What to Look for in a Downlight Supplier: A 7-Step Checklist From the Inspection Side

I'm the person who has to sign off before anything ships. Roughly 200+ batches a year, indoor downlights, outdoor wall fixtures, and everything in between. For four years now.

Before March 2023, this checklist had three lines. Then a batch of 400 outdoor wall fixtures for a corridor project arrived with color temperature all over the place — some warm, some blue-white, occasionally in the same box. Visible to the naked eye on the same wall. Crews had to swap out half of them on site. We paid for labor twice and a redo once. The vendor said it fell "within industry standard." It probably did. Many tolerances allow ±300K. Our spec sheet said ±150K, and they'd signed it.

That's what this list is built on. Seven steps. Take them in order. Don't skip step four.

Who this is for: anyone sourcing downlights, wall lights, or architectural fixtures from an OEM or factory for distributor stock, a private label program, or bulk commercial lighting procurement. If you're buying purely off the shelf from a catalog, this isn't the piece you need.

1. Lock the specification before you ask for a single quote

Order matters. Spec first, quotes second. Ask for pricing first and you'll get back quotes written against each vendor's own internal version of "standard."

How detailed? Spell out: color temperature and the tolerance band (±150K, not just "3000K"). CRI, and at what CCT — warm and cool versions rarely perform the same. Beam angle. Cutout size. Trim diameter. Lens or diffuser material. Dimming protocol (0–10V, DALI, TRIAC — these do not interchange). Driver replaceability.

Writing "3000K" and meaning "around 3000K" is how half of these problems start.

2. Ask for the photometric test report, not the product photos

Every serious LED supplier should be able to send an LM-79 test report on request — that's the IES standard test method for electrical and photometric measurements of solid-state lighting products. No report, no commercial project. Simple as that.

For lifetime, ask for LM-80 driver data and the TM-21 projection for lumen depreciation, not the marketing page number. "50,000 hours" on a spec sheet is fine, but it should be traceable to test data, not a wish.

One more thing people forget: ask for the IES photometric file. That's the .ies file your lighting designer needs for calculations. If it doesn't exist, the fixture can't be simulated in a layout, and your specifier will pick something else.

3. The driver decides whether the fixture survives year two

Most warranty returns I've processed weren't LED failures. They were driver failures. I've seen beautifully cast aluminum housings with clean optics, backed by cheap drivers, dying in bulk inside 24 months in markets with less stable grid voltage.

Ask which brand of driver. If the answer is "we make our own" and there's no third-party data behind it, that's your answer.

Then ask whether it's field replaceable. Integrated and potted is fine for some applications, but it means the fixture's life is the driver's life, and you can't service your way out of it. Know which one you're buying.

4. SDCM — the line most buyers skip

This is where we got burned, and it's the step I see skipped most often in bulk commercial lighting sourcing. SDCM measures color consistency within and across production runs. Five-step MacAdam ellipse is the commercial floor, three-step is solid, two-step is premium work.

I assumed "same supplier, same model number" meant consistent color. Didn't verify. Turned out two production runs in 2022 were visibly different side by side, because they were binned separately and nobody asked which bin.

Now our specs say: SDCM ≤3 across all batches, with a test report per production run. One line. Changed everything.

5. Don't let a sample represent a production run

Samples get hand-picked. Production runs get whatever comes off the line that week.

We now pull three random units per incoming batch and check CCT, beam uniformity, housing tolerances, and driver labeling. Every quarter. It's not paranoia — it's cheaper than a rebuild. If you're building a commercial lighting private label, this is the step that protects your name when a customer asks why two boxes look different on the same ceiling.

6. For outdoor wall and bollard fixtures, verify the ingress protection for real

IP ratings are defined under IEC 60529, and any legitimate manufacturer can document which test the product passed. The problem is that IP65 gets stamped on housings where the gasket or cable entry doesn't actually hold up. Ask what changed between the tested unit and the shipping unit. If nothing, good. If something, get it in writing.

Two more questions most people don't ask. Salt spray testing — ISO 9227 to whatever hour rating your coastal projects need. And finish process — if they say "powder coated," ask which powder and what film thickness. Around 80 µm minimum is standard for exterior architectural fixtures. If they can't answer, it's probably paint, not powder.

This matters especially if you carry both branded product and your own line. Put a batch of private-label wall fixtures next to installed bega wall lights on the same facade, and the difference in finish reads instantly — uneven coverage, different gloss level, visible seal lines. That comparison costs a distributor more than the price difference ever saves. Same logic applies to bega light fixtures and any other established architectural line you stock alongside: the reference is already on the wall.

7. Put delivery certainty in the contract, not just delivery dates

Here's what four years of this actually taught me. A late shipment in March 2023 stalled an installation crew mid-project. We recovered the cost through penalty clauses. We did not recover the customer's confidence.

So now every PO includes a specific ship date, liquidated damages per week of delay, and a rolling release schedule. It feels heavy. It also means the vendor treats the date as real. When we implemented this protocol, on-time performance on our top programs went from inconsistent to something we could actually plan around.

This is also why I'll pay a rush premium when the stakes justify it. Not for speed — for certainty. When we paid roughly $400 extra on a batch to guarantee a fixed delivery window, we weren't buying faster shipping. We were buying the difference between "probably arrives" and "arrives by this date." Cheap and uncertain costs more than expensive and confirmed. That's been true every single time I've tested it.

Things that go wrong before you even ask for pricing

• No .ies file provided, and nobody follows up on it
• Incoming inspection that checks packaging but not a single unpacked unit
• Mixing drivers from different suppliers across one fixture family, then discovering the dimming protocol doesn't match
• Approving a sample, receiving a run that doesn't match it, and accepting it anyway to keep the schedule
• Printing 5,000 private label boxes before the product passes inspection

One last note. In bulk lighting procurement, most negotiations start with unit price, because unit price is easy to compare. That's usually not where the money is. The money is in the batches you have to redo, the customers who notice the color shift, and the installers standing around waiting. Price is a number. Everything else is the cost.