What Is a Mezzanine Floor? 7 Questions Commercial Build-Out Clients Ask Us

I coordinate commercial tile and materials orders — mostly Daltile, mostly restaurants and hotels. Nine years of it, somewhere north of 300 rush jobs. And here's the thing nobody tells you when you take this job: roughly a third of the calls I get aren't tile questions. Clients treat their rep like a general contractor's hotline, which is fine by me, but it means I've answered "what is a mezzanine floor?" more times than I've quoted a backsplash.

So I started keeping a list. These are the seven that actually come up, roughly in the order they come up.

• What is a mezzanine floor, and does it change my tile spec?
• Is "Daltile concrete masonry" even a real product?
• Daltile Color Wheel Arctic White — can I mix batches?
• Are sound proofing panels worth it in a hard-surface room?
• Why does the highball glass order blow up every bar build-out?
• How do I run a rush order without paying disaster prices?
• What's the one step almost every project skips?

What is a mezzanine floor, and does it change my tile spec?

Short version: it's an intermediate level inside a room, not a full storey. The version that matters for permitting is the International Building Code, Section 505. A mezzanine stops counting as a story as long as it doesn't exceed one-third of the floor area of the room it sits in. Past that threshold, it's a story — and your egress, sprinkler design, and structural review all shift with it. That's the part people miss. They draw a mezzanine, forget the threshold, and the drawings come back red.

Tile-wise, three things change. Deflection, transitions, and slip rating.

The Tile Council of North America Handbook puts ceramic at L/360 on the substrate and natural stone at L/720 — and mezzanine framing is usually lighter than slab-on-grade. A second-floor landing over a dining room is going to move more than the kitchen floor does, no matter how good the tile is. The transition at the mezzanine edge is where installs go sideways, because that's a trim profile and a height-difference question that needs settling in the drawings, not the week of.

I'm not a structural engineer, so I can't tell you what your framing carries. What I can tell you is get the deflection number from whoever stamps the drawings before you order anything heavier than 3/8-inch porcelain.

Is "Daltile concrete masonry" even a real product?

Daltile doesn't sell concrete masonry units. Nobody in the tile business does — CMU block is a masonry product with its own supply chain, its own yards, its own delivery logistics. When clients search "daltile concrete masonry," they usually mean one of two very different things.

Either they mean the concrete-look porcelain lines — porcelain printed and textured to read like poured concrete or block — or they're tiling over an actual CMU wall and want to know if tile will bond to it.

Those are not the same order. The porcelain is a stock or quick-ship item depending on the line and the size, and the sample usually tells you everything. Tiling over CMU is a substrate conversation: mortar bed or bonded membrane, and the block needs to cure first. Fresh CMU shrinks. Tile doesn't. Guess which one wins.

Ask your rep which of the two you're actually quoting. I've watched a purchase order go out for the wrong one — it got caught at the sample stage, and only because somebody bothered to look at the wall.

Daltile Color Wheel Arctic White — can I mix batches?

Color Wheel is a mosaic program and Arctic White is one of the brighter neutrals in it. The thing that bites people isn't the color. It's the dye lot.

If you're doing a 40-seat dining room in one phase, order what you need plus 10 percent and don't think about it again. If you're doing three phases across eight months — which is most franchise builds — you're looking at three different production runs, and mosaic whites are not forgiving about it. Arctic White next to a slightly warmer run reads as a mistake in a room with good daylight.

Looking back, on a hotel job in 2023 I should have ordered all three floors at once and warehoused the overage. At the time the client didn't want to pay storage. We ended up tearing out a feature wall because the phase-two tile didn't match. Storage would have run about $600. The tear-out ran well past that, and I don't need to do the math out loud.

Ask for the dye lot number in writing before the PO goes out. If your rep can't give you one, you're not far enough along to be ordering.

Are sound proofing panels worth it in a hard-surface room?

A porcelain floor has an NRC of roughly zero. It reflects everything you put at it. So yes — if you're tiling the floor, the ceiling, and at least one wall, that room is going to be loud, and no tile selection on earth fixes that.

Sound proofing panels — fabric-wrapped absorptive panels, a suspended acoustic ceiling, that whole family — are doing a different job than the tile is. Tile is your finish. The panels are your NRC. Budget both, or the room fails inspection by customer instead of by code.

I'm not an acoustician, so I can't size a treatment for you, and I'd push back on anyone who offers to. What I can tell you from the coordination side is that acoustic scope gets value-engineered out at roughly the same rate the tile upgrade gets value-engineered in — and then everyone is shocked that the space is unbearable on opening night.

To be fair, plenty of operators open loud and fix it later with soft furnishings and a rug budget. That's a real option. It's just a more expensive one, and you only get to make the choice once.

Why does the highball glass order blow up every bar build-out?

Because it's the last thing anybody thinks about and the first thing that's backordered.

Bar glassware is a separate supply chain from your tile and your FF&E package. A highball is typically in the 8-to-12-ounce range depending on the program, and the spec matters — rim thickness, tempered vs. not, whether it survives a commercial dishwasher cycle. But the pattern I see is almost always the same: tile, millwork, lighting, and equipment each get a line in the schedule. Glassware gets ordered three weeks before opening by whoever handles beverage, and then half of it ships late and some of it ships broken.

I'm not going to quote you a breakage rate — it swings enormously by operation and I'm not a beverage consultant. What I will say is order your overage up front and get glassware on the schedule at eight weeks, not three. Stems and highballs are the two categories I've watched delay an opening more than any tile shortage, and that surprises people every time.

How do I run a rush order without paying disaster prices?

You don't run a rush order. You prevent one, and then you run the smaller number of rush orders that slipped through anyway.

When I'm triaging a rush, I'm looking at three things in this order: how many hours are actually left, whether the item physically exists in a warehouse I can reach, and what the worst case looks like if it doesn't arrive. That third one decides everything. A missing bullnose on a Tuesday is an inconvenience. A missing bullnose 36 hours before a certificate of occupancy is a five-figure problem, and I've watched a client eat a delay penalty that dwarfed the cost of the tile by an order of magnitude.

Our house rule now is 48 hours of buffer on anything with a hard deadline attached, and we only got there after a 2022 job where the freight showed up on time and the trim didn't. Granted, buffer costs money and clients push back on it. In my opinion it's the cheapest line item on the whole schedule, and I'll argue that with anyone.

What's the one step almost every project skips?

The pre-order verification walk.

Somebody stands in the actual room with the actual sample and the actual spec sheet, and confirms that the square footage on the PO matches the square footage on the plan, that the trim profile is the one the installer expects, and that the dye lot is written down somewhere a human can find it. That's it. Fifteen minutes, maybe.

The 12-point checklist I built after my third bad order has saved us an estimated $8,000 in rework over the last two years, and every item on it is something that was obvious in hindsight. The problem is never that the information isn't available. It's that nobody's standing in the room looking at it.

This worked for us, but we're a mid-size operation with a fairly predictable project mix. If you're running a dozen simultaneous sites with different superintendents, your version of the checklist is going to look different from mine.

Five minutes of verification beats five days of correction. If you ask me, that's the whole job.