Hardware Problems Almost Never Happen at Installation. They Happen When You Place the Order.

Hardware problems almost never happen at installation. They happen when you place the order.

Here's my position: most cabinet, door, and tooling failures are locked in before anyone picks up a screwdriver. Not at the drill press. Not at the hinge adjustment. At the moment you assumed a panel thickness, or skipped a drawing tolerance, or matched a tool to a compressor by category instead of by spec.

I've been running procurement for a custom cabinet and door shop for nine years now. I've personally made—and documented—16 significant ordering errors, totaling roughly $16,000 in wasted budget. Returns, freight, and on-site rework. Now I maintain our team's pre-job hardware checklist.

The lesson I keep learning isn't that I don't know hardware. It's that every single error happened when I thought I already knew. That's the real danger zone.

Argument 1: A latch's holding force and its fit are not the same thing

In 2019, we built a sideboard unit—24 doors, twin-panel, with spring-loaded compression draw latches at every meeting stile. We don't do many of those. I ordered against a catalog spec—a Sugatsune spring-loaded compression draw latch marked "6N." I assumed 6N meant "good enough."

It arrived. We sent it out. The installer called and said every door needed a rubber mallet to close, and one customer complained about finger pain. The catch? The rubber gasket's compression travel didn't match our door thickness. We were using 18mm MDF with a 3mm solid wood edge. That's 2mm thicker than the number I had in my head when I hit submit. That 2mm doubled the force required to close the latch.

The 6N rating was fine. I wasn't. I skipped the compression travel, the mounting depth, the hole tolerance, and—critically—what position the latch needed to sit at when open versus closed. They're on the same spec sheet. I jumped past them because I trusted my experience to cover it.

That error cost $1,128 in latches (47 units at $24 each), $68 in return freight, a full day of install labor, and one client's trust I can't put a number on.

Argument 2: Sliding door hardware clearance is measured in millimeters, not inches

We took a walk-in closet order in 2021—a rail-and-trolley system, top-hung, no bottom track. Standard Sugatsune sliding door hardware. I reviewed the specs, signed the purchase order, and sent it through.

What went wrong in the field: the rail bracket and the poured concrete lintel had a 4mm gap discrepancy. Invisible to the eye. Enough to make the door rub against the side jamb halfway through the slide.

Where did 4mm come from? Our opening drawing called for 2,410mm finished height. The actual opening was plastered at 2,416mm—nobody re-measured on site. And the supplier's rail system, same model number, had a revised bracket that sat 6mm lower in the current version. Both numbers were correct on their own sheet. Together, they were wrong.

A friend of mine ran into a simpler version of this with a shed door hinge repair the same month. The shed door had sagged until it dragged the floor. He assumed the hinge had failed. It hadn't. The door frame had swelled from moisture, throwing the pivot alignment off. He could have caught it with a straightedge in three minutes. He spent two hours pulling hinges and resetting.

Same problem, different scale. We shipped the right hardware out to a job that wasn't ready for it. It cost us three days of rework and $610.

Argument 3: A 33 gal air compressor and a trim nailer don't automatically match

In September 2022, our woodshop needed a tooling upgrade for a project heavy on trim and baseboards. Someone—I won't say who, but the desk is about two meters from mine—approved buying a 33 gal air compressor along with three trim nailers.

The logic problem: a compressor's tank size is only part of the system. A typical 15-gauge trim nailer needs 2–3 CFM at 70–90 PSI. A 33 gal compressor delivering 3.5 CFM at 90 PSI is fine—provided you set pressure correctly and the nailer has an adjustable regulator.

But we didn't verify the nailers' max rated pressure. Two of the three were rated for intermittent use only, up to 90 PSI. Our compressor runs fine at 110–120 PSI. The result was a 33 gal air compressor turning a $2,400 tooling investment into a paperweight. The nailers started jamming, driving fasteners at wrong angles, and destroying trim pieces until we had to reorder.

Why? Because the question "what type of nail gun for trim and baseboards" doesn't have a single-model answer. It has a system answer. Gun spec, compressor tank and pressure ratings, hose diameter—all of it has to line up on paper before it lines up in practice.

But isn't checking every spec slowing you down?

This is the objection I hear most often. "We don't have time to verify every spec." I get it.

But let me put numbers on it. A wrong order costs 4–6 labor hours in shipping and return processing. A rework visit costs 1–3 days. A lost client costs—you get it. A hardware spec check takes me—I've timed it—22 seconds per line item on our fastest day.

I still sometimes feel like verification is overkill. Some things are obvious. And that's exactly when something "close enough" ends up costing us—documented—$3,200.

Here's what I can tell you without hedging: without the checklist, our rework rate was 8%. With it, it's 1.2%. That's not over-caution. That's basic math.

My position doesn't change

Five hours of prevention beats five days of correction. Every time.

Hardware doesn't give second chances. It's easy to blame the installer, the client, the factory. But the truth is—in my case, every single time—it started with a number I skipped because I assumed I already knew it.

So here's my position, unsoftened: check everything before you pick up a tool. Your next order will thank you.