I'm the quality and brand compliance manager at a mid-size color cosmetics brand. I review every batch that leaves our contract manufacturers before it reaches a retail buyer or a distributor — roughly 240 SKU batches a year, mostly eye and face products.

In 2024, we rejected 11% of first deliveries. Not because the factories were bad. Because what arrived didn't match what was agreed.

Below is the checklist I run before approving a makeup primer sourcing order. Seven steps, in order. It applies whether you're bringing in an eye makeup primer, a face makeup primer, or a face primer with SPF under your own label.

The 7-Step Pre-Approval Checklist

Step 1: Write your own spec sheet before you request a single quote

This is the step people skip because it feels slow. It isn't. It's the cheapest four hours you'll spend on the project.

Your spec sheet needs, at minimum:

  • Texture and viscosity range — not 'lightweight,' but a number or a reference product you can point to
  • Finish: matte, satin, dewy, or clear
  • Tinted or untinted, with a color reference if tinted
  • Fill weight and net content, with tolerance
  • Primary and secondary packaging format
  • SPF claim, if any, and the market you're selling into
  • Shelf life and PAO (period after opening) you're willing to print on the label

Most factories will send you their own spec sheet for their standard primer base. To be fair, that's useful — it tells you what they can make cheaply and reliably. But it's their spec, not yours. If you approve a batch against their sheet, you've agreed to their tolerances, not the ones your brand needs.

What I mean is: the spec sheet is the spine of the contract. Everything downstream — sampling, testing, inspection, reorder — hangs off it.

Step 2: Test physical samples the way your customer will actually use them

Swatching on your wrist under office lighting tells you almost nothing about an eye makeup primer. The eye area behaves differently: thinner skin, more movement, constant contact with eyeshadow, liner, and whatever your customer layers on top.

For an eye primer makeup sample, I want:

  • Eight or more hours of wear, ideally on at least three people with different skin types
  • Layered under the specific shadow and liner formulas you plan to sell it with — not a generic test shadow
  • A creasing check at hour four and hour eight, photographed in the same lighting

For a face makeup primer sample, the variables change: how it sits under foundation, whether it pills, whether it alters the finish of the foundation on top. A primer that looks great alone can make your best-selling foundation separate.

Ask about the sample's production context too. Was it made on the pilot line, or hand-filled in a lab? If I remember correctly, one of our worst mismatches came from a hand-filled lab sample and a production run of the same formula — same formula, completely different texture on skin.

Step 3: Lock packaging color and print before the pilot run

This is where a lot of brands lose money quietly. The formula is fine. The carton looks slightly off, and nobody catches it until it's on a shelf next to your other SKUs.

Industry standard color tolerance is Delta E < 2 for brand-critical colors. Delta E of 2–4 is noticeable to trained observers; above 4 is visible to most people. (Reference: Pantone Color Matching System guidelines)

Put the Delta E target in the packaging purchase order. Not in an email. In the PO.

Two more print-side numbers worth writing down:

  • Commercial offset printing needs 300 DPI at final size. Large-format display pieces viewed from a distance can run lower, but your carton artwork is not that piece. (Industry-standard minimum for offset.)
  • Board weights convert roughly: 80 lb cover ≈ 216 gsm, 100 lb cover ≈ 270 gsm. Conversions are approximate, so specify the gsm you need rather than the lb equivalent if your printer works in metric.

Why does this matter for primer sourcing specifically? Because the unit is small, the print area is small, and small print hides registration problems until you're staring at 10,000 of them in a row.

When I compared our Q1 and Q2 cartons side by side — same printer, same artwork file, different board supplier — I finally understood why the approval step exists. The Q1 box looked fine alone. Next to Q2, it looked grey.

Step 4: Get the SPF test report, and check which formula it belongs to

This is the one most buyers skip, and it has the largest downside.

If you're selling a face primer with SPF, the SPF number on the carton is a regulated claim. It needs a test report. The report needs to reference the exact formula you're buying — same ingredients, same percentages, same manufacturing process. A report for the factory's roughly similar base does not transfer.

It's tempting to think an SPF 30 primer is just a primer with sunscreen mixed in. But the SPF value depends on which UV filters, at what percentage, dispersed in what base — and a change as small as a preservative swap can invalidate the test. I've watched a supplier send a 2022 report for a 2024 formula and genuinely not understand why we pushed back.

Ask for: the test lab name, the test method, the report date, the formula code, and written confirmation that the formula code matches the one on your specification. If the market you're selling into requires registration or notification, check the current rules directly with the regulator — requirements change, and I'm not the person to give you a legal read on it.

Step 5: Run the cost math at your real order quantity, not the sample quantity

Unit price is one line. It's rarely the line that decides the project.

Ask for, in writing:

  • MOQ per SKU, and MOQ per component — the cap, the bottle, the carton, the shipper
  • Fill weight with overfill allowance. Filling 15.5 ml instead of 15 ml across 10,000 units is 5,000 extra ml of product you're paying for
  • Tooling, plate, or setup fees, and whether they repeat on reorders
  • Lead time for the first run and for reorders, quoted separately

Had two hours to decide on a rush reorder once. Normally I'd have gone back and requoted the carton, but there was no time. We approved the existing PO and paid a setup fee twice in one quarter. In hindsight, I should have pushed the ship date by ten days.

Step 6: Approve a first article, not a first batch

Ask for photographs or video of the first 20–30 filled units off the line, with a fill weight reading, before the run ships. Some factories will do a pre-production sample instead. Either works. What doesn't work is finding out at the port.

Check against your spec sheet: fill weight, closure torque, label placement, carton color, batch code format, and whether the claims copy on the carton matches what was approved.

Sampling catches formula problems. First-article inspection catches assembly problems. They are not the same test.

Even after we signed off on the first article, I kept second-guessing. What if the units in the photos weren't the ones being packed? Didn't relax until the container arrived and the fill weights matched the sheet.

Step 7: Put the reorder terms into the contract

The first order is the easy one. The fourth order is where quality drifts.

Include a sealed, dated golden sample held by both parties; the color tolerance in Delta E; the fill weight tolerance; the requirement that SPF documentation be reissued if the formula changes; and a written rejection and rework process that says who pays for what.

Fast, consistent, cheap. Pick two — and write down which two you picked.

Common mistakes that cost the most

Approving the sales sample as the production standard. Sales samples are often hand-filled or pulled from a different line. If you don't know how the sample was made, you don't know what you approved.

Not keeping a golden sample. Without a sealed reference, every future dispute becomes a memory test. Keep one at your office and one at the factory, both signed and dated.

Verbal spec confirmations. 'Yes, we can do 300 DPI' is not a spec. Neither is 'standard tolerance.' Get the number.

Assuming the SPF report covers your formula. It usually covers the one they tested, which is not always the one you're buying.

Five minutes of verification beats five days of correction. Every time I've skipped a step on this list, it was because the timeline was tight — and every time, the fix took longer than the step would have.

Pricing references and regulatory notes here are for general guidance only. Actual costs vary by vendor, specification, and order volume, and cosmetics requirements differ by market — verify current rules with the relevant authority before you commit to a claim.