In 2021, I thought the category label was the spec

When I wrote my first eyeliner sourcing spec sheet, I started with one line: "pencil eyeliner." Line two: "waterline eyeliner." Line three: "winged eyeliner." I thought that counted as doing the work.

Three years and roughly 40 first-article inspections later—we've rejected about one in five deliveries—I now think a good chunk of those rejections were my own spec's fault. Not the supplier's. Mine. We received a batch of "winged eyeliner" where the wing tail feathered to a 1.5mm width on the swatch card. The vendor said it met "industry standard." Technically, sure, it was a pencil. Functionally, it was not the line we'd promised anyone.

That's the shift I want to put down here, because I think it's the one most inside-brand buyers are still walking through slowly:

Eyeliner sourcing has moved from "product category" to "failure scenario." The buyers still ordering their supply chain by "winged," "pencil," "brown," and "waterline" are buying hope.

Category labels are for marketing. In a spec sheet, they're close to worthless.

Argument 1: Winged eyeliner is not a product—it's an execution outcome

Take what "winged eyeliner" actually means. Fine line. Clean tail. No feathering on the edge. Fast dry-down so the tail doesn't transfer onto the eyelid. Film flexibility so it bends a little instead of cracking like chalk.

That's at least four separate specs right there:

  • Tip geometry—a 0.5mm twist-tip and a felt-tip pen write completely different tail shapes
  • Formula rheology—too thin wicks, too thick drags
  • Film flexibility—hard cracks, soft smears
  • Dry time—somewhere between 2 and 15 seconds and that's a wide band

None of those four can live inside "winged eyeliner." So when the supplier quotes you "winged" with a broad category SKU, they're quoting an execution envelope you and they are both assuming you share.

We got hit with this in 2023. A supplier repoured our existing pencil formula into a so-called "winged" package. The tip was fine. The formula wasn't flagged. The wing swatches feathered against our internal Q3 2024 standard on 11 of 40 samples. We rejected the whole 8,200-unit run. They redid it at their cost. Every winged eyeliner contract since then carries an explicit tip and rheology line.

Same logic applies to eyeliner styles as a group. Natural, smoky, cut crease, inner waterline—the industry has names for all of these. Each one is a different specification package for failure. A natural look tolerates imperfect fine-line precision. A wing doesn't. Cataloguing them as five variants of "eyeliner" is how buyers end up specifying for one style and reconciling against another.

Argument 2: Waterline eyeliner and pencil eyeliner are not the same spec

This one is more counterintuitive. Because the shapes look adjacent, procurement people often buy them as "the same category."

They are not.
Waterline eyeliner formulas need to handle:

  • Application to mucosa or near-mucosa—lower migration tendency, eye-mucosa tolerance assessment
  • High-humidity resistance to sweat without transfer into the tear film
  • Anti-clumping behavior after extended wear

Pencil eyeliner formulas—including traditional wood pencils, automatic pencils, and gel pencils—need to handle:

  • Different wax bases
  • Sharpen/delivery behavior entirely
  • A higher hardness floor so the structure doesn't fail under pressure

I'm not a formulation chemist. I can't tell you why one wax base behaves better than another—but from an inspection standpoint, I can tell you which failure modes stop me at first check.

If you buy both as "eyeliner," the supplier will probably cover both with a single formula platform and tell you it's fine. Given three-to-four runs out of ten, you'll be right. The rest of the time, you'll be reconciling it into the next quarter.

Brown eyeliner has the same problem. A "brown" color number across runs is not visible by eye on paper—until you put pigment load and base structure into the spec. We got burned by a supplier with two different factories producing the "same brown" at a measured ΔE of about 4 against our own 1.5 tolerance. Same label. Different tinting shop. Different batch.

Argument 3: 2020 vs. 2025

This isn't just my observation. The whole game of eyeliner sourcing has shifted in the last few years.

In 2020, demand was predictable, custom color was a luxury line item, and buying on the category label plus the price worked out well enough. In 2025, the same buyers are asking for custom brown, for different tip geometries, for waterline-plus-winged dual-use packaging—and specifications, spelled out.

We ran a small audit on our own SKU line last year (Q3 2024 data). Our spec sheets went from an average of 6 lines to an average of 19. First-article rejections dropped from about 32% to under 18% as a result. We didn't change vendors. We just changed what we were asking for.

That's the real point I want to land. The industry isn't evolving because the supply chain got better. It's evolving because buyers who've stopped trusting "winged" as a spec are pushing it there.

According to the EU Cosmetics Regulation (EC) No 1223/2009 (current consolidated text available at ec.europa.eu), eye-area products fall under stricter preservative and irritancy provisions than general skin-contact cosmetics. Verify current requirements at the official EU database—this changes with annex amendments.

"But this is slow and awkward"

I get it. Writing a 19-line spec is a whole lot more work than picking "pencil eyeliner" off a catalog line.

But ask yourself which one is actually slow. You spend 45 minutes writing down a functional winged spec once, or you spend two weeks cleaning up an 8,200-unit redone batch while explaining to the brand team why the line that was promised looks like it was drawn with a wet brush.

I don't want to overstate this. This worked for us, but we're a mid-size B2B brand with predictable ordering and a settled vendor list. If you're sampling a new category, or dealing with big seasonal spikes, some of these lines can wait—the calculus might be different, and I'd write the priorities differently.

And some people will say, "But my supplier guarantees it." Sure. A guarantee is not a spec. Only what's written into the contract counts as a spec.

Category labels still have a place—just not in the spec

I'm not saying "winged eyeliner" should be removed from a catalog. It's fine. It works on a marketing page. It works in a buying conversation. That's how customers talk.

The spec layer is a different thing. The spec layer needs three things:

  • A use scenario—where and how the line lives
  • A failure mode—which condition counts as failure
  • A tolerance range—how far it can drift before it's dead

"Winged eyeliner" covers none of those three. All three together are what a supplier can actually work with.

So the first line of my spec sheet no longer names a category. It names a requirement: "Fine line, dry under 8s, no transfer on re-application, 38°C/85% RH for four hours, edge feathering ≤0.3mm, 3mm tail holds without recoil."

Different beast. And no, "winged" doesn't cover it.