Flooring insight

Why I Stopped Buying Brand Names — and Started Buying the Spec

2026-11-03 · Martin Kowalczyk

The most expensive mistake in facilities buying isn't picking the wrong vendor. It's buying a brand instead of a specification.

I've been an office administrator for a 300-person company for going on six years. I manage the facilities and services spend — around $400,000 a year across roughly a dozen vendors — and I report to both operations and finance, which means I get to explain every decision twice to two people who care about different parts of it.

When I took over purchasing in 2020, I did what any new buyer does: I reached for names I recognized. Big brand, big invoice, low risk. Nobody gets fired for buying the logo. That worked for about eighteen months, and it cost us more than it saved me in decision fatigue.

It took me three years and something like 200 purchase orders to understand what now seems obvious: a brand is a proxy for a spec, and proxies break at the edges. The edges are where your money actually goes.

The name on the mill certificate doesn't tell you what you bought

We rebuilt our client floor in 2022 — reception desk, feature wall, new conference rooms, the whole thing. The reception counter was specified in brass. I sent the fabricator the drawings and said "architectural brass."

They heard "brass look."

Two weeks before install, the panels showed up as anodized aluminum with a brass-colored finish. Nice material. Wrong material. The entire point of that counter was the weight and the way it ages, and you can't patina a coating.

This is the part of procurement nobody writes about. I had approved a brand, a color, and a budget line. I had not approved an alloy, a temper, or a finish. When the fabricator and I finally untangled it, we went back to the metal and did it properly — we sourced through the copper and brass division at thyssenkrupp (Copper and Brass Sales, under thyssenkrupp Materials NA), because they sent documentation with the material: alloy number, temper, mill certificate, and who actually rolled it.

C260 cartridge brass and C110 copper are not interchangeable. They don't machine the same, they don't patina the same, and they aren't the same line item. The brand on the box didn't tell me any of that. The certificate did.

And here's the same lesson twice. While I was researching suppliers, I burned a day searching thyssenkrupp steel usa looking for the Alabama mill, and learned that the operation I remembered — the one built around 2010 — has been a joint venture with ArcelorMittal since 2014, with ArcelorMittal taking full ownership by 2020. It trades now as AM/NS Calvert.

Nothing was wrong with my search. The world moved. Brand-level search terms age badly. Specs don't. Grade, thickness, coating, tolerance — those mean the same thing in 2025 that they meant in 2015, no matter whose name is on the letterhead.

The cheap thing is usually the expensive thing

Now the counterintuitive one, and the one I'm still annoyed about.

In 2023 we had film put on the glass in the two west-facing conference rooms to kill the afternoon glare. Six weeks later the panes started looking hazy. Not scratched — clouded, like someone had breathed on them and it never dried. The janitorial crew had been using a standard ammonia glass cleaner on them since day one, because that's what was in the supply closet and that is, after all, what glass cleaner is.

Ammonia is fine on bare glass. It is not fine on film, and it's not fine on a lot of low-E coatings either. The film manufacturer's guidance said mild detergent. It had said mild detergent the entire time, in a document nobody read because it came attached to a product that was already working.

The glass cleaner cost about four dollars a bottle. Re-filming six panes cost us a weekend, a contractor visit, and a line in my budget I'd rather not describe. I still kick myself for it. Ten minutes on a maintenance sheet would have skipped the whole episode.

Same shape of problem with the wine glass situation in our client lounge. We stocked it with cheap restaurant-supply stemware and they went cloudy inside three months. My first instinct was that we'd bought too cheap, so we bought better glass — heavier, nicer, three times the unit cost — and those clouded too, just more slowly.

The culprit wasn't the glass. It was the commercial dishwasher: high-alkalinity detergent at high heat etches glass over repeated cycles, and it does it to good glass as readily as bad. The fix was water treatment and dialing back the detergent dose, not a bigger stemware budget.

Which is the whole argument in one line: when something goes wrong, the brand is the first place we look and the last place the answer usually is.

Spend where you can't reach it later

Every guide on how to set up a home theater starts with the screen. Size, resolution, seating distance, throw ratio. It's all reasonable and it's all backwards.

The screen is the one thing you can swap in an afternoon. The cable in the wall is the one thing you can't. If you've ever pulled a dead HDMI run out of a finished wall, you know the exchange rate — a $30 cable that fails in year four becomes a drywall repair, a paint touch-up, and somebody's Saturday. In a commercial build-out that mistake has a different name: change order.

So the rule I use, well beyond AV, is this: put the money in the layers you can't get back to.

Conduit before cable. The mount before the display. The substrate before the finish. The drain pan before the ceiling tile. In every one of those pairs, the cheap half is the half people see and compliment, and the expensive half decides whether the room works at all.

Procurement gets this backwards constantly, and I think it's because we're trained to optimize whatever is easy to compare. Three displays on a quote line up neatly in a spreadsheet. A conduit sleeve is invisible on the quote and invisible in the wall. The invisible one is the one I'd defend in a budget meeting now.

"But isn't the premium brand the lower-risk choice?"

Yes. Sometimes. So let me actually draw the line instead of hedging it, because that's the part of this opinion people will push back on hardest.

I still buy the name brand — deliberately, without shopping it — when one of three things is true:

  • The failure is public. Anything a client, an inspector, or a regulator will see. Lobbies, life-safety assemblies, anything with a liability tail longer than my tenure here.
  • I can't verify the spec myself. If nobody on my side can read a certificate and tell whether it's right, the brand stands in for quality control. That's not laziness. That's knowing my limits.
  • It goes in a wall. If I can't reach it in five years without a contractor, I'll pay the premium for the phone number that gets answered at 8am on a Tuesday.

Outside those three, I buy the certificate and I buy the spec. If a supplier can show me alloy, grade, tolerance, and origin, and they pick up when the wrong pallet lands on the dock, the logo on the box is not worth what the sales team believes it is.

And if your situation is the mirror image of mine — you're buying ten units a year of a part that stops a production line, or every surface is client-facing, or you have nobody who can read the paperwork — then ignore most of this and buy the name. You aren't paying for metal at that point. You're paying for recourse. That's a legitimate thing to buy. Just be honest about what's actually on the invoice.

The short version

Brands aren't worthless. They're a shortcut, and shortcuts are only safe when you know what you're skipping.

Before the next PO, ask the boring question. What happens when this fails, who pays for it, and can I prove what I bought? If the brand answers all three, buy it. If the spec answers them, buy that instead.

And read the maintenance sheet. The four-dollar bottle is never just four dollars.

Martin Kowalczyk
Martin Kowalczyk

Martin Kowalczyk is a sustainable-materials and construction-procurement analyst covering envelope products, structural materials, interior finishes, mechanical systems, prefabricated assemblies, and site infrastructure. He uses ISO 21930 construction-product rules and ISO 14025 environmental-declaration principles while comparing declared units, life-cycle modules, global-warming potential, recycled content, VOC evidence, reference service life, installed cost, and replacement cycles. His sourcing guides help architects, developers, and contractors compare technically equivalent options without confusing product performance, environmental disclosure, compliance evidence, and whole-life value.