Why I'm comparing these two paths at all
I've been handling lighting specification and procurement orders for eight years — commercial retrofits mostly, some industrial, a handful of outdoor and specialty jobs. I've personally made, and written down, 11 significant sourcing mistakes worth roughly $46,000 in wasted budget. That list is taped inside a cabinet in my office now. This comparison comes straight off it.
When people ask me about sourcing lighting, they usually frame it as brand versus no-brand. That's not the decision. The real fork is this:
Path A — Branded lamp, standard fixture. You spec a socketed or standard-base fixture and buy branded lamps separately. In commercial work, cree led bulbs are the common reference point — people spec them because the binning and the published test data are predictable, and because a spec reviewer will recognize the name.
Path B — Integrated or OEM-supplied assembly. You buy the fixture with the LED module, driver, and optics already matched and documented as a system by a light fixture oem. You're not buying a lamp. You're buying a light-producing assembly and a paper trail.
Both paths work. Both fail. What decides which one fails is the same four things either way, so let's take them one at a time.
Dimension 1: Photometric consistency, lot to lot
Path A buys you binning discipline. The reason a named lamp costs more isn't marketing — it's that the manufacturer sorts emitters into tighter color and flux bins and then ships you what the label says. In a multi-phase project where the second batch arrives four months after the first, that's the whole ballgame.
Path B depends entirely on the OEM's process control. A good one buys from the same tier-one emitter suppliers and holds the bin. A less careful one buys whatever's available and calls it "warm white."
And here's the part that gets missed: Path A only protects you if you hold the lamp SKU stable. I watched a project manager swap lamps mid-job to save about $0.80 per lamp on a 600-lamp order — roughly $480. Two months later, 140 fixtures in one corridor read visibly warmer than the 460 installed earlier. The correction cost us around $3,100 in labor and replacement lamps, plus a very awkward walkthrough.
Verdict: Path A wins on consistency, but only under a locked-SKU policy. Path B matches it when the OEM will commit to a bin and put it in writing. If they won't name the bin, they don't have one.
Dimension 2: The paper trail — and the causality I had backwards
For a long time I assumed you pay more because the components are better. The assumption is that price causes quality. The reality runs the other way, at least in commercial lighting: consistent, documented performance is expensive to produce, and the price follows from that. What you're actually buying on Path A isn't a nicer diode — it's repeatable measurement and traceability.
When I do recessed lighting sourcing or qualify a new track lighting supplier, these are the documents I ask for, for the exact SKU being quoted:
- LM-79 photometric test report from an accredited lab (lumens, wattage, efficacy, CCT, CRI, beam distribution)
- LM-80 lifetime data on the emitter, plus the TM-21 projection if anyone is claiming a rated life
- UL listing — UL 1598 for the luminaire, UL 8750 for the LED equipment, or an equivalent NRTL listing
- DLC listing status if utility rebates matter to the project
- Driver datasheet with the actual model number, not "constant current driver"
If a supplier can't produce those for the SKU they're shipping, they aren't cheaper. They've just moved the risk from their balance sheet onto yours.
Under FTC guidance on advertising claims, a claim has to be truthful, not misleading, and substantiated with evidence before it runs. In lighting terms, the lumens, wattage, and life numbers aren't decoration — they're claims. The FTC's Lighting Facts label (16 CFR Part 305) has required brightness, estimated yearly energy cost, life, light appearance, wattage, and mercury content on bulb packaging since 2010.
Verdict: This dimension has nothing to do with brand. Path A usually arrives with the documents already done, which saves you two weeks. Path B can match it — I've had OEMs hand over LM-79 reports that were cleaner than the brand-name folder. But you have to ask, and you have to check that the report matches the SKU on the PO, which is where the failures hide.
Dimension 3: Serviceability from year two onward
Path A looks simple: the lamp is the service part. Burn one out, replace it. That's real, and it's a genuine advantage in high-bay and parking applications where access is a lift rental every time.
But there's a catch nobody prices in. In 2019 I standardized a 90-fixture retail job on a specific lamp. In 2023, three years into the client's service agreement, that lamp line was discontinued. I ended up buying remaining stock from three different distributors, and one close-out lot that was a different bin. The client has a corridor with slightly mismatched light and a maintenance log nobody wants to look at.
Path B sounds worse on paper — replacing an integrated module means a module, a driver, or the whole fixture. But good integrated fixtures are often more serviceable than cheap socketed ones, because the driver is field-replaceable and the module is connectorized. I've swapped a driver in four minutes with a screwdriver. I've also opened a budget integrated fixture that was fully potted with no access panel, which is the worst possible version of both paths.
Verdict: "Socketed versus integrated" is the wrong question. The right question is: what is the service part, what's its part number, and can I still buy it in year five? Ask for a service-parts commitment in writing. If the answer is "the fixture," price the whole fixture into the TCO.
Dimension 4: MOQ, lead time, and the cost of changing your mind
Path A is flexible. Low MOQs, distributor stock, next-day availability on common lamps, and you can adjust count mid-project. You pay for that flexibility in unit price and in having two supply chains to manage instead of one.
Path B is the opposite. MOQs are real, tooling and sample approval take weeks, and once the PO is cut, changing anything is expensive. The hidden cost isn't the unit price — it's the change order.
I once wrote "match existing fixtures" in a spec and approved a sample against it. The supplier read "close to existing." We found out when 60 fixtures went up next to the original batch and one run of track read noticeably greener. Both of us had used the same four words and meant different things. That mistake cost about $2,100 in module swaps and a week of electrician time — more than the $1,400 we'd saved by going OEM-direct on that line in the first place.
To be fair, that one was mine, not the supplier's. A written color and CCT target with a tolerance would've prevented all of it.
Verdict: under roughly 50 fixtures with a loose spec, Path A almost always wins. Above 200 fixtures on a fixed spec, Path B usually wins on total cost — but only if the spec is locked before the PO, not after.
How I actually decide now
It took me five years and about 400 fixture orders to stop treating "brand" as the deciding variable and start treating "documentation plus service parts" as the deciding variable. Here's the current decision rule:
- Small job, fast turnaround, mixed existing lamps: Path A. Branded lamps and a standard fixture. Don't overthink it.
- Large new construction, single spec, phased over months: Path B. Lock the bin, lock the driver, get the reports before the PO.
- Retrofit where the original lamp is already discontinued: Path B, integrated. You're never matching the old light anyway, so stop paying for the attempt.
- Track lighting specifically: add one dimension to the list — whether the track, the heads, and the connectors come from the same production system. Mixing brands on a track run is where adapter and polarity problems live.
My honest read, after eight years of this: in about 60% of the projects I've managed, the lowest quote ended up costing us more. Not because cheap is bad — because the quote only covered the fixture, and it never covered the rework, the mismatched corridor, the discontinued lamp, or the four hours the electrician spent on the phone with me. Total cost of ownership includes all of that.
So I don't ask "which is better, branded lamps or OEM-direct?" anymore. I ask two questions: who's holding the documentation, and who's holding the service parts in year five. Whoever answers both gets the order.
