Pertex Fabric, PU Leather, and Luxury Curtain Fabric: The 9-Point Quality Checklist I Run Before Any Order Ships

2026-11-20 · Jane Smith

Who this checklist is for

If you sign off on first-article samples, approve fabric spec sheets, or release suppliers before a purchase order goes out, this is for you. Nine checks, in order. You can run the whole thing in about 90 minutes per SKU.

Some context on where it came from. Over four years I've reviewed somewhere north of 800 first deliveries — waterproof breathable shells (Pertex Quantum Pro and similar), PU leather components, and curtain fabric for hospitality and retail fit-outs. Different categories, same failure modes.

In 2023 our first-delivery rejection rate was 34%. Not catastrophic stuff. Small deviations that pass a hand feel and fail a production line. After I put this checklist in place, that number dropped to roughly 11% (as of Q1 2025, at least). The bigger win was time: first-article approval went from five days to two, because problems got caught before they became email threads.

Here's the list.

The nine checks

1. Confirm which product series you are actually sampling

Pertex is not one fabric. It's a family: Quantum, Shield, Equilibrium, Unlimited, and then sub-variants inside each. Quantum Pro is not the same as base Quantum. From what I've seen in spec sheets, Pro carries a PU coating for greater hydrostatic resistance and is used as a shell layer; base Quantum is softer and lighter, geared toward downproof linings.

I've received a PO that said "Pertex" and nothing else. The mill sent a Quantum cloth because Quantum was what they had in stock. Technically the right brand. Wrong fabric for the design. That order was usable, but we had to re-cost the whole jacket because the downproof rating didn't hold.

Before any order leaves your desk, write down: series name, weight in g/m², coating or finish type, weave construction, and color reference. If the finish matters to performance (DWR, calendering, lamination), name it explicitly.

2. Ask how the waterproof number was generated

You'll see "20K/20K" on a lot of spec sheets. The first number is usually hydrostatic head in millimeters, the second is moisture vapor transmission in g/m²/24h.

Here's the thing: the number means nothing without the method and the condition.

Hydrostatic resistance is tested to ISO 811 or AATCC 127. What matters is whether the result was measured before or after washing. A 30% drop after five home launderings is normal, not a defect. But if your supplier quotes the pre-wash figure and your customer wears the jacket in a downpour, you own that gap.

Put the requirement into the contract: test method, lab name, test date, washed or unwashed, and whether the result is fabric-only or includes seams. Seams are almost always the weak point on a finished garment, and almost never tested at the fabric stage.

3. Do not compare breathability numbers from different test methods

This one burns people constantly.

JIS L 1099 B1, JIS L 1099 B2, ASTM E96, and ISO 11092 are not measuring the same thing under the same conditions. One lab reports 15,000 on B1. A different lab reports 5,000 on B2 for the same roll of cloth. Both reports are honest. The numbers are not comparable.

So when a supplier says "15,000 breathability," the only correct response is: which method, at what temperature, at what relative humidity. If the report doesn't state all three, treat the figure as marketing copy, not data.

Practical rule I use: whatever method the buyer wrote into the spec, that method is the only one I will compare against. Everything else goes in the file and gets ignored for evaluation purposes.

4. Weigh the fabric. Do not trust the hand feel.

Lightweight series live and die on grams per square meter. "It feels light" is not a measurement.

Cut a circular sample, weigh it on a GSM cutter and balance. Every batch. It takes four minutes.

Set the tolerance in writing before you order, because mills interpret "40 g/m² nominal" generously. A ±8% band is common in the trade; some suppliers will argue for wider. Decide what your construction can absorb and put that number in the PO.

Then again, hand feel still matters for a different reason — see check 8.

5. Take PU leather apart before you approve it

What is PU leather, actually? A base fabric plus a polyurethane coating. Sometimes a release paper or embossing step on top for grain.

