-
Month 12: the chair seats started peeling
-
The problem wasn't buying cheap. It was buying the wrong thing
-
So what is PU leather, actually?
-
"Luxury" is a tactile word, not a specification
-
How our process pushed the cost into the base fabric
-
What not solving it actually cost
-
How I write the spec now
-
When this approach is overkill
Month 12: the chair seats started peeling
It started at the edges. The seam where the armrest meets the backrest—the spot everyone brushes getting in and out—went white first, then curled inward in strips. I picked at a piece with my fingernail. The fabric underneath was intact. Only the top layer was failing.
By month 18, not a single chair in the conference room looked salvageable. The curtains were less obvious but the same story: they looked great when we hung them, and by the second year the drape had gone flat, the folds wouldn't hold, and the side facing the window had visibly faded.
My director saw the chairs during a client visit. He didn't say anything at the time. After they left, he asked me one question: how much did these cost?
I pulled the contract. They weren't cheap. Not cheap enough to explain what I was looking at.
That's when I started asking the question I should have asked before we placed the order: what are these things actually made of?
The problem wasn't buying cheap. It was buying the wrong thing
I'm the office administrator for a 400-person company across three locations. I manage roughly twelve vendor relationships and about $900K in annual spend on everything that isn't IT or office supplies. I took over purchasing in 2021, and for the first couple of years I assumed the hard part of the job was process—approvals, invoicing, reconciling.
So my first assumption about the peeling chairs was simple: the supplier cut corners. That assumption cost me about two weeks of contract review and increasingly tense emails.
Their response was technically airtight. The product matched the specification we signed. I went back and read that specification, which I had approved without really reading it. It said, in full: "PU leather, black, 1.2mm."
That's it. That was the whole spec.
The real problem wasn't in the contract. It was that I didn't know what PU leather is.
So what is PU leather, actually?
PU leather has three layers, and only one of them shows up in a quote.
The top is the polyurethane coating—that's what you touch. The middle is the adhesive layer bonding the coating to the backing. The bottom is the base fabric: usually a woven, knitted, or nonwoven substrate.
Writing "PU leather" on a purchase order is roughly as precise as writing "fabric." The term describes the top layer and nothing else.
All three things that actually fail live in the parts you can't see:
- The coating system—formulation, thickness, surface treatment
- The base fabric—fiber type, weight per square meter, construction
- The peel strength between coating and backing
What I eventually learned: the chairs weren't failing because the coating "aged." They were failing because the peel strength was too low. Daily friction at the seams created shear stress, the adhesive layer gave out first, and the coating separated from the backing. The backing itself was fine. It never got a chance to matter.
Here's the detail that still bothers me. When we chose the sample, we chose the one that felt best in the hand. Hand feel comes from the coating. Whether it survives five years comes from the base fabric.
The fibers inside that base fabric never appear on a product catalog. This is the layer where Kuraray fibers operate—their PVA fiber (Kuralon) isn't there to deliver hand feel, it's there for tensile strength, dimensional stability, and weather resistance. You will never see it in a sample book, because it's buried below the part you're judging.
"Luxury" is a tactile word, not a specification
The curtains were worse in a way, because they failed slowly.
The supplier described them as luxury curtain fabric. That phrase has no definition in the industry. In practice it usually means: good drape, substantial weight, some sheen, feels thick in the hand. All of that comes from finishing processes, and it's easy to produce once.
Whether it's still there in three years depends on something else entirely: yarn tenacity, color fastness, UV resistance, and flame-retardant certification.
None of those four were in our specification. Not one.
Something I learned afterward, and it took me a while to accept: the sample that feels best is often not the one that lasts. The finishing that makes fabric feel good and the yarn structure that makes fabric hold up are two different things, and they sometimes work against each other. A soft finish usually means lower twist and a looser structure—great drape, weaker resistance to pilling and stretch.
I didn't believe it at first. Then I pulled the test reports on two sample lots and compared them side by side. I believed it.
To be fair, this isn't an argument that expensive is automatically better. It's an argument that "luxury" has no verifiable meaning in a procurement document. It's a sales word. It cannot be accepted at inspection.
