Short answer: liquid concealer is an emulsion problem, contour is a pigment-loading problem, and a stick is a compaction and drop-test problem. All three can be made on the same site, but they sit in different MOQ tiers and fail for different reasons — so pick the format before you pick 24 shades.
Concealer is the category where briefs most often arrive over-specified in one dimension and under-specified in another. A brand sends a reference photo, a 30-shade grid and a target price, but no decision on whether the product should be a fluid in a doe-foot applicator or a cream in a pot. Those two products share a name and nothing else: one is an emulsion with a pigment load around 8–15% that must not separate in a thin film, the other is a structured anhydrous or water-in-oil cream that must hold a peak in the pan and still melt on contact.
This guide walks through the three formats as manufacturing systems, then the shade architecture that determines your SKU count, then the process and test points where projects actually slip.
Three formats, three manufacturing systems
The table is the decision view. The sections after it explain the trade-offs that the table compresses.
| Format | Base and structure | Typical pigment load | MOQ tier | Defining test | Production window |
|---|---|---|---|---|---|
| Liquid concealer | Oil-in-water or water-in-silicone emulsion | 8–15% | Emulsion: from 1,000 pcs/SKU | Shade drift and syneresis after 3 months at 40 °C | 10–15 days |
| Cream concealer or contour | Anhydrous or water-in-oil, structured with waxes and fillers | 15–30% | Emulsion: from 1,000 pcs/SKU | Pay-off versus creasing; pan stability | 10–15 days |
| Stick concealer or contour | Wax-based, high solids | 20–35% | Stick: from 3,000 pcs/SKU | Drop test and sweat resistance at 45 °C | 10–15 days |
Two things follow from this. First, the stick tier is three times the emulsion tier, so a 12-shade stick launch is a materially larger commitment than a 12-shade liquid launch. Second, the same shade number does not produce the same colour across the three bases: pigment load, film thickness and the refractive index of the base all shift the apparent depth. If you plan to launch a liquid and a stick in "the same" 12 shades, budget for two approvals, not one.
Shade architecture decides the SKU count before chemistry does
Most concealer ranges are built as a depth × undertone grid: four to six depths, each with a neutral, a warm and a cool variant, plus one or two correctors. That is 12–20 SKUs from a single base. Each SKU is a separate batch and a separate stability sample, so the grid is a cost decision as much as a merchandising one.
Three practical rules we use:
- Build the grid on depth first, undertone second. A range that spans light to deep with only neutral undertones serves fewer faces than a five-depth range with three undertones, and it costs less to produce.
- Fix the titanium dioxide to iron oxide ratio before matching the first shade. TiO2 drives opacity and the white cast; iron oxides drive depth and warmth. Changing that ratio halfway through the grid forces every earlier shade to be re-matched.
- Approve under two light sources. D65 and A-light, side by side against a retained Golden Sample. A shade that matches in daylight and goes grey under warm retail light will come back as a return, not as a complaint you can answer.
For contour, the grid is usually six shades or fewer, and the failure mode is different: too much pigment reads as dirt on fair skin, too little reads as nothing on deep skin. A contour range built from the same iron oxide blends as the concealer, with a lower TiO2 level, keeps the undertone family coherent.
Pigment dispersion is where the money goes
A concealer that looks right in the lab and separates in the bottle is almost always a dispersion problem, not a formula problem. Pigments arrive as agglomerates; the job of milling is to break them down and wet them out so that every particle is coated by the oil phase.
- Three-roll milling suits high-viscosity pastes and gives good colour strength, but throughput is low and heat build-up can degrade heat-sensitive actives.
- Bead milling suits fluids and gives tighter particle size distribution, at the cost of a cleaning step between shades that must be validated — cross-contamination between a deep and a fair shade shows up as a grey cast.
The measurable acceptance criterion is colour difference against the approved standard: we work to ΔE ≤ 1.0 on the finished bulk, measured on a drawdown rather than on the bulk surface. Anything looser drifts visibly across a 3,000-piece run. Dispersion time, mill temperature and the order of addition all belong in the batch record, because a shade that is right on batch one and wrong on batch six is usually a process record that was not followed, not a pigment that changed.
The emulsion side: water-in-silicone versus oil-in-water
Water-in-silicone gives the dry, non-migrating film that most long-wear concealers aim for. It also brings three constraints: the silicone phase must be built at a controlled temperature, electrolytes in the water phase can break the emulsion, and preservation is harder because the continuous phase is not water. Oil-in-water is more forgiving to manufacture and easier to preserve, but it creases more readily under the eye unless the oil phase is built with volatile carriers and film formers.
