Product Data for Make-up: Get Shade Ranges and Variants Under Control

One foundation, forty shades: how make-up turns into a variant explosion, why no two brands name a colour the same way, and how attribute mapping keeps the whole range filterable.

Jakob Feinböck, ProductbayJuly 4, 20267 min read
☝️Key takeaways
  • Make-up is a shade problem: one foundation can ship in 40+ shades, each a separate variant with its own number, name, GTIN and image — a variant explosion, not a single SKU.
  • The naming is inconsistent across every brand: „230 Natural Beige“ vs. „N30 Sand“ vs. a bare hex value — so the same beige lands three times under three labels.
  • Standards like GTIN and eCl@ss identify and group products but carry no normalized shade taxonomy — undertone, depth and finish must be derived.
  • Productbay uses attribute mapping and AI to normalize every supplier's shade field onto one structure and group variants into families — so the whole range stays filterable.

One foundation. Forty shades. Each shade is its own sellable unit — with its own number, its own name, its own GTIN, and ideally its own swatch image. Multiply that across a lipstick range, a concealer line and a dozen brands, and a make-up assortment stops behaving like a catalog and starts behaving like a spreadsheet with a thousand near-identical rows.

Product data for make-up is a shade problem: the colour variants of a single product explode into dozens of records that no two suppliers name the same way. That's what separates make-up from the rest of a beauty range — skincare and fragrance are usually one SKU per product, but a foundation is a whole matrix. This is the sharpest sub-problem of the broader beauty & cosmetics data challenge.

Why do make-up shade ranges explode into a variant problem?

In most of a beauty catalog, one product equals one record. Make-up breaks that rule. A single foundation lives as a range of shades, and every shade is a full variant that has to carry its own data:

  • Its own identifier: a shade number (like „230“ or „N30“) and usually a shade name („Natural Beige“, „Sand“) — two labels for the same variant, neither standardized.
  • Its own GTIN/EAN: each shade is a separate sellable unit, so matching and updates happen at the shade level, not the product level.
  • Its own image or swatch: the customer buys the colour, so a range without per-shade visuals is unsellable — and images almost always arrive detached from the data feed.
  • Its own attributes: undertone (warm/cool/neutral), depth (light to deep), finish (matte, dewy, satin) — the very things a shopper filters by.

So one product balloons into 20–50 variants that all need to be grouped as one family, yet kept individually addressable. Handle that in Excel and a single foundation launch is an afternoon of copy-paste — before you get to the next brand.

Why does no two suppliers name a shade the same way?

There is no shared shade language in make-up. Every brand invents its own, and it arrives in your feed exactly as the brand thinks about it — not as your shop needs it:

  • Number-led: „230“, „30W3“, „N30“ — operationally clean, but meaningless to a shopper and impossible to sort without a mapping.
  • Name-led: „Natural Beige“, „Sand“, „Golden Honey“ — evocative, searchable, but wildly inconsistent between brands for the same real-world colour.
  • Visual-only: a hex value or a swatch image and nothing else, leaving the attribute blank entirely.

Import all of that unfiltered and your filter for „skin tone“ never groups anything, the same beige appears three times under three names, and a foundation range can't be sorted light-to-deep. The naming chaos is the reason the shade problem hurts — and it's a mapping problem, not a data-entry one.

Which standards help — and where do they stop?

The standards a beauty retailer relies on do real work at the identification layer, but none of them carries a normalized shade taxonomy. Here's the honest split:

Data layerWhat the standard deliversWhere it stops
Unique identificationGTIN/EAN identifies each shade as a sellable unitSays nothing about what the shade looks like
Category classificationeCl@ss groups make-up into product categoriesNo shade attributes (undertone, depth, finish)
Content aggregationICEcat / Open Icecat may carry brand copy for big linesThin for indie brands and fast-rotating shade drops
Shade namingNo standard exists; every brand names its own way
Variant groupingMust be derived from the supplier data itself

In short: GTIN and eCl@ss let you identify and slot a product, and that matters. What they don't give you is the shade structure — the undertone, depth and finish that make a range filterable, and the grouping that turns 40 loose records into one product with 40 variants. That has to be built from the messy supplier field.

How does Productbay solve make-up shade variants?

The core of it is attribute mapping plus variant grouping, run across every supplier at once — and that's what Productbay is built for:

  • Consolidate & match: import each source — supplier CSV, Excel, feed URL, FTP, API — and match at the shade level by GTIN/EAN, so the right variant updates instead of creating a duplicate.
  • Map the shade field: each brand's raw shade value (number, name or both) is normalized onto one internal structure — undertone, depth level, finish — so „N30 Sand“ and „230 Natural Beige“ can finally sit in the same filterable dimension.
  • Group into families: the 40 shade records of one foundation are grouped as a single product with 40 variants, each keeping its own GTIN, image and shade label.
  • Enrich with AI: derive missing undertone or finish, write consistent shade and product descriptions, translate via DeepL, and attach the right swatch image per variant — always with a review step before publishing.

The result is a make-up assortment where every shade is addressable, every range is sortable light-to-deep, and a shopper's „warm, medium, matte“ filter actually returns the right products. For the full beauty picture see the beauty & cosmetics overview; Productbay is built for specialist retailers running multi-supplier catalogs — from single shops to large chains.

Frequently Asked Questions

Let's look at your make-up data process

Shade matrices, inconsistent naming, one image per variant — make-up is the hardest corner of a beauty catalog. See how Productbay maps every supplier's shades onto one structure and groups the variants in a 30-minute walkthrough.

Get started