A foundation is not one product. It is one base formula and a set of colour recipes, and the arithmetic of foundation shade range development with a Chinese OEM lives in the gap between those two things. The base is developed once; each shade is a separate colour match, a separate approval, a separate filling changeover, and — depending on how the plant counts its minimum — possibly a separate 2,000-unit commitment. A twelve-shade launch is not one order with twelve colours. It is twelve small orders that share a kettle.

This article covers how shades are built from pigments, how a match is judged and by what tolerance, what a shade actually costs you in minimum-order terms, how many shades to launch first, and which questions to put to the plant before development begins. It stays on the manufacturing side; regulatory status of colourants in your destination market, and any claim about how the product performs on skin, are your obligations as the brand owner.

How a shade is built: one base, four pigments

Almost every foundation shade in the world is made from the same four colourants: titanium dioxide (CI 77891) for opacity and lightness, and three iron oxides — yellow (CI 77492), red (CI 77491) and black (CI 77499) — for hue and depth. The base formula (a W/O silicone emulsion, an O/W emulsion, an anhydrous stick or a cushion fluid) carries them. A shade recipe is the ratio of those four pigments to the base.

That has a consequence buyers do not always expect: the base and the shade range are developed on different clocks. The base is a formulation project — texture, wear, stability, pack compatibility — and once approved it is frozen. The shades are colour projects layered on top. If the base changes after shades are matched, every shade drifts, because the same pigment load in a different emulsion reads differently on skin.

Two details from the compounding room that decide whether shades stay consistent:

  • Pigment pre-dispersion. Iron oxides and titanium dioxide agglomerate. Dropped straight into a kettle they give a grainy, streaky shade that changes as it is stirred. Plants that do colour seriously grind pigments into a dispersion — on a triple-roll mill, a bead mill or by buying pre-milled pigment pastes — and then dose the paste into the base. Ask which of those three this plant uses. The factory page for Yadun Cosmetics lists base makeup within its first product line, plus compounding and a lab among plant functions, but does not publish dispersion equipment. Pin it down.
  • Coated versus uncoated pigments. Surface-treated pigments (silicone, amino-acid, or other coatings) disperse better in W/O and silicone bases and resist colour shift with humidity. They also cost more, and they are not interchangeable with uncoated grades — a substitution changes the shade. Specify the pigment grade on the formula sheet, not just the colour index number.

How a match is judged, and by what tolerance

You will send the plant a reference — a competitor product, a physical swatch, or a description — and receive a sample back. Whether it "matches" is a measurement question, and the method should be fixed before the first sample, not argued about after the third.

Method What it measures Strength Weakness Ask the plant
Visual, uncontrolled light Impression Fast Metamerism — matches under one light, fails under another Not acceptable as the sole method
Visual, light booth (D65 / store light / incandescent) Colour under defined illuminants Catches metamerism Still observer-dependent Which illuminants the booth has
Instrumented, spectrophotometer, ΔE Colour difference in CIELAB Repeatable, contractable Reads a drawdown, not skin Instrument, ΔE formula, tolerance
Skin drawdown panel Colour on a range of skin Real-world Slow, subjective How many skin tones, who reads it

The instrumented method is the one you can write into a contract. Colour is measured on a drawdown — a film of the product applied at a fixed thickness on a card — and compared with the reference as a ΔE value. Across the colour-cosmetics trade a ΔE of around 1 is treated as a tight match and something near 1.5 as an acceptable production tolerance; the specific number is negotiated and depends on the ΔE formula used (ΔE*ab and ΔE2000 give different values for the same pair). Set the formula, the instrument and the tolerance on the specification, and ask which of them the plant can actually measure. Spectrophotometer capability at this plant is not published.

Three insider points on matching:

  1. Match wet-to-dry, not wet-to-wet. Foundation changes colour as the water or volatile silicone evaporates — the "dry-down" shift, usually darker and warmer. A sample that matches the reference in the jar can be a shade off after ten minutes on skin. Approve on a dried drawdown at a stated time, and record that time.
  2. The reference has to be physical and current. Colour descriptions and screen images are useless. A competitor product bought at retail is a real reference, but check its batch date; mass-market foundation shades are reformulated more often than their names change.
  3. Mass tone versus skin tone. The colour of product in bulk (mass tone) is not the colour on skin. Two shades can look identical in the jar and diverge on a drawdown because of opacity differences. Always judge on the drawdown.

What a shade costs you in MOQ terms

This is the number that decides the whole range, and it is the number most plants do not volunteer.

Yadun's published minimum is 2,000 units. What is not published is whether that 2,000 applies per finished SKU (each shade a separate 2,000), per formula (2,000 spread across shades sharing one base), or somewhere between — for example a per-shade floor set by the smallest kettle charge. The three readings produce very different launch budgets:

MOQ interpretation 6-shade launch 12-shade launch 20-shade launch
2,000 per shade 12,000 units 24,000 units 40,000 units
2,000 per base, shades split freely 2,000 units 2,000 units 2,000 units
2,000 per base with a per-shade floor (illustrative: 500) 3,000 units 6,000 units 10,000 units

The middle row is what buyers hope for and rarely get in full, because each shade is a separate compounding charge: the base is made once, but pigment dosing, mixing, colour check and filling changeover happen per shade, and there is a minimum bulk quantity below which the kettle cannot homogenise properly. Many plants land on the bottom row in practice — a formula-level minimum with a per-shade floor — but the floor is a per-plant number. Ask for it in writing and negotiate from there; the CMH guide to MOQ negotiation with Chinese factories covers the general tactics, and the shade floor is where they apply.