That definition is exactly why "PU leather" is useless as a purchasing specification. The base could be a 100 g/m² knit or a 300 g/m² woven. The coating could be water-based or solvent-based. The grain could be embossed or cast. Two suppliers can both ship "PU leather" and one will fail in month three.

I learned this the expensive way. First upholstery order, I specified color and "canvas-like texture." Sample arrived with a light knit backing. Our spec called for a woven base at roughly 400 N/5cm tensile. What showed up measured around 180. We rejected 4,000 cushions and the redo set the launch back six weeks.

Now every PU leather order states: base fabric construction, base weight, coating chemistry (water-based vs solvent-based), coating thickness in millimeters, total weight, grain method, and peel strength. Plus a smell test on arrival, because solvent-based coatings off-gas and that matters for anything that ships in a sealed carton.

6. Judge color in a light box, never in the office

Office lighting lies. Conference room lighting lies harder.

For brand-critical colors we hold the internal tolerance at Delta E ≤ 1.0. Around 2.0 the difference becomes visible to a trained eye. Above 4.0 most people will notice it without being told. Those thresholds aren't from a regulation, they come from our own brand standard, and yours may be tighter or looser — the point is to write the number down before you argue about a batch.

Two habits that save trouble:

7. Test dimensional stability before you cut anything

Curtain fabric is the most unforgiving category here, because the whole product is a large, hanging, gravity-loaded surface. Cut to size, wash or dry-clean, and the drop changes.

Reference ISO 6330 for home laundering procedure, or AATCC 135 if your market expects that route. Measure warp and weft shrinkage separately. For a luxury curtain quality we work to around 2% warp / 1.5% weft after the specified care cycle — that's our internal spec, not an industry floor, so set your own number based on panel size and heading type.

PU leather has a different dimensional risk: heat. It relaxes and deforms under the wrong storage temperature. Storage condition belongs in the spec too, not just the test report.

8. Compare every batch to a retained reference sample (the step most people skip)

This is the one I'd keep if I could only keep one.

Most teams check the first article carefully and then trust the rest of the run. Batch three is where quiet drift shows up.

We got hit in 2023. A shell fabric passed full spec on lots one through three. Paper conforming. But lot three felt different in the hand. Slightly stiffer, slightly less drape. Nothing failed. Then we started seeing a return rate creep from 2% to 5.5% on finished jackets. Took two quarters to trace it back to the cloth.

Now every approved batch gets a retained swatch: labeled, dated, sealed, filed. Each incoming lot gets held next to it under the light box and compared by hand and eye. Fifteen extra minutes. It has caught three drifts since.

Honestly, the simple act of having a physical reference changed how suppliers talk to us. They know lot four gets compared to lot one.

9. Close the documentation file before you release the order

Last step, and the least glamorous.

You want: dated test reports from an accredited lab, lot traceability, fiber content, country of origin, and care labeling. If the product sells in the US, textile fiber content and origin disclosure fall under 16 CFR Part 303. That's not optional.

Watch the environmental language. Under the FTC Green Guides (16 CFR Part 260), claims like "recyclable," "eco-friendly," or "sustainable" need competent and reliable evidence, and the claim has to match how the material is actually used. If a quote sheet says "eco PU leather" with no third-party documentation behind it, that phrase does not go on your product page. Ever.

File it all together. If a customer complaint lands eighteen months later, the paperwork is your only defense.

Four mistakes that cause most rejections

Trusting the quote sheet instead of the test report. Sales documents describe intent. Lab reports describe the batch you're actually receiving.

Comparing numbers across test methods. See check 3. It happens in almost every failed supplier negotiation I've sat through.

Never looking under the coating of a PU leather. The base is where the structural failures live. If you don't know the backing, you don't know the material.

No retained samples. Without a reference swatch, "slightly different from last time" is a feeling, not an argument.

Look, none of this is exciting. It's a weigh scale, a light box, a filing cabinet, and a checklist. But it's the difference between finding the problem at receiving and finding it after 8,000 units are already in a container. Not glamorous. Effective.

Leave a reply