Cree LED Sourcing FAQ: 7 Questions I Should've Asked Before My First Bulk Order

I handle commercial lighting procurement — distributor accounts, a few contractor programs, a couple of OEM relationships. Nine years now. In that time I've personally made, and written down, 14 significant mistakes. Roughly $41,000 in wasted budget, give or take. (Should mention: some of that was freight, which I never count properly.) Most of them happened because nobody had told me what to check before I signed the PO.

What follows is the FAQ I wish someone had handed me in 2017. Seven questions. Five are the ones I actually get asked by new buyers. Two are the ones nobody asks until after the order has shipped.

1. What should I verify on Cree LED bulbs before ordering at volume?

Four things, and 'lumens' is only one of them.

Delivered lumens, not nominal output. A fixture rated at 4,000 lumens is not the same as a fixture that delivers 4,000 lumens to the floor after the optic and the diffuser take their cut. Ask for the photometric file — the IES file — not the marketing number.

LM-80 data and the TM-21 projection behind the lifetime claim. Every 50,000-hour or 100,000-hour number you'll see comes from an LM-80 test and a TM-21 extrapolation. If a supplier can't tell you which LM-80 report their number came from, the number is decoration.

Chromaticity tolerance. This is the one that got me. ANSI C78.377 sets a tolerance band around each nominal CCT, which means two fixtures both labeled 4000K can sit at opposite edges of that band and look visibly different when they're side by side on the same ceiling.

In 2019 I ordered 400 bulbs — no, 420, I'm mixing it up with the parking garage job — for a corridor retrofit. Same part number, same CCT, same order. Half of them had a cooler cast than the other half. Not dramatic. Enough that the facilities manager noticed, and once he noticed, he couldn't stop noticing. Swapping the visibly mismatched fixtures cost about $2,100 plus a weekend of labor.

Dimming compatibility. If the existing dimmers are older, test compatibility before the volume order, not after. This is a boring check and it is the one that most often fails.

2. How do I choose a light bar for wholesale without getting burned?

Ask for three documents before you look at a price list: the IES file, the driver spec, and the warranty terms in writing.

The IES file tells you what the bar actually does. The driver spec tells you whether the driver is replaceable or potted into the housing — which is the difference between a maintenance line item and a full fixture replacement in year four. The warranty terms tell you whether you're covered for the fixture, the driver, or 'the fixture excluding the driver,' which is a phrase I have learned to read very carefully.

Two more checks that people skip:

• Bin codes and batch consistency. If you're buying the same SKU quarterly, ask whether the color bins hold across shipments. Cheap variance here shows up as a wall of slightly different whites two years later.
• Date codes. Old stock isn't defective, but it has already aged on the shelf. If the service life of the electrolytic capacitors in the driver matters to you — and it should — ask for the production date.

To be fair, most of this is standard practice for a good supplier. But 'good supplier' is exactly what you're trying to verify, and price lists don't tell you.

3. A bulk downlight quote — is that a different product?

Sometimes yes, sometimes it's the same fixture with different wording. And you cannot tell which from the quote alone.

I said 'the 6-inch downlight we spec'd last quarter.' They heard 'a 6-inch downlight.' Result: different trim, different junction box, and a fixture that didn't fit the ceiling cutouts we'd already cut. That was a $3,200 order and a three-day delay while we hunted for adapters that, as far as I can tell, nobody manufactures.

What I do now: I never order off a description. I order off the full manufacturer part number, and I make the supplier confirm it in writing on the quote. If they can't or won't put the exact part number on the document, that's a signal.

On a true bulk downlight SKU — the volume version of a fixture — check the driver wattage and the wiring configuration specifically. Those are the two places where a bulk version most often differs from the spec-sheet version, and they're the two that determine whether it installs into your existing circuit.

4. Are Cree LED grow lights worth it over cheaper boards?

Depends on what you're comparing and what you're measuring. Compare these three numbers, in this order:

1. Efficacy, in µmol/J. Not wattage. A 640W fixture at 2.1 µmol/J and a 720W fixture at 2.6 µmol/J are not the same purchase, even if the PPF on the box looks close.
2. PPF and the footprint it's rated for. PPF without a coverage area is half a spec.
3. DLC listing, if you're claiming a utility rebate. DLC maintains the Qualified Products List for horticultural lighting. Verify the listing is active at designlights.org, and check the current technical requirements version — a lapsed listing is a dead rebate.

Honestly, I'm not sure why two fixtures with nearly identical PPF and efficacy numbers can perform so differently under the same canopy. My best guess is spectrum distribution and how aggressively the driver derates as the fixture heats up. I've never fully understood the second half of that, and if someone has real insight, I'd like to hear it.

What I can tell you: thermal design is where cheaper boards lose. Run them at full power in a sealed room and compare output at 1,000 hours versus 100 hours. The ones that hold their output are the ones you want to buy twice as many of. At least, that's been my experience.

5. What does nobody warn you about on bulk orders?

Controls. Everyone budgets for the fixtures and nobody budgets for making the fixtures talk to each other.

The surprise on our last warehouse job wasn't the price difference between vendors. It was that the fixtures arrived, installed fine, powered up fine, and then wouldn't talk to the existing gateway. Wrong firmware revision, and the manufacturer's commissioning tool assumed a controller we didn't own. The hardware was fine. The system was not.

Three questions to ask before the PO:

• What protocol does this actually speak — 0-10V, DALI, Zigbee, something proprietary?
• What firmware revision ships, and can it be updated in the field?
• Is commissioning in the quote, or is it a separate line item I'm going to find out about later?

If the answer to the third one is 'separate,' get the number. I should add that in my experience it's rarely small.

6. My spec sheet is from 2019. Is it still valid?

Parts of it, no.

In 2019, 100 to 110 lumens per watt was a respectable commercial spec. As of January 2025, that same number would be a red flag on a new order. Controls that were add-ons five years ago now ship as standard on a lot of commercial lines, which changes the wiring, the commissioning, and the labor estimate. DLC technical requirements have moved through several versions since then, and a spec that qualified under an old version may not qualify now.

What was best practice in 2019 may not apply in 2025. That's not a knock on the older specs — they were correct for the products that existed at the time.

What hasn't changed: thermal management, driver quality, and the fact that a cheaper fixture with a non-replaceable driver is usually the more expensive fixture over ten years. The fundamentals are the same. The execution isn't.

Practically: re-verify any spec sheet older than two years. Re-check the DLC listing. And re-check whether the controls protocol you standardized on back then is still the one your supplier is shipping today.

7. What's the single most expensive mistake you've made?

A $3,200 order, referenced above, that I approved without the exact part number in writing. But that's the specific one. The pattern behind it is worse.

Every expensive mistake I've made came from the same place: I was confident about something I hadn't verified, because I'd verified it once, a year ago, on a different job. Same supplier. Same product family. Different SKU.

The checklist I keep now has 22 items on it. We've caught 47 potential errors with it in the past 18 months — most of them small, a few that would have been expensive. In my opinion, the value of a checklist isn't the items on it. It's that it forces you to check the thing you're sure about.