The Chromallogy State of Complexion · 2026

The industry
as one line

We ran our shade audit on every foundation and concealer shade in the category, pooled into a single range. Here is what it surfaces. And what it would surface on yours.

3,144
shades audited, across 90 brands
202
doing genuinely unique work
18/100
the category's score, graded as one range
The whole field, one grade

One range. Every shade. Still an 18.

Pooled into a single range, every foundation and concealer shade the category has produced still reaches only a fraction of real skin. The shortfall is shared rather than particular, which means everyone is missing roughly the same territory, in the same direction, at once. What follows is the anatomy of that single 18, read along the one axis that matters, the full depth of real skin.

The shape of it

Every audited shade, in one space

Plot all 3,144 shades in color space and the problem takes on a shape. The whole category crowds into one warm, light region of the map, and the open volume around the cloud is the territory almost nobody is making. This is the industry's own shades, in their true colors. It is not our reference map.

Light ↑   Deep ↓ Drag to turn

Each point is one real audited shade, positioned by its color and shaded in that color. Vertical is depth, light at the top. The emptiness is the gap.

Depth

A range built for the top of the face

Nearly half of every shade ever made sits in the two lightest bands, while the three deepest bands hold less than a quarter between them. The range thins exactly where representation has been contested for decades, which makes this the oldest gap in complexion and, somehow, still the most common.

Very Light
0
Light
0
Light-Medium
0
Medium
0
Medium-Deep
0
Deep
0
Very Deep
0
0 of all shades fall in the two lightest bands.
0 is all that is left for the three deepest bands combined.
Undertone

One direction, presented as a range

A range built well carries a real spread of undertone at every depth. This one defaults instead. The light half runs 93.6% warm, then the deep half inverts into mostly cool and neutral, so undertone never really travels down the range so much as trades one default for another around the middle.

We call it the undertone cliff, and it is the quietest failure in complexion, because the bottle is always on the shelf. It simply matches nobody precisely, having been built toward a direction rather than for a person.

Cool
Neutral
Warm
Very Light
7
33
887
Light
3
29
533
Light-Medium
4
28
413
Medium
2
39
299
Medium-Deep
16
91
197
Deep
43
83
123
Very Deep
79
87
51

Intensity shows concentration of shades, not skin color. The warm column owns the light half of the range and thins toward the bottom, where cool and neutral finally appear. That inversion is the cliff.

Chroma

Light shades too loud, deep shades too flat

Chroma is intensity, measured against how saturated skin actually reads at a given depth, and the range gets it backwards at both ends. Light shades run louder than skin calls for, while deep shades are built too flat to sit correctly on skin. The deepest customers are underserved twice, first by scarcity and then by miscalibration, and it surfaces as returns and the specific complaint that a shade went ashy.

Very Light
skin
Light
Light-Medium
Medium
Medium-Deep
Deep
Very Deep
Built louder than skin reads Built flatter than skin reads Where skin actually reads
The gaps

Every omission points the same way

We name gaps the way a person describes their own skin. Do that across the whole category and one thing is impossible to miss: every unserved region points warm and golden. Ninety brands guessing independently all missed the same territory, which is less bad luck repeated than a shared blind spot in how the category decides what to make. Ranked by the coverage a single well-placed shade would add, the most valuable correction available is, without exception, warm, golden, and deep.

Golden, warm, deep skinDeep · unserved
0
Golden, warm, medium-deep skinMedium-Deep · unserved
0
Golden, warm, medium-deep skinMedium-Deep, high intensity · unserved
0
Golden, warm, medium skinMedium · unserved
0
Golden-rose, warm, deep skinDeep · unserved
0
Golden, warm, very deep skinVery Deep · unserved
0

Coverage gain is the movement in the 0–100 score from adding one well-placed shade to that region. On your own line, this same ranking arrives with the placements attached.

Redundancy

202 shades doing the work of 3,144

The category did not under-produce. It over-produced, in the wrong places. Only 202 shades do unique work, while the rest sit close enough to something already on the market that a colorimeter cannot tell them apart. It added thousands of shades and closed almost none of the gap, because the new shades kept landing on top of the ones it already had.

0 / 0shades doing work no other shade does

Every one of the 3,144 audited shades, in its true color, sorted light to deep. Hover any shade to inspect it. Most of what you see is a near-duplicate of something already beside it.

r = 0.00
shade count vs. coverage, across the audit

The category spent eighty years answering how many shades to make, and produced 3,144 of them and an 18. Adding shades, on the evidence, is if anything faintly associated with covering worse, because the new ones keep landing where the range is already crowded.

The question was never how many. It was which, and for whom.
What you'll receive

Two numbers tell you where you stand. The audit tells you what to do.

Everything on this page is the category read: aggregate, anonymous, structural. On your own range, the same method runs at two depths. One is free and tells you where you sit. The other is the plan.

Start here · Free

Shade Snapshot

Two scores, back within 72 hours.
Coverage
/100
Industry Grade
A–F
  • Where your range sits against the category
  • The one number a retailer will eventually ask for
  • No cost, no commitment, no data required

Pricing is set to scope on a short call. Start with the free Snapshot and decide from there.

The category read is on this page. The rest is your range.

Start with the free Snapshot, or book a conversation.

Book a conversation
Appendix · A note on the numbers

New work, the same method

Earlier this year we powered the SeeMe Index, which scored brands one at a time and averaged the results to an industry figure near 33. This is separate, ongoing work on a larger and still-growing dataset, run with the same method. It reads lower because it is measured differently. Rather than averaging brands, it pools every shade into one range and grades that range against the best you could do with that many shades.

0industry average, scored brand by brand
0the field pooled into one range

A forty-shade brand is judged against the best possible forty-shade range. The industry, holding more than three thousand, is judged against something close to perfect coverage of every skin tone that exists. The category has far more than enough shades to cover everyone. It scores an 18 because those shades pile into the same light, warm corner instead of spreading across the map. That is the whole argument, sitting inside one number.