Large Chandelier Procurement for Commercial Projects: An 8-Point QA Checklist

Large Chandelier Procurement for Commercial Projects: An 8-Point QA Checklist

If you're the one signing off on a large chandelier for a hotel lobby, a restaurant dining room, or a high-end office atrium, this is the checklist I use. Eight steps, in the order I run them, the week before any lighting purchase order gets approved.

I've been the quality and brand compliance manager at a commercial lighting company for the past six years. I review every fixture package before it ships to our design-build and hospitality clients—roughly 200+ unique fixtures a year. Since we tightened our verification protocol in 2022, I've rejected about 14% of first deliveries. Most of those rejections were avoidable. The cost of catching them at the loading dock is roughly one-tenth the cost of catching them after install.

So, eight steps. No background section. Let's go.

Step 1: Weigh the fixture against the real ceiling, not the drawing

Most specifiers check the architectural reflected ceiling plan (RCP) and call it done. That's the mistake. The RCP shows what was designed. It doesn't show what the structural engineer approved after the second change order.

A large chandelier—say a Visual Comfort Cadence chandelier in the 48-inch size—can weigh 80 to 120 pounds once you add canopy, crystal, and trim. The mounting load isn't the same as the fixture weight, either. Off-center canopies on sloped ceilings can double the torque on the junction box. If the box was rated for a 50-pound ceiling fan, you're already over spec.

Before I sign off, I confirm three numbers in writing from the GC: fixture weight, canopy type, and junction box rating. If they don't match the spec sheet, the fixture doesn't ship.

Step 2: Match finish lots across the entire package

This is the one most buyers miss completely. It's also the one that generates the loudest client complaint, because the problem is visible to anyone who walks into the room.

When you order a Cadence chandelier alongside chandelier sconces for the same space, they need to come from the same production batch. Hand-applied finishes—especially the antique brass and plaster finishes in the Visual Comfort range—vary between batches. Two identical SKUs from different lots can read as two different products under hotel lighting.

I request the batch numbers on the PO and hold the shipment until the vendor confirms all components ship from the same lot. On a 22-fixture restaurant package we ran in 2023, the brass between two batches read as warm and cool respectively. The interior designer flagged it before install (thankfully). If she hadn't, that would've been a repaint of the entire ceiling coffer.

Step 3: Check crystal and alabaster consistency, piece by piece

Crystal and alabaster are natural or hand-cast materials. They vary. That's the point, and also the problem.

On a Talia Visual Comfort chandelier—which uses hand-cut crystal drops—I ask for a photo of the actual units, not catalog images. Crystal color temperature under the specified dimming protocol matters more than the catalog suggests. Warm-white LEDs make cool crystal read as gray. That's a $6,000 fixture that looks cheap in the finished space, and it's a complaint you can't fix after install without removing the fixture.

For alabaster, I check veining consistency across the visible faces, not just the top-down view. Alabaster is directional. What looks uniform from above can look blotchy from the actual seating angle.

Step 4: Confirm the dimming protocol on the actual driver

I've written this up before and I'll say it again: the dimming spec on the catalog sheet is the theoretical maximum. What matters is the driver sitting inside the canopy.

Most large chandeliers in the Visual Comfort range ship with a 0-10V or TRIAC driver, but the specific driver varies by SKU and by production year. If the project runs a DALI control system, a Lutron HomeWorks panel, or—worse—a builder-grade wall dimmer, you can get flicker, buzz, or a fixture that only dims to 30%. None of that shows up in the pricing quote. All of it shows up at commissioning.

I ask for the driver datasheet before the PO. If the driver isn't compatible with the control system, the fixture gets rejected at the loading dock, not during a design review two months later.

Step 5: Verify canopy and ceiling geometry

This is unsexy and it's where timelines die.

Sloped ceilings, coffered ceilings, and ceilings with recessed HVAC returns all change the canopy requirement. A standard 5-inch canopy doesn't cover a 7-inch opening. Visual Comfort makes sloped-ceiling adapters for most collections, but they're a separate line item with a 3-4 week lead time if they're not bundled into the original PO.

I measure the actual ceiling opening on site—not the dimension on the drawing—and confirm the canopy covers it with at least 1 inch of overlap in every direction. If it doesn't, the order changes before the fixture ships.

Step 6: Lock the electrical rough-in before you place the chandelier order

Here's the cost question everyone asks late: how much to install recessed lighting versus the chandelier rough-in. For a commercial space, recessed lighting install runs roughly $150-$350 per can (labor plus fixture) as of January 2025, depending on ceiling type and whether new circuits are needed. The chandelier rough-in is often cheaper—$200-$500 for the junction box, whip, and control—but only if it's planned in the same electrical phase as the rest of the ceiling work.

In Q2 2023 I watched a project save nothing on the chandelier itself and then spend $4,200 reinstalling drywall because the rough-in was missed. That's not a lighting problem. It's a scheduling problem that masquerades as a savings. I get an electrical rough-in sign-off before the chandelier PO goes out—no exceptions.

Step 7: Pair the chandelier sconces for visual continuity, not just SKU match

Sconces and chandeliers from the same collection often use different light sources and different finishes. That's fine. What's not fine is when the sconce reads warmer or brighter than the chandelier at the same dim level.

I pull two units from the same batch—one chandelier, one sconce—and power them side by side at 50%, 75%, and 100%. If they don't read as the same family, one of them replaces before the shipment goes out. The test takes 45 minutes. It has saved me three redesigns in the last two years, which is a lot more than 45 minutes.

Step 8: Inspect packaging and freight protection

Large chandeliers ship in crates. Crates get dropped. Crystal chips, plaster cracks, and bent brass arms are the top three freight failures, in that order.

I require the freight carrier to sign a condition report at delivery, and I photograph every crate before opening. If there's damage, the report supports the claim. Without it, the vendor's response is some version of "must have happened on site," and you're paying for the replacement. I want to say we've claimed back roughly $18,000 in freight damage over four years purely because the report was signed, but don't quote me on the exact figure—it's probably a bit higher.

Common mistakes and notes

The most expensive mistake I see isn't a spec error. It's skipping the sign-off chain. All eight steps above take about a day of work on a mid-sized project. Skipping them costs weeks.

A few final notes before you take this into your next project meeting:

  • Order all fixtures for a single room at once, in one PO. Split orders are the leading cause of finish mismatch, even within the same year of production.
  • Never accept "within industry tolerance" as a finish-variance argument above ΔE 2.0 for matched sets. Get the number in writing on the purchase contract before you countersign.
  • Prices above are as of January 2025 and vary by market, union status, and project type. Verify local labor rates against a recent contractor quote before you estimate.
  • The 14% rejection figure isn't a criticism of any specific manufacturer. It's the reality of hand-finished decorative lighting at commercial scale. Plan for it in the schedule.

I used to think this level of checking was overkill. It took me three years and about 150 rejected deliveries to understand that the checklist isn't about distrust. It's about making sure the fixture the designer specified is actually the fixture the client receives. Those are two different outcomes more often than most people in this industry want to admit.

Linh Tran
Linh Tran

Linh Tran is an LED driver and smart-lighting controls analyst specializing in dimming, sensors, switches, wireless control, and connected lighting systems. She checks IEC 61347 controlgear safety boundaries and IEC 62386 protocol functions, then measures power factor, THD, inrush current, standby load, dimming range, addressability, and diagnostic behavior. She writes engineering guides for teams comparing drivers, motion sensors, smart bulbs, and control strategies across new installations and interoperability-sensitive retrofits.