I Ordered 10 Tubelite Entry Doors. Six Came Back Wrong. Here's the $8,400 Lesson.

In February 2023 we picked up a multi-tenant retail storefront job. Eight units, sixty-seven curtain wall panels, ten entry doors. The spec called out the Tubelite storefront line, the authorized distributor turned a quote around in two days, and I approved it in about twenty minutes.

I'd been doing storefront and curtain wall purchasing for nine years at that point. For the first eight of those years, I had never once looked carefully at a door. Windows are rectangles with glass — you order them by size and series and the factory figures out the rest. Doors are not that. A door is a set of hardware openings that happens to be wrapped in a slab.

That realization cost me $8,400.

The submittal came back red

March 14 we sent out the shop drawings and the hardware schedule. Three days later the architect kicked back six doors.

The problem wasn't dimensions. It was hardware.

Three doors were specified with panic devices; we'd quoted standard locksets. Two more had a finish mismatch — brushed stainless on the elevation, anodized on our sheet. I was irritated. Isn't hardware the distributor's job? No. The distributor configures what you hand them, and what we handed them was "standard." Standard, it turns out, means whatever the last buyer ordered.

We redid the hardware schedule, resubmitted, got approval in April. I figured that was the end of it.

The expensive part hadn't happened yet.

The part I got wrong

On April 18 we released the door order. Alongside it came the factory opening sheets — one page per door opening, showing hinge locations, lock height, closer preparation, sweeps, thresholds.

Three things I did not do before signing:

• I didn't climb a ladder and measure the actual openings.
• I didn't read the opening sheets line by line against the architect's hardware schedule.
• I didn't ask what specific hardware was included in the distributor's "standard" package.

Then I signed. The factory built exactly what I signed. That's the whole point of an opening sheet.

Here's the part I still find funny: everything I'd ever read said to lock in early. Book it, freeze the spec, get in the production queue. The door we locked in earliest was the only one that ended up needing to be rebuilt. Early only helps if you locked in the right thing.

It's tempting to think a door is just a window with hardware. But the factory cuts those holes once, and the thing you sign is the thing they follow.

What surfaced at the site meeting

May 22, weekly coordination call. The site super flipped to the door sheet and asked, "Which way does this single door swing?"

I said left-hand in. He said the elevation showed right-hand in.

Three pages of opening sheets had been sitting on my desk for a month. I never cross-checked the hand against the elevation, because the submittal round had flagged locksets, so I assumed everything else was clean. That assumption is where the money went.

And hand isn't a field fix. The slab comes prepped from the factory. Getting it wrong means a new slab with new preparations, not a hinge swap in the parking lot.

Delivery day

June 12, the doors showed up. By the standard of "can this go on the building as delivered," six out of ten failed.

• Two handed wrong. The single door plus one leaf of a pair — that second one nobody had caught at all.
• Three with the wrong closer prep. Our sheets called for surface-mounted closers. The elevations specified concealed. Those got crated and sent back to the plant.
• One threshold problem. The concrete at the entry was running about 3/4" off the architectural floor height, which put us over the ADA limit on threshold height (ADA 2010, 404.2.5 — 1/2" max, measured vertically). The GC's answer was a ramp outside. The ramp and the door sweep then spent two weeks fighting each other.

Factory turnaround on rebuilt slabs was three and a half weeks. The job did not have three and a half weeks.

The bill

Here's what landed on our account:

• Two replacement door slabs, re-prepped to corrected hardware: $5,900
• Three sidelights... sorry, three doors crated back to the plant for re-machining, rush processing, and freight both ways: $1,650
• Field solution on the threshold plus a replacement sweep: $850

Total: $8,400.

Storefront installation started twelve days late. The GC sent a formal letter. Once the storefronts weren't in, the drywall crews couldn't get in behind them. Those twelve days were just on my side of the ledger — the knock-on for the GC was longer.

We ate the invoice. We signed the opening sheets. There wasn't much of an argument to make.

My one regret: I never had the site super co-sign the opening review. A single extra signature would have given us something to work with on the schedule claim. Instead it was entirely our problem, entirely our money.

The five-point check I run before signing anything now

1. Hardware by individual door, not by order. Sit with the architect's hardware schedule and the opening sheets side by side, door number by door number. It takes longer than you want. It takes less time than a rebuilt slab.
2. Hand-of-door confirmed by whoever is standing in the opening. Not by the sheet. Go to the elevation and walk it.
3. Threshold and finish floor heights measured on site before the order goes in. If the concrete is off, you want to know while the door is still a set of drawings. Compare against the 1/2" vertical threshold limit early — not after the concrete is poured.
4. If it's a steel-framed building, check embed tolerances and deflection limits at the same time. Steel erectors and facade anchorages are usually working to different numbers, and that gap shows up as a gap in your system, not theirs. This one isn't just about doors — it moves the whole elevation.
5. Get "standard hardware" itemized in writing. Hinges, closers, locks, panic devices, by model number. A quoted package with a single lump sum tells you nothing about what's actually inside it.

When this checklist is overkill

Real talk: not every job needs this.

If you've got eight identical storefronts, all doors swinging the same way, finished floors already poured and verified, and a hardware package with nothing unusual in it — then a door-by-door review is wasted motion. Order off the catalog sheet and move on. You'll be faster and you won't lose anything.

This list is for jobs that drift. Where the site is still moving, where openings aren't all the same, where the design team has specified different hardware at every exit. If that's your project — and especially if you're anchoring storefront or curtain wall onto steel columns and beams — the hour you spend reconciling hardware against opening sheets doesn't really compare to what it costs to get it wrong.

The doors eventually went in. I walked the building after. They swing fine, the hardware's clean, everything works.

They just didn't need to cost $8,400. Most mistakes don't look like disasters when you make them. They look like a form you signed on an ordinary Tuesday.