What this checklist covers — and who should use it

I review commercial lighting orders before they reach the installation crew. In 2024, that was roughly 190 deliveries—metal halide retrofits, decorative fixtures for hospitality jobs, smart-controlled spotlights for office retrofits. I flagged or rejected about a quarter of first shipments.

Not because the products were junk. Because 'junk' isn't the only thing that fails inspection. Spec drift, wiring mismatches, and smart-control gaps cause the same headaches—and they're harder to spot.

Here are the seven checks I run on every commercial lighting delivery. No fluff. Just what to look at, in order.

Step 1: Verify the metal halide retrofit matches the fixture—not just the lumen count

When a project calls for a GE Lighting metal halide light replacement—say, a 66249 series kit or an equivalent—the easy approval is matching lumens and calling it done. That's how we got burned in 2023.

We approved 40 retrofits based on a lumen match. The LED array fit the housing. But the beam angle was 15 degrees wider than the original spec. The warehouse got brighter. The aisles got glare. We redid the whole order at our cost.

Now the check is three-part:

  • Kelvin temperature across the batch (not just the sample)
  • Beam angle against the original fixture spec
  • Ballast compatibility—many metal halide housings use specific ballast types, and 'it works with most' isn't a spec

Checkpoint: Compare the retrofit datasheet line by line against the existing fixture spec. If the vendor says 'equivalent,' ask for the numbers.

Step 2: Check decorative fixtures against the approved sample—not the catalog photo

Metal chandeliers and decorative pendants are the category where 'looks close enough' costs the most. A catalog render is not a sample. A sample is a sample.

In Q1 2024, we received a batch of decorative pendants for a hotel lobby. The finish was noticeably different from the approved sample—Delta E of 4.2 against our spec. Normal tolerance for brand-critical finishes is Delta E < 2. Delta E of 2–4 is noticeable to trained observers. Above 4 is visible to most people.

The vendor claimed it was 'within industry standard.' We rejected the batch. They redid it at their cost. Now every contract includes a finish verification clause and a signed sample retention policy.

Checkpoint: Open at least one unit. Put it next to the approved sample under daylight-balanced lighting. If you can see a difference, your customer will too.

Step 3: Test the app spotlight on a real device, on a real network

This is the step most buyers skip. You can't verify app-controlled spotlight functionality by reading a datasheet. We test three things:

  • Pairing time on a standard iOS and Android device (typically 30–90 seconds for Zigbee)
  • Latency between app command and fixture response
  • Dropout rate over a 48-hour test period

The most frustrating part of smart lighting QA: 'it works' and 'it works reliably' are different things. You'd think a pairing protocol either works or doesn't. But we've seen fixtures pair fine once and fall off the mesh within 24 hours.

Checkpoint: Pair three units. Leave them connected overnight. Check the mesh status the next morning. If any dropped, the whole batch needs testing.

Step 4: Confirm line vs load wiring before the electrician shows up

Line vs load light switch confusion causes more callbacks than any other wiring issue we see. Line is the incoming power. Load is the wire to the fixture. Reversed, and the smart switch won't work—or it works intermittently.

On a commercial job last fall, the installing electrician called from site: 'The app controls the right fixture, but the wall switch does nothing.' Twenty minutes later, we found line and load swapped at the switch box. The fixture worked because it was always powered. The switch just couldn't interrupt it.

Now our checklist includes a continuity test: with the switch off, the load wire should read zero voltage. If it reads 120V, something's reversed.

Checkpoint: Label both wires before installation. If you're not the one installing, ask the electrician to confirm the test. Get it in writing.

Step 5: Run a light output and color consistency spot check

We sample 10% of every delivery with a handheld spectrometer. Two things to check:

  • CCT consistency. We require ±150K variance across fixtures in the same delivery.
  • CRI against spec. Especially for retail and food service—a CRI of 80 vs 90 changes how merchandise and produce look.

If I remember correctly, our color inconsistency rejection rate was around 12% in 2023. We tightened the spec language—specific CCT bin requirements, not a range—and dropped it to 6% by Q3 2024. The fix wasn't better vendors. It was clearer requirements.

Checkpoint: Test the first unit, the middle unit, and the last unit in the shipment. If they don't match each other, stop.

Step 6: Inspect physical condition and packaging—before the crew touches anything

Obvious, right? Except it's the step people skip when the deadline is tight. I've seen crews install fixtures with chipped housings because 'it was in the back and the install window was Friday.' That chip becomes a warranty claim in six months.

Our rule: if the packaging is compromised, open and inspect every unit. If the housing shows any finish damage, quarantine it. No exceptions. The cost of a replacement fixture is always less than the cost of a return visit.

Checkpoint: Photograph the packaging before opening. If you find damage, photograph that too. Vendors respond faster when you have evidence.

Step 7: Verify documentation and certifications

The last step isn't glamorous, but it's the one that protects you six months later:

  • UL or ETL listing
  • DLC listing if rebates are involved
  • Warranty terms in writing—not a link to a page that might change
  • Installation manual in the box, not emailed later

We once had a delivery with no installation manual. The vendor said they'd 'send it over.' They didn't. We spent three days tracking down the correct wiring diagram because the fixture had been revised without a model change.

Checkpoint: If documentation is missing, don't sign the delivery slip or accept the shipment as complete. Once you sign, your leverage drops.

Common mistakes — and what they cost

Three things I've learned the hard way:

Don't treat the delivery date as a quality signal. In March 2024, we paid $400 extra for rush delivery on a smart spotlight order. The alternative was missing a $15,000 install window. The fixtures arrived on time—with two units that wouldn't pair. Rush delivery bought certainty on timing, not certainty on quality. We still had to test everything. If you're paying for speed, budget for inspection time on the other end.

Don't skip the 10% sample because it's the same vendor. Same vendor, different batch. Our Q1 2024 audit found a color shift in a repeat order we would've missed if we'd assumed consistency. Vendors change components. Production lines shift. Your sample is the only thing that tells you what actually arrived.

Don't accept verbal specs. If the vendor says 'it's equivalent to the GE Lighting 66249 series,' get the actual spec sheet and compare the numbers yourself. 'Equivalent' means whatever the vendor wants it to mean until you verify it. I still kick myself for not documenting a verbal promise about metal halide retrofit compatibility. If I'd gotten it in writing, we'd have had grounds to dispute the replacement cost when the ballast didn't match.

The checklist isn't perfect. But since I implemented it in 2022, it's cut our warranty claims by roughly 40%—and it's kept more defective fixtures out of customer ceilings than I can count. That's the whole point: catch it before the install crew does.