There is no single right way to buy LED lighting
About once a month someone forwards me a lighting quote with a one-line message: "Is this good?" My answer is usually another question. Good for what? A $3.10 LED bulb and a $14 downlight can both be good buys, depending entirely on whether you're matching a lamp that already exists inside a fixture, or choosing the fixture yourself. Get that distinction wrong and the savings evaporate downstream, usually in a way nobody notices until the second order.
I'm the office administrator for a 220-person company. I manage facilities and MRO ordering — roughly $180,000 a year across about ten vendors — and I report to both operations and finance, so when a lighting order goes sideways I hear about it from two directions. When I took over purchasing in 2021, we were buying bulbs four different ways depending on who placed the order. In our 2024 vendor consolidation project I finally got the categories separated properly.
What follows is that separation. Three situations, three different approaches, and some honest guidance on figuring out which one you're actually in — because I've wasted money applying the wrong one in both directions.
One boundary up front: I'm not a lighting designer and I'm not an electrical engineer. I can't tell you whether 3500K or 4000K looks better in your reception area, and I can't sign off on a photometric layout. What I can tell you is how to keep the buying side of it from costing you money.
The three situations
- Situation A: You're replacing lamps, not choosing lighting. Something burned out. You want the same thing that was there.
- Situation B: You're changing the lighting in a space you control — a retrofit, a refresh, two floors of downlights.
- Situation C: You're committing to a product line. Multiple sites, a new build, private label, or a spec you'll have to live with for the next five years.
The expensive mistake is almost never picking the wrong vendor. It's applying Situation C diligence to a Situation A purchase (you pay three times the unit price for oversight you didn't need), or applying Situation A diligence to a Situation C purchase (you save four dollars a fixture and inherit a warranty problem).
Situation A: You're matching, not choosing
This is 80% of our lighting volume by order count and maybe 15% by dollars. Corridors, stairwells, the loading dock, the parking lot floods. The goal is boring: buy the same thing again, from someone who will take it back if you get it wrong.
What actually matters here
Base type first, always. We once ordered 60 lamps with an E26 base for a corridor where the fixtures took GU24. That was a few hundred dollars of lamps that fit nothing, and the return freight came out of my budget because the PO clearly said E26 and the vendor shipped exactly what I ordered. Nobody was wrong except me.
Then lumen output, Kelvin, and whether the thing is genuinely dimmable with the dimmers you already have. That last one is where the language gets slippery. I asked a rep for "dimmable 3000K, same as before." They heard "these are dimmable-rated lamps." Result: the lamps were rated dimmable, our wall dimmers were old TRIAC units from a 2009 fit-out, and the whole floor buzzed at anything below about 60%. The lamps weren't defective. The pairing was.
For outdoor floods specifically — this is where I'd point anyone toward a catalog rather than a guess. A Sylvania flood light that's an integrated LED fixture isn't a bulb replacement at all. You're buying a fixture, which means you're back in Situation B whether you wanted to be or not. Check whether there's a lamp inside the housing before you assume there is.
The part that surprises people
For small quantities, going direct to a manufacturer is usually worse. I know that sounds backwards — everyone wants to cut out the middleman. But a factory does not want your 40-piece order, will not stock your odd color temperature, and will not have a returns counter when four lamps show up dead in the box. Your local distributor charges a markup and in exchange eats that problem. On a $900 order, a distributor who takes returns the same week is worth more than $60 of per-unit savings.
Bottom line: in Situation A you're buying availability and returns, not performance. Price the distributor, not the factory.
Situation B: You're changing the lighting in one space
Now it's a project. Forty to three hundred fixtures, one building, and a visible outcome that people will comment on. This is where downlight sourcing and spotlight catalog work actually starts to matter, and where I learned my most expensive lesson.
You need files, not a price sheet
Any supplier doing real work in this space will hand you IES photometric files without being asked twice. Those files are what let a designer or contractor show you what the room will look like before you buy it. If a supplier can't produce them, that's not a paperwork gap. It's a red flag about whether the performance numbers on the spec sheet mean anything.
