Flooring insight

Three Kinds of Spec Risk: Why Your Kohler Flow Restrictor and Your Sound Proofing Panels Need Different Reviews

2026-11-13 · Akiko Fujimori

Here is the thing I tell every new hire on my team: there is no universal submittal checklist. A Kohler shower head flow restrictor and a box of sound proofing panels are not the same kind of risk, even when they arrive in the same envelope on the same Tuesday.

In our Q1 2024 quality audit we reviewed 214 line items across four projects and rejected about 15% of first deliveries. Very little of that rejection came from products being bad. It came from us reviewing a compliance item the exact same way we reviewed a comfort item.

So now I sort every package into three buckets before I open the first spec sheet:

  1. The item that fails an inspection. There is a code, a federal rule, or a listed assembly behind it. Wrong means rework, not a complaint.
  2. The item that breaks a chain. It is fine on its own, but it has to match something already installed or already warrantied.
  3. The item that makes a claim. The performance number is real, but it was measured under conditions nobody in the building will ever reproduce.

Same envelope. Three different review speeds. Here is how each one actually plays out.

Bucket 1: If it fails an inspection, the spec sheet is not the authority

Shower head flow restrictors and door latches both live here, and this is where I burned the most time early in my career.

On the flow side, the number is not a suggestion. Under 10 CFR 430.32(p), showerheads manufactured after July 1, 2018 are capped at 2.0 gallons per minute at 80 psi, down from the old 2.5 gpm ceiling. EPA WaterSense goes slightly further and requires a labeled model to sit between 1.8 and 2.0 gpm, so it still performs like a shower instead of a misting wand.

WaterSense labeled showerheads are independently certified to meet both the federal flow limit and a minimum performance floor. Source: epa.gov/watersense; 10 CFR 430.32(p).

What goes wrong is rarely the product. It is the maintenance decision six weeks after turnover, when someone on the property team decides the pressure 'feels weak' and pulls the Kohler shower head flow restrictor out because it is a two-minute job with a pair of pliers. Then we come back for the 11-month walk and measure 2.6 gpm at the fixture. That is a failed inspection on a fitting nobody remembers touching.

The other version is quieter. We specify the OEM restrictor, the distributor ships an 'equivalent' one that threads differently and seats two millimeters proud of the face, and the installer tightens it down anyway. On forty units.

Which is exactly what happened to me in 2022. I wrote 'as specified' on the approval. The supplier read 'equivalent acceptable.' We were using the same words and meaning different things, and I only found out when eight heads came back off the third floor because the spray disc would not seat. We caught the rest at the loading dock. The contract line now carries the manufacturer and part number in full, and 'equivalent' is not a word we use anymore.

Door latches are worse, because the failure mode is invisible. A latch that passes the eye test can be the wrong latch entirely on a fire door assembly. IBC Section 716 requires positive latching, NFPA 80 governs the installation, and a Grade 2 latch from a reputable line is not automatically a listed component of that specific assembly. On accessible routes, the 2010 ADA Standards, Section 309.4, cap operating force at 5 pounds, one hand, no tight grasping, pinching, or twisting.

We rejected 1,400 latches once. The submittal showed a Grade 2 product that technically met the spec language, but the fire-rating paperwork for that latch and closer combination did not exist. Redoing it cost roughly $18,000 in hardware and a nine-day slip on the door schedule. I still kick myself for not asking for the listing document at submittal instead of at inspection. If I had, we would have caught it in week two instead of week fourteen.

Bottom line for this bucket: you are not reviewing quality, you are reviewing paperwork. If the listing, the flow rate, or the operating force is not written down and traceable to a standard, the item is not approved, no matter how good it looks in the box.

Bucket 2: If it has to match something already installed, cheap is a trap

This is the corner shower question, and Kohler is the cleanest example I have.

