The email landed at 6:40 on a Tuesday morning in March 2024. Subject line: “Photos attached — please advise.”
I was eleven days out from a furniture install on a 214-room hotel refresh, and my upholstery supplier had just dropped 340 yards of chenille fabric that didn't match the swatch we'd approved back in October.
I opened the photos on my phone. Warehouse lighting, phone camera, bad angle. I could still see it.
That one email turned into a nine-day schedule hit, an $18,400 replacement run, and a permanent rewrite of how our firm writes fabric specifications. It also settled an argument I'd been having with myself for about four years.
The spec looked great on paper
We had around 61,000 square feet of soft goods moving through that project — guest rooms, lobby, two restaurants, a rooftop bar. The design team wanted chenille for the lounge seating and banquettes, and honestly, I got it. Chenille has that dense, soft hand that reads as expensive in a rendering and photographs even better.
I pulled five candidates out of a chenille fabric catalog and sent them to a third-party lab for abrasion and lightfastness testing. Two failed our minimums. Abrasion is measured by ASTM D4157 — the Wyzenbeek method. Hospitality seating usually needs to clear 30,000 double rubs, and anything within a few feet of a bar or a luggage path, I won't sign off under 50,000. Lightfastness runs through AATCC 16 and is measured in hours; our public-space minimum is 60.
The winner was a chenille from a mill we'd used before, sourced through their OEM program. Good price (the chenille fabric OEM route usually comes in 15–25% under a branded equivalent, because you're paying for the yarn and the weaving, not for the name on the back of the swatch). Good lead time. Solid lab numbers.
Then we got to color.
The project carried a brand blue — Pantone 286 C — and our design director cares about it more than she cares about almost anything else. So we wrote a tolerance into the purchase order: Delta E under 2. For context, industry standard color tolerance for brand-critical colors is Delta E < 2. A Delta E of 2 to 4 is noticeable to a trained observer. Above 4 is visible to most people. Reference: Pantone color matching guidelines.
At the time, writing a Delta E number into a PO felt a little obsessive. I'd like to publicly rescind that opinion.
Then the bulk yardage arrived
Lab dip approved October 2023. Bulk shipped March 2024. Our technician read it against the retained standard and came back with Delta E 3.8.
Not catastrophic. Not invisible either. On a 24-yard banquette, 3.8 means a guest sitting in the lobby at 4pm can see two slightly different blues depending on which side of the room they're looking at. On a brand-standard hospitality project, that's a fail.
The mill's response was, and I'm quoting from memory here, “that's within industry standard.” Which is true if you're selling generic goods and not true if you're selling to someone who wrote Delta E < 2 on the paperwork and signed it.
We rejected the 340 yards. They redyed at their own cost. We ate nine days.
And while all that was happening, the mockup room handed us problem number two: after about eight weeks of real use, the banquette at the lobby bar was pilling at the seat edge.
Here's something vendors won't tell you: chenille pilling isn't a defect. Chenille is a pile yarn — the short fibers that give it that plush hand are the same fibers that migrate and ball up under friction. You can finish it better. You can spin it tighter. You can blend it. But if you put a chenille seat where people slide in and out a hundred times a day, you're going to see pills eventually. It's a trade-off you're choosing, not a flaw you're discovering.
The phone call where somebody told me no
Mid-project, I sat down with the team at Carnegie Fabrics LLC to talk through the high-traffic portion of the spec. I'd been circling Carnegie Fabrics for a while, mostly because spec sheets kept landing on my desk with the same handwritten note from three different designers: look at Xorel.
I expected a pitch. What I got was a boundary.
“We're not going to tell you chenille is a bad fabric — it's a great fabric for the right seat. But anything within three feet of an entrance, a bar top, or a rolling suitcase, we'd put Xorel there instead. And if you want a custom print on a 100% chenille base at a 500-yard minimum, that's not our strength. Here are two mills that do it really well.”
I remember sitting there thinking: you just handed part of my order to somebody else.
That's exactly the point.
What they were steering me toward was the Carnegie Fabrics Xorel specification — a proprietary high-performance woven, polyethylene-based and spec'd as PVC-free, cleanable with bleach and water. It's been in healthcare corridors and airport lounges for decades for a reason. It's not soft like chenille. It doesn't have that plush hotel-lobby hand. Honestly, the first time I put my palm on a sample, I wasn't excited. It felt technical. Industrial, even.
Then we did the side-by-side.
Our mockup suite had two identical banquettes — same frame, same foam, same distance from the window. One in chenille, one in Xorel. Nine months of rotating staff use on both. When I finally looked at them together, the chenille had pilling at the seat edge and visible fading on the arm nearest the glass. The Xorel looked like it had been unwrapped the week before.
Same room. Same abuse. Different answer.
What we changed in the spec book
After that project, we rebuilt our soft goods specifications around five rules.
- Zone by exposure, not by aesthetics. High-touch public space, entrance zones, bar fronts and luggage paths get performance wovens. Guest room seating and quiet lounge areas can carry chenille, and should, because it's comfortable and it looks right.
- Write the test method and the number into the PO. Not “commercial grade.” ASTM D4157 at 50,000 double rubs. Not “colorfast.” AATCC 16 at 60 hours. A spec sheet that can't be measured can't be enforced.
- Treat dye lot as a line item. Pre-bulk dye lot approval, retained samples on both sides, Delta E under 2 read by a third party, and a written remedy if it isn't.
- Keep upholstery and drapery in separate documents. There's a reason the sheer curtain fabric specification guide on a contract project runs 18 pages — NFPA 701 flame resistance for drapery, railroading direction, shrinkage after cleaning, seam placement, fullness ratio. Blending it into the upholstery spec is how you end up with beautiful sheers that fail inspection.
- Ask every vendor what they don't do. This is the one that changed the most for us.
The numbers, eighteen months later
We moved 870 yards of the public-space spec into Xorel and kept chenille — same OEM program, same mill — in the guest rooms.
Xorel ran roughly 42% more per yard than the chenille OEM price. On 870 yards, call it $10,700 in extra material cost.
Eighteen months in: zero cushion replacements in public space. Two replacements in guest rooms, both stains, neither structural. On the previous project — the one with the $18,400 first-round replacement — we were at six swaps by month eight.
So the “expensive” option cost less. Not because it was cheaper per yard, but because I stopped paying for the same square footage twice.
The part I actually learned
It took me about four years and close to 300 spec sheets to understand that a vendor who tells you where the edge of their competence is will usually get the rest of the job right. The ones who say yes to everything are the ones who quietly subcontract the piece they can't handle, and you find out at install.
I'd rather work with a specialist who knows their limits than a generalist who overpromises. Every time. Carnegie lost a chenille print order that afternoon, and they've had every public-space upholstery and drapery specification we've written since.
It cuts the other way, too. We're a procurement and compliance team, not a testing lab. We don't run abrasion tests in-house and we don't pretend to. We pay a third party, we read the report, and we tell clients when a scope sits outside what we're set up to verify. Knowing which parts we're not great at is the reason the parts we are good at hold up.
If you're writing a fabric spec this quarter, ask every vendor on your shortlist one question: “What's the part of this project you'd rather someone else take?”
The answer tells you more about them than the catalog will.