There is no universally better choice. There is a better choice for your claim:
- Claim is long wear and transfer resistance → water-in-silicone, with a film former and a volatile carrier.
- Claim is hydrating and comfortable → oil-in-water with a modest pigment load and a humectant system.
- Claim is both → expect two rounds of stability testing, because the two claims pull the base in opposite directions.
Emulsion basics, including phase ratios and homogenisation, are covered in our emulsion manufacturing explainer; the point here is that the base decision should be made before the shade grid is signed off.
Process: hot pour, cold blend and filling temperature
Cream and stick formats are hot-poured: the bulk is held above its wax melting point, filled into the pan or the stick shell, then cooled under controlled conditions. Cooling rate determines crystal structure, and crystal structure determines whether the stick glides or drags.
- Fill temperature too high → pigment settles and the top of the stick is leaner than the bottom.
- Fill temperature too low → the bulk thickens in the nozzle, fill weights scatter, and the surface shows flow marks.
- Cooling too fast → brittle wax crystals, and a stick that fails the drop test at the tip.
Liquid concealers are cold-filled, which moves the risk to viscosity and to the applicator. A doe-foot applicator that delivers 0.12 g per dip needs a bulk viscosity that lets the product flow back into the reservoir between uses; too thick and the reservoir empties unevenly, too thin and the product runs off the foot.
The test panel that predicts returns
Stability and compatibility testing is not a formality for this category, because the product is applied to thin, mobile skin and carried in a small package with a high surface-to-volume ratio.
| Test | Condition | What it catches |
|---|---|---|
| Accelerated stability | 3 months at 40 °C, plus 3 freeze–thaw cycles | Syneresis, shade drift, pH movement |
| Package compatibility | Filled unit stored inverted and upright at 40 °C | Applicator swelling, label lift, leakage |
| Preservation challenge | ISO 11930 protocol | Under-preserved water-in-silicone bases |
| Wear and crease panel | 8-hour wear on 12–20 panellists | Creasing under the eye, oxidation on oily skin |
| Drop and heat test (stick) | Drop from 1 m, 45 °C for 24 h | Wax structure and tip strength |
Our stability testing guide covers the design of these panels in more detail, and the preservative challenge explainer covers why a water-in-silicone base cannot be signed off on a preservative level alone.
Where projects actually slip
Four patterns account for most of the delays we see on colour cosmetics:
- Approving shades before approving the base. Every shade matched on a trial base has to be re-matched when the base changes. Fix the base first, then the grid.
- Matching a reference texture without matching its package. A cream that performs in a jar can fail in a thin-walled stick shell, because the package determines the thermal history and the fill weight.
- Ignoring fill tolerance at high pigment load. A 30% pigment cream is dense; a ±3% fill tolerance on a 3.5 g unit is a visible difference in how many applications the customer gets.
- Underestimating shade approval rounds. Two rounds are typical, three are common for deep shades where the undertone is hardest to hit. Each round is a lab batch plus a wear panel.
What to send in the brief
A brief that gets a firm quote in 24 hours contains: the format and the package type, the shade count and the depth range, the reference product with its texture and finish (not just its photo), the claim you intend to make, the destination market, and the annual volume per shade at MOQ. With those five inputs, the base, the pigment system, the MOQ tier and the production window can be settled in one reply.
Samples follow the standard route: 7 days including two rounds of modification, against a retained Golden Sample, before the bulk order is released. The wider cosmetics manufacturing range shows the formats we run in the same plant.
FAQ
Q: Can a liquid and a stick concealer share the same shade numbers? A: They can share the numbers but not the formula. The stick carries a much higher pigment load and a wax structure, so each shade must be matched separately against the same reference. Budget for two approvals.
Q: Why does a water-in-silicone concealer separate in the bottle? A: Usually because the silicone continuous phase was built outside its temperature window, or because electrolytes in the water phase were not accounted for. It is a process and phase-balance issue more often than an emulsifier issue.
Q: What pigment load gives opaque coverage without a cakey film? A: Most liquid concealers sit between 8% and 15% total pigment, with titanium dioxide doing the covering and iron oxides doing the depth. Above that range the film becomes visible in fine lines unless the base is rebuilt.
Q: How many shades can be run in one batch cycle? A: Shade count does not change the per-batch MOQ; each shade is its own batch from 1,000 pcs for emulsions or 3,000 pcs for sticks. A 16-shade range is 16 batches, which is why the grid is a cost decision first.
Q: Does the applicator change the formula? A: Yes. A doe-foot reservoir, a paddle and a brush each impose a different viscosity and yield point. Confirm the package before the bulk viscosity is fixed, not after.