For downlights, the questions that burned us were physical, not electrical. Cut-out diameter. Whether the housing is IC-rated (i.e., safe to touch insulation in the ceiling) or not. Junction box depth. And, critically, whether the driver is field-replaceable. We installed 120 recessed units in 2022 where the driver was sealed into the housing. Two failed in year three. Both times we replaced a perfectly good fixture because we couldn't replace a $30 component inside it.
For spotlights and accent track heads, the catalog numbers to compare are beam angle and center beam candlepower, not just wattage. Two 10W heads can look completely different on a wall depending on whether one is a 15° narrow spot and the other is a 40° flood. And check the color consistency language — if the document doesn't mention binning or a color tolerance, assume there isn't one, and assume that when you reorder in eighteen months the new batch will be a slightly different white.
If utility rebates are part of your budget justification — and for us they were — the fixture has to be listed in the current qualified products database for the program, with an ID you can look up yourself. Do not take a rep's word on this. One of our vendors submitted 2022 listing numbers in a 2024 quote, and the rebate that made the project pencil out never materialized.
This was accurate as of early 2025 based on what we ran into. Qualification lists and rebate programs change on their own schedule, so verify current listings with the program administrator before you sign.
Where the risk actually sits
The upside of the cheaper fixture was about $2,800 across the project. The risk was a driver design we couldn't service, in a ceiling grid we'd have to re-open. I kept asking myself whether $2,800 was worth being the person who explains to finance why we're re-doing a floor two years after finishing it. We went with the more expensive one. I'd make the same call again, and I'd still feel like it might have been the conservative choice rather than the smart one.
Also worth saying plainly: no single fixture line is compliant for every market, every voltage, and every project type. Anyone who tells you otherwise is selling, not advising. Match the listing to the job.
Situation C: You're committing to a product line
Multi-site rollout, new construction, or private label. Here the question shifts from "is this fixture good" to "is this manufacturer going to be the same manufacturer in three years." That's a different evaluation entirely, and it's the one most people do badly because they evaluate samples instead of evaluating process.
How to evaluate a lighting manufacturer
Ask for these and verify the ones that can be verified:
- Safety listing file number (UL or ETL), not just the logo on the box. Look it up.
- Qualified products listing ID for commercial rebate eligibility, if that matters to your project.
- LM-80 test data and the lumen maintenance claim (L70, L90) with the hours attached. A lifetime number without test data behind it is marketing.
- Driver make and model, and whether it's replaceable in the field.
- Warranty terms in writing — the term, what it covers, who pays freight on a claim, and what the remedy actually is (replacement? credit? nothing?).
- Change notification clause. Will they tell you before they swap a driver or a bin? This one is worth more than the sample.
- Who holds the listing under a private label arrangement. If the certification lives under the factory's name and the relationship ends, you can be left holding inventory you can't legally sell into some jurisdictions.
- MOQ, lead time, and what happened to their lead time during the last component shortage. Ask for specifics, not reassurance.
Here's the counterintuitive part: a great sample tells you almost nothing. Any factory can hand-polish ten units for a buyer. What tells you how the relationship will feel in year three is the boring paperwork — the change notification process, the warranty claim path, and whether the person answering your email has authority to make a decision. Samples prove capability. Clauses prove behavior.
Looking back, on our first private label conversation I should have asked about listing ownership in the first meeting instead of the fourth. At the time it seemed like a legal detail, and I was focused on unit cost and packaging. It turned out to be the deal-breaker — or rather, the thing that would have been one if we'd caught it later.
So which one are you in?
Three questions, and they're usually enough:
- Are you matching a fixture that already exists, or defining a new one? Matching goes to Situation A, full stop. Defining goes to B or C.
- Will you need to reorder this exact SKU in three years? If no — a one-time floor, a leased space — you're in B. If yes, and especially across more than one site, you're in C whether or not the budget feels like a C.
- Who absorbs the cost if this fails? You, out of an operating budget, with a distributor who takes returns? That's A or B. A client, a landlord, a spec you have to certify? That's C, and the diligence level should match.
If you're not sure on any of the three, assume B. It's the safest default — you'll get real spec documentation without paying for a supplier audit you don't need yet.
The last thing I'd say is that none of this is about finding the cheapest lamp. It's about spending your attention in the right place. Ninety percent of our lighting orders should take ten minutes and a distributor phone call. The other ten percent should take three weeks. Confusing the two in either direction is how facilities budgets quietly leak.