-
Step 1: Look at the actual roll, not a photo
-
Step 2: Get the numbers, not the adjectives
-
Step 3: Test how knits recover before you commit
-
Step 4: Press-test every swatch before the fabricator touches the order
-
Step 5: Lock the dye lot in writing before the PO goes out
-
Step 6: Test the claim before you repeat it
-
Notes, and the mistakes I still see
I've been handling fabric specification and purchasing for a mid-size commercial interiors firm for nine years. In that time I've personally made — and documented — 31 significant mistakes, totaling roughly $14,000 in wasted budget. Every single one of them came from skipping a check that takes less than five minutes.
This is the checklist I now run before every order, whether it's 12 meters of velvet for a hotel lobby or 300 meters of basic polyester for furniture production. Six steps, about 20 minutes total. It's caught 47 potential problems in the last 18 months.
If you specify, buy, or approve decorative fabric for a living, it's for you.
Step 1: Look at the actual roll, not a photo
Screen color is a suggestion, not a fact. I don't care how calibrated your monitor is or how professional the supplier's photography looks — the same fabric shot under a 5600K studio light and viewed on your laptop in a room with warm lamps is a different color.
Velvets are the worst offenders. Pile direction changes the apparent color, and a velvet photographed with the nap running down can look two shades lighter than the same velvet cut with the nap up. If you're specifying a velvet, a mohair, or anything with a directional pile, you need to see the sample with your own eyes and tilt it under both daylight and the lighting the end user actually has.
I learned this in 2021 with a warm taupe velvet for 40 dining chairs. The photos were accurate — under studio light. The roll that arrived had a distinct green cast under the restaurant's 2700K fixtures. We re-selected, ate the $2,400, and I haven't spec'd a pile fabric from a photo since.
When we're working with romo textiles, I book a slot at the romo fabrics showroom and pull the actual sample boards off the wall. That's not unique to romo — most established fabric houses keep a showroom or a sample library. The point is that the visit costs 45 minutes and it replaces a three-week redo.
Step 2: Get the numbers, not the adjectives
"Heavy-duty," "commercial-grade," "premium" — none of these are specifications. They're marketing words. Ask for the technical sheet and look for actual figures: composition, weight in both gsm and oz/yd², denier for synthetics, weave or knit construction, abrasion results in double rubs (Wyzenbeek or Martindale — the two aren't directly comparable, so ask which one and what the rub count is), and lightfastness on a defined scale.
The 900d polyester fabric trap is a good example of why this matters. I once put two samples side by side, both described by their suppliers as "900D polyester," and the measured weights differed by 38 percent. One was quoting the denier of the individual yarn, the other was quoting the total construction. Neither was lying, technically. Ask for the label on the roll, not the label on the marketing sheet. (Note to self: check the roll label, every single time.)
Here's something vendors won't tell you unless you ask directly: the showroom sample and the production roll can come from different dye lots — occasionally even different mills under the same pattern number. That's not a scandal. It's manufacturing. But it's exactly what catches people on reorders, which is where Step 5 comes in.
Step 3: Test how knits recover before you commit
Knit ponte fabric is popular for banquettes, seat cushions, and anywhere you want a smooth hand without a woven's stiffness. It's a double knit, which makes it more dimensionally stable than a single jersey. But "more stable" is not the same as "stable." Recovery varies wildly between suppliers, even at the same measured weight.
Honestly, I'm still not sure why some suppliers' ponte recovers better than others at an identical spec. My best guess is yarn tension during knitting and how the finishing line is run. If someone has a cleaner explanation, I'd genuinely like to hear it.
The test that works: cut a 100mm square of the swatch, mark 50mm, stretch it to 100mm, hold for 30 seconds, release, then measure after one minute and again after 24 hours. If it hasn't returned to within 3mm of the original mark, it's going to sag on a cushion. Ugly word, uglier outcome.
In 2022 I spec'd a knit ponte for banquette backs on a restaurant project. Sixty covers. Three months in, the backs looked like they'd grown two inches — not because the fabric stretched, but because it stretched and stayed there. The redo was $1,800 plus a client conversation I'd rather not repeat.
One more thing: upholstery-grade ponte and garment ponte are different animals. If a supplier hands you a fabric sold by the meter for apparel, double-check that it's actually rated for the end use before it goes into a contract spec.
