I review eyeliner samples most weeks. Buyers ask me the same handful of questions over and over — usually after a first delivery has already gone sideways. So no long intro. Seven questions, seven straight answers, sourced where I can source them and pulled from our own audits where I can't.

Quick context on who's answering: I'm the quality and brand compliance manager at a beauty company doing private label and white label work. Every eyeliner that carries a customer's name goes through me first — roughly 200 SKUs a year across pencil, kajal, waterline, and liquid formats. In our 2025 review cycle, about 17% of first-article samples got rejected. Not because the suppliers were bad. Because the spec was vague.

1. What specifications should you lock before you ask for a quote?

Start by being honest about what you're actually buying. Eyeliner sourcing is really four sourcing jobs wearing one name: a wooden pencil, a retractable pencil, a soft kajal stick, and a liquid liner with a felt or brush tip. They share a category name and almost nothing else in manufacturing terms.

For a pencil eyeliner, the specs that bite you later: core diameter (to 0.1 mm), wax-to-oil ratio, pigment load, wood type and sharpener compatibility, total length. For kajal eyeliner, it's softness rating and smudge window. For waterline eyeliner, it's the preservative system and which colorants you're permitted to put near the eye. For liquid, it's tip type, flow rate, and whether the reservoir is fiber or valve-driven.

In my first year doing this, I made the classic rookie mistake: I assumed "standard pencil" meant the same thing to every vendor. It doesn't. One supplier sent 2.0 mm cores, another sent 3.5 mm. Both were "standard." That cost us about $2,200 in retooling and two weeks of schedule.

2. Why is brown eyeliner so hard to keep consistent batch to batch?

Because brown isn't one color. It's a blend of iron oxides — red, yellow, black — and the ratio drifts. A batch that reads warm brown in a viewing booth looks gray under office lighting and orange in a phone photo. Black forgives this. Brown eyeliner doesn't.

We now run every brown through a color check against a master standard under D65 and warm-white illumination, with a Delta-E tolerance written into the contract. Usually ≤1.5 for bulk, ≤2.0 on the consumer-visible surface. That number stopped being negotiable for us in 2024.

Then there's the cost question. A cheaper pigment blend might save $0.11 a unit. On a 40,000-unit run that's $4,400. Sounds like free money — right up until the second shipment doesn't match the first and the retailer returns the whole launch. The math I run: is $4,400 worth a restock fee and a delisted SKU? Every time I've asked that out loud, the answer was no.

3. What has to be tested on a waterline eyeliner?

More than on any other format, because this sits on mucous-membrane-adjacent skin. At minimum: a preservative efficacy test run in your actual packaging (not a lab jar), accelerated stability at 40°C/75% RH for 8–12 weeks, a repeat-insult patch test on human panelists, and an ophthalmic safety assessment for the colorants used.

In the EU, colorants near the eye fall under Regulation (EC) No 1223/2009 — Annex IV lists what's permitted, and "permitted in cosmetics" is not the same as "permitted in the eye area." Check the annex, not the supplier's summary slide. If a supplier can't tell you which annex entry their colorant sits under, that's already your answer.

One more thing. Waterline pencils are softer by design. Softer means more prone to breaking inside the barrel. We do a three-point bend test on 30 samples per lot and reject anything that snaps below spec. Sounds excessive until you ship 20,000 pencils that break on the first sharpening.

4. What's different about sourcing winged eyeliner and liquid formats?

The precision claims live or die on the applicator, not the formula. Felt tips wear. Brush tips splay. Valve systems clog if they sit. Ask for the applicator manufacturer's name — if the factory won't tell you, treat that as a flag, because the applicator often has a longer lead time than the fill itself.

For winged eyeliner styles specifically, what buyers underestimate is the drying curve. Fast-dry gives crisp lines but can crack. Slow-dry gives control but smudges. It's a real trade-off and someone has to decide which one your customer actually wants. Your brand isn't all things to all users — pick.

And on claims: per FTC advertising guidance (ftc.gov), claims like "smudge-proof," "24-hour wear," or "waterproof" have to be truthful and backed by competent evidence. That means testing, documented, on file, retrievable. We started requiring claim substantiation reports with every launch in 2023. Two suppliers pushed back on it. We dropped both.

5. Which batch tests should never be skipped?

Short list. Accelerated stability. Micro testing on the finished fill. Core-to-barrel adhesion on pencil formats. Payoff and laydown panels across at least three skin depths — a shade that looks rich on light skin frequently goes chalky on deep tones, and that's a formulation problem, not a marketing one.

Honestly, though, the test that has killed more of my launches than any formula issue is decoration adhesion. We had a 12,000-piece run where the printed logo rubbed off with a fingernail. Supplier called it "normal wear." It wasn't — the packaging spec called for a tape pull test at 4B minimum under ASTM D3359. They'd tested against nothing.

6. What do most specification sheets leave out?

Packaging interaction. Almost every time.

Three things routinely missing: cap pull-off force (specify it in newtons), print or logo adhesion on barrel and cap, and the fit tolerance between the sharpener cap and the pencil body. If the cap releases at 4N when your spec says 8N, it opens in a makeup bag. If it takes 20N, your customer breaks a nail on first use. Both failures come from the same blank line in the spec sheet.

Also write down the acceptance method, not just the requirement. "Tape pull test" means nothing. "3M 610 tape, 90-degree pull, 4B minimum per ASTM D3359" means something — something a third-party lab can fail a batch against.

7. What do buyers not think about that they should?

SKU creep. Every addition to your eyeliner styles list — a new brown, a slightly different winged tip, matte versus satin finish — means another pigment batch, another line clean-down, another QC setup, and often a fresh MOQ that only pays off if that exact color actually sells.

The numbers said trim the line to seven colors; they covered 94% of sell-through. My gut said keep all fourteen, because every one of them had someone internally asking for it. I went with the numbers. Two years out, the cut SKUs had produced zero measurable lost revenue, and we'd freed up roughly $18,000 in inventory carrying cost plus a shorter changeover queue at the factory. Occasionally the spreadsheet is right. Rarely this cleanly, but occasionally.

If there's one thread through all seven answers: write more down than feels necessary, and write it in a language a lab can actually test against. Everything else is negotiation.