I've been handling smart lighting sourcing for 9 years. I've personally made — and documented — 11 significant mistakes, totaling roughly $27,400 in wasted budget. Two of those were the same mistake in opposite directions: I went too "private label" on one product line, and too "OEM" on another. Both cost me real money.
Here's the thing nobody told me at the start: "OEM vs private label" is not one decision. It's three. And the right answer flips depending on which one you're actually in.
The rough split:
- Under 1,000 units, no validated demand yet → Scenario A
- 2,000–10,000 units with an existing customer base → Scenario B
- Project-based work with contract compliance requirements → Scenario C
If you don't know which one applies to you yet, read all three, then use the checklist at the end. You'll notice something counterintuitive along the way: the default "just start with private label" advice you'll hear from most people is wrong roughly half the time.
Scenario A: You're under 1,000 units and haven't validated demand yet
Recommendation: private label — but with one non-negotiable condition. You spec the driver and control components yourself, even if the housing is off-the-shelf.
Everyone tells you not to bother with OEM for a first run. That's mostly right. But "private label" isn't the same as "whatever the supplier already uses."
Here's what happened to me. In early 2022, I sourced 600 ceiling fixtures as a private-label product. The supplier's standard driver was a no-name DALI-compatible unit that worked fine on the bench. Six weeks into installations, we started getting flicker complaints. DALI addressing wouldn't hold across certain dimming curves.
The problem: I assumed "DALI compliant" meant interoperable. It doesn't. Certification matters, and the specific driver model matters just as much as the protocol label on the box.
I'm not a lighting controls engineer — that's not my background. So I can't speak to what happens inside the driver firmware. What I can tell you from a procurement perspective is: in smart lighting, the driver and control layer is the product. The housing is cosmetic. If you private label the fixture but let the supplier pick the driver, you've outsourced the part that actually matters.
What to do instead: spec the driver yourself, even on a small run. Reference a properly certified part built to IEC 62386 — a Tridonic DALI driver, for example, is one of the reference-grade options buyers ask for by name. You'll pay a small premium. You'll avoid a rerun.
What that mistake cost me: $4,300 in replacement drivers plus a three-week delay and two cancellations.
If you've ever had a customer return an entire batch over flicker, you know that sinking feeling. It's not the driver failure that stings. It's that you should have specified the driver yourself.
Scenario B: You have 2,000–10,000 units and an existing customer base asking for specific features
Recommendation: hybrid. OEM the driver and control interface. Private label everything else.
Here's the counterintuitive part. When you hit medium volume, the instinct is to go full OEM — custom housing, custom PCB, custom everything. That's usually a mistake at this stage. You'll spend on tooling for features that two customers asked about and nobody else wanted.
What you should actually do is narrow the OEM investment to the part that drives customer switching behavior. In smart lighting, that's control. Zigbee module integration, DALI drivers, sensor wiring, commissioning software compatibility — this is what buyers walk away over. Housing aesthetics matter less than you think at this volume.
I learned this the hard way. In September 2023 I committed $8,700 to a custom aluminum housing for a mid-volume line. Two customers had asked for a smaller profile. The tooling cost only made sense across the full forecast — which, of course, we missed. We shipped the old housing with a new driver instead, and nobody complained.
What I'd do differently: split the order. Take a proven chassis. Put your OEM budget into the driver and control module. If Zigbee is part of your architecture, this is the stage where the Zigbee supplier choice really matters — a supplier who can hand you certified modules with proper stack documentation saves you a couple of months on integration and certification. Anyone who can't give you that paperwork is a red flag at this volume.
One scope warning: I can only speak to North American commercial installs. If you're dealing with EU ErP compliance or Australian MEPS, the certification burden changes the math significantly.
Scenario C: Project-based work with contract compliance requirements
Recommendation: OEM for the assembly, but with name-brand components specified in the BOM. Do not let the manufacturer substitute — not even "equivalent" parts.
This is where I have the strongest opinion, because this is where I've seen the most expensive mistakes — mine and others'.
When you're bidding on commercial projects — offices, hotels, retail, and especially anything with a written lighting control spec — the actual product isn't the fixture. It's the compliance documentation. You need to be able to prove the driver is DALI-2 certified, the Zigbee module is certified, and the whole assembly passes the tests the spec calls for.
I made this mistake once. We specified Tridonic DALI drivers on a 1,200-unit project. The manufacturer substituted a "100% compatible" equivalent for 300 of those units — after we'd approved the BOM. Nobody caught it until commissioning. The substituted drivers worked fine. They just weren't documented for the spec. We had to redo the compliance submission from scratch.
Cost: $2,100 in documentation work plus a two-week hold on final payment.
The rule that came out of that: if the spec names a component family, the BOM locks to that family, and the contract requires written approval for any substitution — from a named person at your company, not "the supplier's discretion."
Reference point: DALI interoperability is defined under IEC 62386. If your spec just says "DALI," confirm which parts of 62386 apply to your project — control gear, control device, and bus power each have separate parts. Ask for the certificate, not the compliant claim.
How to figure out which scenario you're actually in
Don't answer this by gut. Answer it with four questions:
- How many units will you actually ship in the next 12 months — not how many you hope to? Under 1,000 → Scenario A. 2,000–10,000 → Scenario B. Project-based → Scenario C, regardless of count.
- Who asked for the customization? If a customer asked, they're paying for it (Scenario B or C). If you asked, you're paying for it — and the question is whether that's marketing spend or waste.
- What's the failure mode if a component fails? If a single flicker complaint triggers a batch return, the driver specification is not a place to save money. That pushes you toward Scenario C behavior regardless of volume.
- Who owns the compliance document? If the answer is "the manufacturer," you're already in Scenario C whether you planned to be or not.
One more thing, and this goes back to the small-order question that nobody in procurement likes to hear. Small orders usually get treated as small risk. In smart lighting, that's backwards. A 200-unit order that fails on control compatibility costs you the customer, the review, and the reference. I've kept vendors who treated my $200 orders seriously. They're the ones I still use for $20,000 orders.
If you're under 1,000 units and a supplier won't spec a proper driver because the order is "too small," that's a deal-breaker. Move on. The part matters more than the order size — and any supplier who doesn't get that is telling you something about how they'll handle your first serious problem.
