
Our policyA promise
It means A promise about how I work, not a claim about color.
For example The season swap. No photo, ever.
Each rule has three parts: what to do, why, and a seal.
StrongModerate
Why, and how solid
What each seal means, with an example
Each rule on this site has three parts: what to do, why, and a seal. If a rule has no why I can explain, it doesn't go in. If the why is tradition, I say so. The seal sits next to the rule it belongs to, and a rule that mixes a fact with a tradition is split into two lines.
| Seal | What it means | An example |
|---|---|---|
| Strong | Peer reviewed studies, with consistent results. | Contrast between your features and your skin is a visible signal, and makeup changes it. |
| Moderate | Some studies, or reasonable indirect evidence. | A white base can leave a pale film on deeper skin. |
| Tradition | Stylist practice without studies. Useful, not proven. | The 12 seasons as a frame for shopping. |
| Myth | Contradicted or without basis. I explain it so you can skip it. | The wrist vein test. Gold or silver as a test of undertone. Deep skin means Autumn. |
| Our policy | A promise about how I work, not a claim about color. | The season swap. No photo, ever. |
If I can't point to a specific study behind a seal, I lower it one level. If a seal changes, I update the guide and say so on the site.
If a rule has no why I can explain, it doesn't go in.
Strong, or only Tradition
- 1
- 2
- cheek3
- 4
- 5
- 6
- lips7
- 8
- 9
StrongTradition
Features against your own skin
Compare your eyes, lips and brows with your own skin, in lightness.
Why, and how solid
The rule: Compare your eyes, lips and brows with your own skin, in lightness.
Why: studies of how faces are read find that the lightness gap between those features and the skin around them is something people see, and makeup changes it. Because I measure the gap against your own skin, the method is built to read deep, olive, medium to tan and fair skin the same way. That part is my design, not a finding.
Strong Several peer reviewed studies agree on the signal. Who they cover is in the limits box below.
Undertone is a starting point
Test two colors next to your face, one change at a time.
Why, and how solid
The rule: Treat "this color suits your undertone" as a starting point, and test two colors next to your face, one change at a time.
Why: I found no study that tests whether a season guess is right. The studies on clothing color measure what observers like, not what a color does to your skin. Changing one thing at a time and repeating a test is a standard way to measure, so that part holds.
Tradition Color consultants teach this. The idea that a season suits you comes from them, not from a proven rule.
Seasons are a language
Use the 12 seasons to file three readings together, not as a box you belong to.
Why, and how solid
The rule: Use the 12 seasons to file value, clarity and temperature together, and not as a box you belong to.
Why: skin color varies smoothly between people, so many readers sit near a line between two seasons. That is why a result can say between two, and why the swap covers any season.
Tradition The 12 seasons are a language from color consultants, and not a law of biology.
Why, and how solid
The short list, as rules.
Why no photo
A photo carries the room's light and whatever the camera and filter did.
Strong
that tint is the lamp.
but a window and a gray strip can help.
Why, and how solid
A photo carries the color of the light in the room and whatever the camera and the filter did to it. That is why the same face can get different seasons from different apps. A home test can't control light either, so each comparison here is relative: you compare your face to a fixed gray strip and to printed cards, in daylight by a window, and your card says how sure the result is.
Test in daylight, facing a window, with no direct sun on your face, and compare instead of trusting a photo.
Why: a photo's color depends on the light and the camera, and a filter changes it again. A comparison next to your face, in daylight, changes one thing at a time.
Strong Changing one variable at a time and repeating a test is a standard measuring principle.
How the gray strip works
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
The strip is nine neutral grays, from near white to near black.
Tradition
Why, and how solid
How it is built, how to use it, and the table
The strip is nine neutral grays, from near white to near black. Each step is about 11 units darker than the one before on L*, the lightness scale used in color science, so one step is about the same visual distance anywhere on the strip. Step 1 is not pure white and step 9 is not pure black on purpose, so the white of the paper and the black of a printer stay outside the scale.