A corner shower is not a standard alcove. The tray, the door, the wall panels, and the valve trim are effectively one assembly, which is why manufacturers list them as matched sets. I have watched projects save $340 on a tray and spend $2,900 re-cutting glass eight weeks later.

I went back and forth between an OEM tray and a local fabricator's custom base for two weeks on a 62-unit multifamily job. OEM offered a warranty chain and a part number that will still exist in 2032. The fabricator offered a 12-day lead time and roughly 22% savings. I chose the OEM, and not because of the name on the box — because when a pan leaks in unit 214, I want one phone number instead of a three-way argument between a fabricator, a setter, and me.

Here is where process efficiency actually paid off. In 2023 we connected our review sheet to the distributor catalog feed, so finish codes, part numbers, and discontinued flags populate on their own. That cut our clearance cycle from five days to two. Not dramatic, but it killed the hand-typed data entry errors we used to have, the ones where a brushed nickel trim got ordered in polished chrome because somebody mistyped a code.

I want to be fair about the limits, though. On that corner shower, automation got us nowhere. It was a non-standard rough opening in a renovation, and the only thing that worked was a 40-minute call with the local rep and a set of field measurements. There is a real category of purchase where the old way — a person, a phone, and a tape measure — still beats every system I have. Tile and glass is that category.

Bucket 3: If the number came from a lab, test it against the room

Sound proofing panels are the clearest example I have, and the drum set for beginners is the clearest story.

We had an amenity room in a multifamily building where a resident family put a beginner drum kit in a basement room. Within a month there were four noise complaints. The property manager did the reasonable thing: bought panels with an STC rating printed on the carton, mounted them on the shared wall, and called it handled. Two weeks later, complaint five.

Nobody lied on that box. STC is measured in a lab under ASTM E90, in a controlled opening with no flanking paths, no electrical boxes, no ductwork. Put that wall in a real building and test it in place under ASTM E336, and the field number comes in meaningfully lower — commonly by several points, sometimes more. STC also weights speech frequencies heavily and de-emphasizes everything below 125 Hz, which is precisely where a kick drum lives.

The surprise was not that the panels underperformed. The surprise was that $900 of added mass and sealing beat $2,400 of acoustic foam, because foam absorbs sound inside the room and does not stop it crossing into the next one. We sealed the perimeter, added a layer of mass, put the drums on a decoupling platform, and the complaints stopped.

The habit I keep from that one: for this bucket, read the test standard before you read the number. STC and NRC describe two different things, and buying the wrong one is expensive in a way that takes months to discover.

Which bucket is your item in?

Three questions, in this order:

  1. If this item is wrong, does the building fail an inspection or a certificate of occupancy?
  2. If this item is wrong, does a warranty or a single point of responsibility break?
  3. If this item is wrong, is somebody just going to complain for a while?

A yes on the first question stops you at bucket 1: full documentation review, no substitutions without a drawing. A yes on the second puts you in bucket 2: match the chain first, then negotiate price. If the only downside is complaints, you are in bucket 3, and there you ask what standard the claim was measured under before you spend anything.

If I had to compress the whole method: my team used to reject about 15% of first submittals, and after we started bucketing, that number dropped to just under 7%. Not because suppliers got better, but because we stopped sending back comfort items for compliance paperwork and started catching the real ones earlier (mental note: we still need to formalize the bucket 3 checklist — it is the loosest part of the process).

Two hundred line items. Three review speeds. The mistake I made for years was giving all of them the same one.

Akiko Fujimori
Akiko Fujimori

Akiko Fujimori is a structural building-materials analyst covering sawn timber, structural composite lumber, plywood, OSB, masonry units, concrete products, reinforcing components, and metal building systems. She references ASTM D5456 and ASTM C90 within their material scopes while comparing bending strength, modulus of elasticity, compressive strength, density, moisture content, dimensional tolerance, connector capacity, and corrosion protection. Her evaluations help structural designers, fabricators, and procurement teams align declared properties, connection details, exposure conditions, and inspection evidence with project loads.