Most "cement board specification guides" I read are brochures in a lab coat. They tell you what the product is good at. They almost never tell you what we'll reject. One of those is a specification. The other is advertising with better typography.
I'm the quality and brand compliance manager at a construction specialties supplier. I review every technical submittal, cut sheet, and spec guide before it reaches a customer — around 200 documents a year. In 2025 I bounced 31% of first drafts. The most common reason wasn't a wrong number. It was that the document never defined "acceptable." If your spec guide can't tell a contractor when to stop, it isn't doing a spec guide's job.
A spec that never says "no" isn't a spec
Here's the thing about tolerances: they only matter at the point of rejection.
In February 2025 we took delivery of 4,200 fiber-cement panels, nominal 12.7 mm, from a supplier we'd used twice before. Thickness measured 12.1 to 12.2 mm at the corners and 12.5 mm in the middle. Squareness was off on roughly a third of the pallet. The vendor's rep said it was "within industry standard."
Maybe it was. But our published spec said ±0.5 mm on thickness and 1.5 mm on diagonal — and the batch missed both. So we rejected it. They redid it at their cost, and we lost three weeks. Now the thickness and squareness requirements go on the purchase order itself, not just in the guide nobody reads.
The lesson wasn't about that supplier. It was that a spec guide with no rejection criteria is a document with no teeth, and contractors know it. ASTM C1186 (the standard for flat fiber-cement sheets — check you're citing the current edition) doesn't exist to make a product look good. It exists so two people can disagree and have a way to settle it.
What a working spec actually contains
Not much, honestly. Five things:
- The product standard and the edition you're citing
- Dimensions and tolerances, in the units the installer actually uses
- The test method behind every performance claim (fire, water, impact — name the method, not the adjective)
- The documentation that ships with the order
- What you'll reject, and at what measurement
If your guide doesn't have all five, you're asking a contractor to trust you instead of the document. Contractors don't do that, and honestly they shouldn't.
The counterintuitive part: the cheap option costs more per wall
Everyone buys on unit price. I get it — I sit in those meetings.
But unit price is the least useful number in a construction specialties catalog, because the failures that cost real money are the ones you can't see at delivery.
We ran a check with three of our install partners a while back. Same PVC wall panel profile, two versions: one with a factory-sealed edge, one without. We asked the crews which they'd rather install. Nobody could tell. Same weight, same cut, same look on the day.
Eighteen months later, in a humid bathroom retrofit, the difference was pretty obvious (not catastrophic — it was a maintenance call, not a failure). But it was a maintenance call the sealed panel didn't generate.
The seal added about $0.60 per linear meter at the time. On a 3,000 m order, that's roughly $1,800 upfront. One service call on a commercial bathroom — access, labor, downtime — runs higher than that in most markets. Our service partners quoted $2,400–$3,800 per call in Q1 2026, excluding the owner's lost revenue. Verify against your own numbers; labor rates move fast and regional spreads are wide.
So the cheaper panel was the more expensive one. But only if you were measuring over 18 months instead of 18 days.
This is where a structural panels supplier or a panel distributor either helps you or doesn't. The good ones show you the installation sequence, the sealing method, the fastener schedule, and — this is the tell — the things that go wrong if you skip a step. The others send a PDF with a hero shot and a color card.
You can't educate a customer after they've ordered
I'd rather spend ten minutes explaining options than three weeks arguing about a rejected lot. That's not altruism. It's arithmetic.
Think about what happens when a designer searches "cement board specification guide" at 11pm before a submittal deadline. They're not looking for your brand story. They want three things: which standard applies, what the tolerance is, and what documentation ships with the order.
Same with the odd searches. When someone types "construction specialties RSH-5700" into a search bar, they aren't browsing. They picked that designation off a drawing or a competitor's cut sheet, and they want the submittal data behind it. If that page has a hero image and a "contact us for details" button, you've lost them to whoever published the actual sheet.
Honestly, I'm not sure why so many suppliers still hide their tolerances. My best guess is that a lot of them have never measured a production run closely enough to know what their tolerances are. That's a guess, not a fact. But I've read enough cut sheets that say "nominal" and nothing else to suspect it.
The point holds either way: an informed buyer asks better questions, and better questions close faster. Someone who understands what TSCA Title VI means for composite wood (40 CFR Part 770, in force since 2019 — confirm current requirements with your compliance team) doesn't need a two-week email chain about formaldehyde. They ask for the certificate, get it, and move on.
The obvious objection
"If we publish our rejection criteria, customers will use them against us."
Some will. The ones who do were going to be a problem anyway.
The other version of this objection is "we don't want to confuse people with technical detail." I've heard that one at three different companies. What it usually means is that the sales team doesn't understand the detail either, and nobody wants to be the person who asks.
I had this exact argument internally in 2024 when we started putting inspection criteria on product pages. I second-guessed it for about six weeks. What if buyers started quoting our own tolerances back at us on claims? What if a competitor screenshotted the page?
What actually happened: our first-article rejection rate on incoming material dropped from 12% to 4% over three quarters. Not because we got nicer — because buyers started ordering to the spec instead of guessing at it. And the claims conversations got shorter, because there was a shared document instead of two opinions.
I won't pretend that generalizes to every company. It generalizes to companies whose specs are actually correct. If your tolerances are aspirational, publishing them will hurt, and it should.
Where I stop being useful
I'm not a fire-testing engineer, so I can't tell you how a specific panel assembly performs in a rated wall. I'm not a structural engineer either, so I can't sign off on shear values or wind load for a given fastening pattern. If you need that, you need the test report and a stamped calculation from someone licensed in your jurisdiction. Anything I told you past "here's which document to ask for" would be me guessing in a nice font.
What I can tell you, from a compliance seat, is the short list. Ask for: the test report with the lab's accreditation and report date; the standard it was tested to (ASTM E84 for surface burning, ASTM E136 for noncombustibility, EN 13501-1 if the project is European); the current edition of the product standard (C1186, C1288, C1396 — they get revised); a formaldehyde compliance statement if there's any composite wood in the assembly; and a certificate that isn't older than the last formulation change.
If a supplier can produce those in one email, that tells you more about them than any specification guide they publish.
Bottom line
A specification guide that only lists features is a sales document. It's not dishonest, exactly. It's just useless at the moment it's needed — which is when a pallet shows up wrong and two people have to decide who pays for it.
Write the number. Write the tolerance. Write what you'll reject. Then let the customer decide.
Ten minutes of explaining beats three weeks of arguing. Every time.