Step 4: Press-test every swatch before the fabricator touches the order
"How to iron satin fabric" is the question I get most from fabricators, and the reason is simple: satin is uniquely unforgiving. The weave is engineered to reflect light, which means any mark you leave, you leave in the most visible way possible. Direct iron contact, high heat, or uncontrolled steam can flatten the weave or leave a shine mark that no amount of re-pressing will remove.
The protocol I now put in writing on every order that involves satin, silk, or any heat-sensitive finish:
- Test on a hidden corner or seam allowance — never on a visible surface.
- Use a press cloth: cotton muslin or Teflon between the iron and the fabric.
- Start at the lowest heat setting that works. If the iron has dots, start at one dot. Yes, one.
- No direct steam. If wrinkles are stubborn, steam from the back of the fabric or from 15cm away.
- Short passes, iron moving continuously. Setting the iron down even for a second is how shine marks happen.
- Never press a velvet flat with an iron. Steam only, from the back, with the nap lifted.
I know all of this now. In 2020, I didn't put any of it in writing. A fabricator pressed 26 satin seat cushions without a press cloth — no fault of theirs, I never said — and $1,900 of finished cushions came back with visible shine marks on every seat. (Ugh, again.) They had to be replaced, and the delivery slipped three days.
Step 5: Lock the dye lot in writing before the PO goes out
Reorders are where checklists usually fall apart, because everyone's careful on the first order and casual by the fifth. "Same pattern, same colorway" is not the same as "same dye lot."
Attach the dye lot code, roll numbers, and production date to the purchase order, and ask the supplier to confirm them back to you in writing. For a branded line — a romo textile, for example — that means pattern number plus colorway plus dye lot. All three, not just the pattern.
Story time. Eight months after a first order, we reordered 12 panels of what our notes said was "the same" fabric. The supplier had discontinued the original dye lot and matched to the pattern's current standard. The color was close — close enough to approve a swatch, not close enough to put next to the existing panels. We ended up replacing the originals so the room would match. About $3,200, and the single biggest line item on my personal mistake ledger.
Step 6: Test the claim before you repeat it
This is the step most buyers skip. It's also the one that protects you from saying something to a client you can't defend six months later.
Every claim on a spec sheet — "outdoor-safe," "stain-resistant," "eco-friendly," "commercial-grade" — should be substantiated before it goes anywhere near a proposal. Per FTC guidelines, advertising claims have to be truthful and supported by evidence, and environmental claims get extra scrutiny. The FTC's Green Guides (16 CFR Part 260) are specific: a "recyclable" claim is only defensible if recycling facilities are available to at least 60 percent of consumers where the product is sold. Fabric suppliers don't get an exemption just because they sell B2B.
Four questions to ask for any claim you plan to pass along:
- What standard is this measured against?
- What test was run, and by which lab?
- What's the test date?
- Does it cover my end use — upholstery, drapery, outdoor, contract?
If the supplier can't answer, the claim doesn't go in the proposal. It's that simple, and it's the most under-used filter I've come across in this industry.
Notes, and the mistakes I still see
Don't accept absolute claims. "Will never fade" isn't a spec. Lightfastness is a number on a defined scale — get the number. No fabric is immune to the sun forever, and any supplier who tells you otherwise is telling you something about their sales approach, not about their product.
Don't pre-wash decorative fabric out of habit. The "always pre-wash everything" rule comes from an era when sizing and finishes varied enormously between mills and washing genuinely was the safest move available. For a lot of modern decorative fabrics, washing can break the mill finish, change the hand, or shrink the face. Ask the supplier before you wash anything you didn't plan to wash when you bought it.
What changed, and what didn't. Five years ago, half of this checklist was phone calls and faxes. Today most established suppliers publish technical sheets online, and a brand like romo runs a searchable pattern library you can query without picking up the phone. That's a real improvement and it makes the work faster. What hasn't changed is the order of operations: see the physical sample, verify the number, verify the lot, test the press. What was best practice in 2020 may not apply in 2025, but those four checks have survived every shift I've watched.
The whole checklist is roughly 20 minutes per order. The average redo on my own ledger is three to six weeks. Do the math on which one you'd rather spend.