Two further cost lines that appear per shade rather than per range:

  • Packaging components. Bottles, pumps and caps are usually the same across shades, but any shade-coded component — a colour-coded cap, a printed label per shade, a shade sticker — carries its own component minimum, often larger than the fill minimum. Check the component MOQ before you decide to print the shade name on the bottle rather than on a label.
  • Sampling rounds. The factory quotes three days for a standard formulation sample and seven for a custom one. A shade match is a custom sample, and a twelve-shade range is twelve of them, each of which may need two or three rounds. Plan the development calendar in weeks, not days, and read the CMH article on the formulation and sampling process at Chinese cosmetics OEMs for how rounds are normally structured and charged.

How many shades to launch, and in what order

The manufacturing answer and the marketing answer differ, and the manufacturing one should be heard first.

  • Start with the base, one shade. Develop and approve the base on a single mid-range shade. Stability, pack compatibility and dry-down behaviour are properties of the base; test them once, then freeze it.
  • Match anchor shades next. Three to five anchors spread across light, medium and deep with distinct undertones (cool, neutral, warm) establish the pigment ratios. Every intermediate shade is then an interpolation between anchors, which the colourist can do far faster than a fresh match.
  • Fill the gaps last. Intermediate shades are added by ratio, sampled, and checked against the drawdown standard. This is where a range goes from six to twenty without restarting development.
  • Build the range on a skin-tone scale, not on guesswork. Ranges designed against a measured lightness axis — the Individual Typology Angle (ITA°) is the usual dermatological scale — distribute shades evenly instead of clustering in the middle. Ask whether the plant's colourist works to a scale or to "light, medium, dark".

The factory page describes the plant's 2,000-unit minimum as suited to brands testing several SKUs before committing to volume. In shade-range terms that means: launch the anchors as your pilot, sell through, then extend. A first order of twenty shades from a plant you have not worked with before commits you to twenty colour approvals and twenty changeovers before you know whether the base sells.

Keeping shades consistent from lot to lot

A range is only as good as its second production run. Colour drift between lots is the most common complaint in base makeup, and the controls are cheap compared with a recall:

  1. Golden standard per shade. A sealed retained sample of the approved shade, stored cool and dark, that every production lot is compared against — by instrument, on a drawdown, at the stated dry-down time. Ask how long the plant retains standards and whether they are refreshed.
  2. Pigment lot approval. Iron oxides vary by lot. A plant that colour-checks each incoming pigment lot against the previous one, and adjusts the shade recipe by tinting rather than by dumping the batch, is the plant you want. Ask to see the incoming-inspection record for a pigment.
  3. Tinting in-kettle. Most lots are adjusted at the end of compounding by adding small quantities of pigment paste to bring ΔE inside tolerance. Ask whether this is done, who authorises it, and whether the adjustment is recorded on the batch record.
  4. Lot code and batch record with the shipment. Each shade, each lot, traceable to a batch record that shows the pigment lots used. Without it a shade complaint six months later cannot be diagnosed.

These controls are what the plant's QC function should be doing; the format and the lab are listed among this plant's functions, but the procedures are not published, so the questions are the evidence.

Common questions

Can the plant match a competitor's shade exactly?

Close enough to pass an instrumented tolerance, usually. An exact match is limited by pigment availability, by the base — a different emulsion gives a different dry-down — and by the age of the reference. Approve on ΔE and on a skin drawdown, not on the word "exact".

Does the base type change the shade work?

Yes. Silicone-based W/O foundations, the most common long-wear format, need coated pigments and shift less with humidity; O/W formulas are easier to compound but can shift more on dry-down. If the same range is planned in more than one format — a fluid and a stick, say — each format needs its own shade matching, because the pigment load that gives a colour in one base gives a different colour in another. The companion article on sunscreen and primer formats covers how those base systems differ on the filling line.

What does "ODM shade library" mean, and does this plant have one?

Some plants hold a set of house foundation bases already matched into a range of shades, which you can brand as-is or extend. That shortens development considerably. Whether this plant holds such a library is not published; ask for the shade cards and the base formula codes if it does. The broader landscape of colour cosmetics OEM in China describes how house libraries are typically offered.

How do I brief a shade without a physical reference?

Send one. If you genuinely have none, ask the plant to build anchors against a lightness scale and supply drawdowns on contrast cards plus skin swatches on a panel, then select from those. Written descriptions produce a first sample that is wrong in a way nobody can measure.

What to ask the supplier next

Put these to the plant before shade development starts, and get the answers on the quotation:

  1. Does the 2,000-unit minimum apply per shade, per base formula or per finished SKU? If per base, what is the per-shade floor in units and in bulk kilograms?
  2. How are pigments dispersed — triple-roll mill, bead mill, or purchased pastes?
  3. Which pigment grades (coated or uncoated) will be specified on my formula sheet?
  4. Do you have a spectrophotometer? Which ΔE formula, and what tolerance do you propose?
  5. What light booth illuminants do you approve shades under?
  6. What drawdown thickness and dry-down time will approval be judged at?
  7. Do you hold house foundation bases or an existing shade library, and can I see shade cards and formula codes?
  8. How long do you retain golden standards, and how are incoming pigment lots approved?
  9. How many sampling rounds per shade are included, and what does an additional round cost?
  10. Which shade-coded packaging components would carry their own minimum?

Yadun Cosmetics is a fair example of the type of plant these questions are written for: a Baiyun District contract manufacturer operating since 2009 with 100–500 staff, listing base makeup alongside sunscreen in its first product line, offering both OEM and ODM terms, with a 2,000-unit minimum and custom sampling quoted at seven days. The factory profile gives the scope and the terms; the per-shade answers above are what turn a scope into a launch budget.