-
1. Verify actual lumens, not "equivalent wattage"
-
2. Check color temperature consistency—not just accuracy
-
3. Confirm CRI with R9 values (both matter)
-
4. Test smart bulb protocol compatibility—in the actual ecosystem
-
5. Dimming compatibility: the hidden dealbreaker
-
6. Measure physical dimensions and mounting specs
-
7. Validate operating temperature ratings against real conditions
-
8. Verify certifications—not just the logo on the box
-
9. Audit packaging and label accuracy
- 10. Read the warranty terms like a contract (because they are)
I review lighting products before they reach customers—roughly 500 unique SKUs annually across commercial and residential lines. In our Q1 2024 quality audit, I rejected 23% of first deliveries. Most rejections weren't defects—they were specification mismatches that would've turned into field returns or reputation damage.
Here's the 10-step checklist I've refined over four years at GE Lighting. It applies whether you're specifying high bay fixtures for a warehouse or buying a G25 LED bulb for your kitchen. No fluff, no theory. Just what I check and where vendors typically fail.
1. Verify actual lumens, not "equivalent wattage"
"Equivalent to 100W" is the most dangerous phrase in lighting. It's not a lie—it's a conditional truth that hides the fine print. When I compared first-article test reports from two vendors on identical spec sheets, the actual lumen output differed by 22%. Same claim, different reality.
My rule: the spec sheet must list measured lumen output, not a calculated equivalence. And I don't accept vendor-supplied test data as final—we pull random units and send them to a third-party lab. Have I been accused of being paranoid? Yes. But the one time I skipped that step ("we've worked with them for three years"), the batch came in 15% under spec. That mistake cost us a $6,800 redo and a delayed shipment of GE high bay lighting.
2. Check color temperature consistency—not just accuracy
A single 2700K lamp can be anywhere from 2650K to 2850K and still pass. Put ten of them in a row and the variation becomes visible. In a chandelier with six shades, it looks broken.
What I do: pull 5–10 units from the batch, energize them, and compare chromaticity coordinates. Tolerance of ±150K is my threshold. Most brands fail this because their chip binning isn't tight enough. The ones that pass—surprise—charge slightly more. But the cost of a callback on a 200-unit installation dwarfs the unit-price difference.
3. Confirm CRI with R9 values (both matter)
CRI above 85 works for residential. Commercial—especially retail, hospitality, galleries—needs 90+ with R9 (red rendering) at least 50. Plenty of bulbs claim CRI 85 but have R9 in negative territory. Red objects look gray.
High bay lighting in warehouses has a related trap: many products sacrifice color rendering for raw output. We measured a 12% drop in barcode scan success at a logistics client after they switched to a cheaper CRI 70 fixture. Same lumens, worse outcome. That data point became our sales pitch for the next quarter.
4. Test smart bulb protocol compatibility—in the actual ecosystem
The top complaint on smart bulb reviews isn't burned-out LEDs—it's connection failures. Zigbee, Wi-Fi, Bluetooth Mesh are not interchangeable. "Works with everything" is not a specification. No brand—including mine—achieves 100% cross-platform compatibility.
What I require: pairing, grouping, scene control, and OTA firmware updates must work end-to-end in the specific ecosystem the product claims to support. We only print protocols and versions that passed live testing. If a vendor says "it should work," it doesn't.
5. Dimming compatibility: the hidden dealbreaker
LED bulbs and legacy dimmers fight each other more than the industry likes to admit. Inrush current and minimum load ratings cause flicker, buzz, or dead silence when connected to older forward-phase dimmers.
We publish a dimmer compatibility matrix per product line—specific brand names and model numbers. Not "works with most standard dimmers." That sentence means nothing. When I specify requirements for our $18,000 lighting package, the matrix is part of the contract. Vendors who can't provide it don't get the order.
6. Measure physical dimensions and mounting specs
This is the step most people skip—and the one that causes the most field failures. A shade chandelier with an opening diameter off by 2cm doesn't fit the canopy. A high bay hook with a 1mm thread pitch mismatch won't hang.
My protocol: first-article inspection with calipers against the engineering drawing. Tolerance is ±1.5mm. Last year we rejected a full batch because the mounting ring thread didn't match the drawing. The supplier said "it's industry standard." The installer said "it doesn't fit." The installer's assessment wins.
7. Validate operating temperature ratings against real conditions
LEDs are temperature-sensitive. High bay lighting in a warehouse ceiling can hit 50–60°C ambient in summer. Outdoor fixtures (the kind you're researching when you look up how to choose outdoor lighting fixtures for your home) need cold-start reliability at −20°C in northern climates.
I've seen a bollard light rated for −20°C that failed to start at −15°C. The vendor's response: "that's lab-condition data." My response: "then your lab conditions are wrong." We now require a cold-chamber test report for every outdoor product line we carry.
8. Verify certifications—not just the logo on the box
UL, ETL, DLC—every certificate has a scope and an expiration date. In one case, a vendor recycled an old certification number for a new housing design. The logo was there; the listing wasn't.
Five-minute check: pull the certificate number from the issuing body's public database. Match the manufacturer name, model coding, and construction details. We've walked away from three suppliers who failed this. The cost of a recall or a failed inspection is 10–20x the five minutes it takes to check.
9. Audit packaging and label accuracy
The first thing a customer sees isn't the light—it's the box. Lumen count, color temperature, model number, certification marks—every data point must match the product inside. We ran a blind test in 2024: identical bulbs in two packaging sets, one with sharper print and more precise labels. User perception scores on lighting quality were 17% higher for the better-packaged set. Same LED die, different box.
10. Read the warranty terms like a contract (because they are)
"5-year warranty" means nothing until you find the freight clause. If the buyer pays return shipping, one replacement can cost 40% of the fixture price. Commercial procurement teams do that math.
My standard: warranty must state who pays freight, response time, and whether replacement is repair or swap. If the answer is "we'll figure it out," we move on.
A few things that still catch people
Don't approve based on samples. Batch-to-batch variation is real. Pull from production stock, not the vendor's presentation kit.
Don't confuse "industry standard" with "my standard." Industry standards are floors, not ceilings. Your acceptance criteria belong in the contract, with measurement methods and tolerances spelled out.
Don't accept "other customers use it this way." Your application is not their application.
Every time someone asks me to "just let this one through," my answer is the same: five minutes of verification beats five days of correction. I've run that math more times than I'd like to admit.