How our process pushed the cost into the base fabric
This part is structural, not the fault of any one supplier.
Office refurbishment procurement usually runs the same way: three quotes, priced per square meter or per chair, award to the middle-low option.
That process has a predictable outcome. The coating is visible, touchable, comparable. The base fabric isn't. So when a supplier needs to move on price to win, the only place left to move is the part nobody can inspect.
Fifty grams less backing. A thinner adhesive layer. A shorter-staple blend. None of it detectable in a sample. All of it invisible until month 12, when the material tells you itself.
What I do now: for anything involving upholstery or curtains, I request the base fabric technical data sheet before the sample. The order matters. Samples confirm color and appearance. They do not confirm durability.
What not solving it actually cost
The direct cost was one bill. Reupholstering the conference room plus three reception areas. That didn't come out of the annual budget—it was an emergency request, and I had to explain it twice to finance.
That's the cheapest part of the story.
The second cost was credibility. My director didn't see a supplier's name when he looked at those peeling chairs. He saw whoever signed the purchase order.
The third cost is the one I least like talking about. After the refurbishment was signed off, we had to file material and fire-safety documentation. We had no flame-retardant certificate for the curtain fabric—it wasn't in the contract, so the supplier never provided it. Retroactive testing took six weeks. During those six weeks, that room couldn't host anything requiring a formal reception.
If an inspection had landed in that window, the problem would have been considerably larger than peeling chairs.
Add the three together and it exceeded several times the price gap between the quote we took and the one with the sample that felt slightly less impressive.
I still kick myself for only asking about the coating. If I'd asked about the backing, we'd have caught it before the order.
How I write the spec now
Four changes. That's all.
One: write to the use case, not the hand feel. Chairs say "high-frequency contact, daily use, five-year replacement cycle." Not "black, 1.2mm, soft hand." The second set describes a sample. The first set describes a situation.
Two: for upholstery, always ask about the base fabric. Fiber type, weight, construction, peel strength test report. Those four questions alone filter out a share of suppliers—not because their material is necessarily bad, but because in some cases they don't know either.
Three: for curtains, demand composition, weight, and certification. Yarn content, grams per square meter, color fastness rating, flame-retardant standard. Requirements differ by market—Europe and North America aren't aligned, and hospitality projects have their own set. This has to be confirmed before signature, not at handover.
Four: data first, sample second. Reverse the order and the hand feel will pull you off course every time.
None of it is complicated. The hard part is thinking of it the first time.
When this approach is overkill
I don't think this is universally right, and I'd rather say so.
If the space is short-cycle—a trade show booth, a temporary showroom, a leased site you'll hand back in two years—screening on base fabric data sheets is self-inflicted delay. A cheaper PU holds up fine for one use cycle, and the cost math is reasonable.
Same for low-traffic spaces that open a few times a month. Durability isn't the binding constraint there; appearance and budget carry more weight. Choose on appearance.
There's also a boundary on PU leather itself that no amount of sourcing skill fixes. It hydrolyzes in humid, poorly ventilated environments. That's not a quality issue, it's the chemistry of the material. In a damp climate with weak air circulation, PU upholstery has a shorter life than it would in a dry room. You can't select your way out of that—you can only control the environment.
And the reverse: if you're furnishing a reception area that takes hundreds of contacts a day and needs to last a decade, PU leather shouldn't be on your shortlist at all. That's not a knock on any manufacturer. It's the wrong material for that job.
Synthetic leather as a category does have serious options. Kuraray's Clarino is a different approach—bio-based raw materials, a structure that isn't the same as conventional PU-coated fabric, and durability in a different bracket. It's also a different budget line, and it shouldn't be bid against a 1.2mm coated fabric as if they were equivalents.
Which is where I've landed on the word "luxury": it tells me nothing. I want composition, weight per square meter, and test reports.
After the chair incident, I went back and inventoried every upholstered and curtained surface we own. Forty-seven locations.
Thirty-one of them, I still can't tell you what the base fabric is.
That's the cost of not asking.