You hold the strip beside your cheek, lips, brows and eyes, and pick the step closest in lightness. The test reads how many steps separate each feature from your cheek. Because it counts steps against your own skin and not against a norm, it is built so that a gap of two steps means the same on deep skin as on fair skin.
| Step | Name | Hex | Lightness (L*) |
|---|---|---|---|
| 1 | Near white | #F6F6F6 | 96.9 |
| 2 | Very light | #D7D7D7 | 86.0 |
| 3 | Light | #B9B9B9 | 75.1 |
| 4 | Light-mid | #9B9B9B | 64.0 |
| 5 | Mid | #7F7F7F | 53.2 |
| 6 | Mid-deep | #636363 | 42.0 |
| 7 | Deep | #494949 | 31.0 |
| 8 | Very deep | #303030 | 19.9 |
| 9 | Near black | #191919 | 8.8 |
A cream background changes how grays look, so the strip on the kit page sits in a white frame.
Hold the strip flat beside the spot you are measuring, in daylight by a window, and squint a little. Pick the step closest in lightness, not in color.
Why: daylight from a window is more even than a lamp, and squinting blurs hue so you compare light and dark, which is what the strip measures.
Tradition The method is mine. The step spacing rests on L*, which is standard.
What this system can't tell you
This system has not been validated yet, and that includes the gray strip method.
Why, and how solid
What has not been checked
I found no study that checks whether a home color test lands on the right season, and none on how often color analysts agree with each other. The 2027 pilot below is how I plan to measure it.
What the test does not do
- It doesn't ask for a photo and it has no upload field.
- It doesn't diagnose you. It gives a likely season, a neighbor and a note on how sure the result is.
- It can't control your light, your mirror or your screen, so each comparison is relative and your card says how sure it is.
- It doesn't tell you how a shade will look on you. Screens and light change that, and the guides link swatches so you can compare.
- It doesn't adjust anything by skin tone. Temperature counts for less only when your own answers contradict each other, and then the card says neutral-leaning and shows both neighbors.
- It doesn't score gold or silver. No study backs it as a test of undertone, so it is a Myth as a diagnosis.
- It doesn't read line or scale yet. Those are a 2027 beta, and they are not part of the free test or the guides.
- It doesn't describe who you are. It describes how colors and outfits read.
- This does not say how people will treat you. Studies document bias in how people are treated by looks, skin color, natural hair and age, and I found no study where a color or an outfit changes it.
- It isn't for anyone under 18, and it is not medical or dermatological advice.
Where the evidence is thin
- Most of the contrast studies I found cover women who are white, Chinese, Latin American or Black South African, and I found no study with South Asian or Middle Eastern women.
- Tests like this one were mostly built with lighter skin in mind, and deep and olive skin are where they tend to miss. That is why this one can say you sit between two seasons instead of forcing a single answer, and why I don't promise accuracy.
How I plan to measure it in 2027
In 2027 I plan to ask 60 to 100 readers of different skin depths and backgrounds to take the test twice and help me measure it. Skin depth and background are volunteered by the reader, and the test doesn't ask for them. The pilot will measure:
- Whether the same reader lands in the same place on two different days, in different light.
- Whether a friend answering about you lands near your own answers.
- How far the gray steps move between a screen and a printed strip. If they move by more than one step, I will drop the strip and measure with objects and clothes.
- How many readers land in the middle of each scale. If more than half do, I rewrite the questions.
- Whether any result leans on a stereotype for a group. If one does, I pull the question and publish the correction.
- Whether a result sounds like you, on a scale of 1 to 5.
Results by skin depth will be published with their sample sizes, and without claiming more than the numbers show. If line or scale don't hold up, I will drop them or fold them into another layer, and say so.
I will publish what the pilot finds, including the results that don't flatter it.
Products and pictures
Products are chosen by color
Chosen by color for the season, each linked once per page.
Why, and how solid
I don't write reviews from use and I don't post my own swatches.
The rule: Read the exact shade name on the store page, and open the swatches the brand or reviewers show before you buy.
Why: brands rename and drop shades, and a swatch depends on the arm, the light and the camera, so several swatches let you judge for yourself.
Our policy A promise about how I work, not a claim about color.
Swatches stay exact
Swatches and palettes are flat blocks in their exact color. A picture on a product card shows the shade and is not a photo of that product.
Why, and how solid
The paper texture and the live washes are made with code. The painted style is the frame, and the swatch is the data.
The rule: Read a picture as the shade, and the code beside it as the color.
Why: a picture changes with the screen and the light, and the code beside it does not.
Our policy A promise about how I work, not a claim about color.
Try it yourself
About 6 minutes, no photo, no email.
