The Quote That Looked Too Good
When I started sourcing grow lights for our greenhouse expansion in early 2024, I thought I had it figured out. Send the same specs to eight manufacturers, compare the quotes line by line, pick the one with the best price-per-unit. Simple, right?
The lowest quote came back at roughly 30% below everyone else. Same wattage. Same claimed spectrum range. Same 5-year warranty language. I nearly signed it that week.
I didn't. And that decision—mostly made out of caution, not wisdom—ended up saving us a costly lesson further down the line. But it also taught me how badly the standard "compare three quotes" approach fails when you're buying horticultural lighting at scale.
What the Spec Sheet Doesn't Tell You
Here's what I didn't understand at the time: a grow light spec sheet is not a product. It's a marketing document. Two manufacturers can list identical wattage, identical PPF, identical "full spectrum" language, and still deliver fixtures that perform completely differently in an actual greenhouse.
The deeper issue is that most bulk buyers—myself included—evaluate lighting manufacturers using procurement logic that works fine for office furniture and fails for photobiology. We compare numbers on paper. But the numbers that actually determine ROI aren't on the first page of the spec sheet.
Three things the cheap quote glossed over:
1. Spectrum delivery vs. spectrum claim. A manufacturer can claim "full spectrum" while delivering peak output in bands that don't match your crop's photomorphogenic needs. Without third-party spectral data (not a marketing diagram), you're trusting a claim you can't verify. Some manufacturers will provide IES files or DLC-listed test reports. Others won't—and that silence is itself a data point.
2. Driver and thermal architecture. The first thing to get cut in a race-to-the-bottom quote is the driver quality and heat sink design. These don't show up in the headline specs. But they determine whether your fixture maintains its output at hour 20,000 or starts degrading at hour 8,000. In a commercial greenhouse running 16-18 hours a day, that gap is enormous. I've since learned to ask specifically: what driver brand, what junction temperature rating, and what's the projected lumen maintenance curve?
3. Compliance and listing status. If your facility needs DLC listing for utility rebates—and many do—a fixture that isn't listed isn't cheaper. It's disqualified. I've watched a colleague discover this after placing a six-figure order. The rebate loss alone erased any per-unit savings.
The Cost Nobody Quantifies Upfront
This is where the "cheap" quote really starts to unravel. The purchase price is maybe 40-60% of what you'll actually spend on a lighting installation over its useful life. The rest hides in places procurement teams don't usually look.
- Installation fit. Fixtures that don't match your existing mounting infrastructure require adapters or custom brackets. That's labor, downtime, and often a change order. One manufacturer we evaluated had a 4-week lead time on mounting hardware—after the fixtures arrived.
- Spectrum mismatch cost. If the delivered spectrum underperforms, you don't just lose yield. You lose the entire grow cycle's worth of time, energy, and inputs that went into those plants. That's not a product cost—it's an operational cost.
- Replacement logistics. A 5-year warranty means nothing if the manufacturer doesn't have a local RMA process or if shipping replacements takes 6 weeks. During peak season, that's a non-starter. I now ask for the actual RMA turnaround they've achieved in the past 12 months—not the policy, the practice.
- Specification drift. This one caught me off guard. Some manufacturers quietly change components (driver, diode bin, housing material) between the sample you approve and the bulk shipment. If your PO doesn't lock the BOM, you might get a different product than you tested.
The most frustrating part of this whole process: the manufacturers who cut corners rarely advertise it. You find out through a failed crop, a rejected rebate application, or a fixture that dims faster than expected. By then, you've already paid.
How I Evaluate Grow Light Manufacturers Now
I've since built a checklist that goes beyond price comparison. It's not glamorous, but it's saved us from at least two bad decisions.
Ask for the same data from every vendor. Specifically: PPF map at 12" and 24", spectral power distribution data from a third-party lab, driver manufacturer and model number, thermal test results, and DLC or UL listing documentation. If a manufacturer can't provide these, that's your answer.
Request a sample under your growth conditions. Not a demo in their showroom—a fixture in your facility, running your photoperiod, on your crop. Two weeks of side-by-side data will tell you more than any spec sheet. A manufacturer confident in their product will support this. The ones who won't probably have a reason.
Calculate TCO, not unit price. Include: fixture cost + shipping + duties + installation labor + expected energy cost over 5 years (at your electricity rate) + estimated replacement rate + rebate value (if applicable) + the cost of one failed grow cycle if spectrum underperforms. The gap between the lowest quote and the best TCO is often 20-40% in the other direction.
Verify the manufacturer, not just the product. Are they actually manufacturing, or are they sourcing from a third party and rebranding? There's nothing wrong with a trading company per se, but it changes your support model, your ability to request custom specs, and your leverage if something goes wrong. Ask direct questions: where are the fixtures assembled? Can they do private label with your specifications? What's their MOQ for custom spectrum work?
None of this is revolutionary. But I've watched too many buyers—myself included—skip these steps because the quote looked good and the timeline was tight. The cheap quote is almost never the cheap outcome.
That initial order I almost placed? I later found out that same manufacturer had been dropped by two other greenhouse operations in our region for driver failures within 18 months. We ended up going with a mid-priced option with verifiable test data and a responsive engineering team. Higher upfront, significantly lower total cost.

