When Construction Specialties Is Just the Start: What a Quality Inspector Wishes Specifiers Knew

The Shipment That Made Me Change How I Review Specifications

In my job as a quality/compliance manager for a construction specialties manufacturer, I review Division 10 packages before they ship—roughly 200 unique items a year. Last spring, I stood in a job trailer with a set of shop drawings that should have been dead simple. The spec said construction specialties rsh 5700. The submittal said the same. The stamped drawings said the same.

Then I walked out to the pallets, measured the first unit, and asked the installer to flip one over. Wrong configuration. The model number matched. The product inside the box wasn't what the project needed.

We rejected the batch. The contractor looked at me and said, "It's the exact model you wrote."

And that's the problem. The model number was right. The spec around it was not.

Learned never to assume that a model number in the contract means the same options as the model number in the catalog. I had made exactly that assumption. Didn't verify. Turned out there was an accessory code appended in the manufacturer's data, and nobody put it on the documents. We didn't have a formal sample-return process before that. I created one after.

The cost of catching it on the dock was a few hours and a freight charge. The cost of catching it after installation would have been a $22,000 redo and a launch delay.

"Construction Specialties" Is a Bin, Not a Product

In the CSI MasterFormat, Division 10 covers construction specialties: wall protection, expansion joints, louvers, sunshades, corner guards, kick plates, door products, and other items that don't fall neatly into other trades. It's a useful organizational bin. But on project documents, I keep seeing the phrase treated as a product.

Specs say things like "construction specialties per Section 10 14 00" and then stop. Or they say "use a qualified supplier." Or they bounce the responsibility around until someone picks a product that doesn't fit the opening, the traffic, or the design intent.

The phrase becomes a shortcut. Every shortcut in a spec is a hole where something can crawl in.

To be fair, I've also seen specifiers do the hard work and get it right. They list the series, the model, the finish, the anchors, and they attach a cut sheet with a revision date. The vendors who respect that level of detail are the ones I keep calling. But the system is stacked against clarity, and I'm not sure everyone wants to fix it.

"Meyer Construction Specialties can quote it" is another version of that shortcut. Maybe they can. I've gotten good pricing from distributors on some projects. The ones that work are the ones that send a full cut sheet, confirm the options, and don't swap the model number when they submit the bid. The ones that don't work just say "we can supply that" and let the job absorb the risk.

The Real Problem Is Under-Specification, Not Brand Confusion

Let me be direct: a brand name is not a specification.

Take Acrovyn wall protection. Acrovyn is a product family, not a model. There are impact-resistant versions, antimicrobial versions, different thicknesses, and several mounting approaches. If you write "Acrovyn wall protection" and stop, you are inviting someone to pick the cheapest-profile product that can carry a logo.

Same with construction specialties rsh 5700. The model number tells me what the vendor thinks you're buying. It doesn't tell me:

• The finish specification
• The mounting hardware
• The performance test requirements
• The acceptable tolerance
• The warranty conditions for field cleaning and maintenance

Let me rephrase: it's not that the model is wrong. It's that the model is only the beginning. I want to say the tolerance on the units we rejected was ±1/16 inch, but don't quote me on that—I'd have to look at the old submittal. The point is, the drawing said "RSH 5700," not "RSH 5700 with these options and this mounting condition." That's how a correct catalog number becomes a wrong product.

A Black Front Door Isn't a Color

The same logic applies to things people think are simple.

A black front door can be a paint-grade fiberglass door, a factory-finished steel door, or a stained wood door. Under different light, those read completely differently. I once watched a homeowner choose a "black front door" from a chip, and the full-size door looked green next to the frame. The supplier was right: the color code matched.

Door handles are another universe. "Brushed stainless steel lever" doesn't tell you the series, the backset, the function, the handing, the strike, or the latch. Builders write "door handles" shorthand all the time, and finish hardware schedules are delayed for weeks while someone tries to guess what was meant.

What It Costs When the Spec Is Loose

Let's talk numbers, because "quality" sounds abstract until it's your line item.

In our Q1 2024 audit, we flagged 22% of first deliveries for issues traceable to incomplete specs—maybe 22%, I'd have to check the summary. The issues weren't always manufacturing defects. They were wrong accessories, missing gaskets, unapproved substitutions, and finish mismatches.

After we added a verification protocol—product data sheet, physical sample, and a color/layout mock-up—the reject rate dropped to 6%.

That gap is a specification problem, not a factory problem.

I remember one wall protection order where the contract allowed the installer to substitute "or equal." The product they installed wasn't defective. Put another way: it was exactly what they bid, and nothing like what the architect selected. The client paid for it to be torn out and replaced. The upside of accepting the first delivery was staying on schedule; the risk was a $22,000 redo after the tenant move-in. I kept asking: is staying on schedule worth that possibility?

The answer ended up being no.

The Maintenance Conversation Starts With Documentation

If you search "how to clean glass stovetop," you'll find a dozen methods, some of which can damage the coating. The right answer is in the manufacturer's manual, not the blog, but you have to read it before you ruin the surface.

Building products are the same. The care and maintenance requirements for Acrovyn panels, aluminum louvers, expansion joint covers, and door hardware are not optional extras. I've seen a corridor stained by the wrong solvent and then blamed on the product. It wasn't a product deficiency; it was a documentation gap. If the submittal doesn't include a care-and-cleaning instruction sheet, ask for one. Read it before the first cleaning crew shows up, not after.

What I Check Before Signing Off

1. Verify the exact model and include a revision. "Construction specialties rsh 5700" is a start. Add the product data sheet with a date and revision. That stops "or equal" guessing.
2. Demand a physical sample for anything visible. Colors fail on screens. A black front door or a wall panel finish needs to be seen in the actual light. It's the cheapest mistake prevention you'll ever buy.
3. Ask about green claims. Per the FTC Green Guides (ftc.gov/green-guides), environmental claims such as "recyclable" must be substantiated. If a supplier says their louver or wall protection is "environmentally friendly," ask for the substantiation. If they can't provide it, you're the one carrying the risk.
4. Use standards where they exist. USPS publishes exact envelope dimensions in its Business Mail 101 guide—letters and flats have clear maximum sizes so nobody gets to "interpret" them. Building product specs would be easier if more of them were that unforgiving. Check the manufacturer's published standards, not just the marketing copy.

The Honest Advice I'd Give Any Specifier

I'd rather work with a specialist who knows their limits than a generalist who overpromises. The project manager who tells me, "this is outside our scope, here's the right person"—I trust them forever. The one who says "we can do it all" and then sends a submittal with missing performance specs? I check everything.

So ask your supplier what they don't do. If they can't answer, that's a red flag.

You're not just buying a product. You're buying the confidence that the product will be the same in the field as it was on paper. That confidence takes work. But the rework costs more.