The phone call that tells you the comparison was wrong
March 2024. A hotel lobby refurbishment, 340 recessed downlights, six days to handover. The main contractor calls me at 7:40 in the morning because the lighting designer has just opened the first carton and the beam looks nothing like the IES file that went into the model.
We had done everything "right." Three quotes, spec sheets, a side-by-side comparison spreadsheet with columns for wattage, lumen output, CRI, color temperature, driver type, warranty. And no column at all for the things that were about to blow up the schedule.
What the sheet didn't say: the photometric data was a "typical" file from a reference build, not the production run we got. Nominal 3000K was closer to 3400K in the batch that shipped. The 38-degree beam was actually 46 degrees, which in a lobby with an 8-meter ceiling is the difference between a wash and a splotch.
We fixed it. It cost four days—well, three and a half once you count the Friday inspection—plus a ballroom we had to shut down twice, and about $9,000 in air freight to bring replacement optics in from a second supplier. Not a catastrophe. But it was avoidable, and what made it avoidable wasn't a better spreadsheet. It was a different question, asked three weeks earlier.
The question everyone asks, and the one they should
When I talk to project buyers and specifiers, the question that opens almost every conversation is: "How do I evaluate downlight manufacturers?" Usually followed by "who's cheapest" or "who has the most certifications."
The question they should be asking is: "Which of these suppliers can prove, in writing, what will actually arrive on the loading dock — and when?"
Those aren't the same question, and the gap between them is where most lighting projects go sideways. It took me about six years and a lot of expensive lessons to see it clearly. I sit between the specifier and the architectural lighting distributor on a lot of fit-outs, so I get to watch both sides of the handoff — and the handoff is where it breaks.
Most buyers compare catalogs. The failure lives in the supply chain.
Here's something vendors won't usually put in the email: the spec sheet you're comparing is a marketing document, not a measurement. It's real, it's often accurate, but "typical" values describe a range, and a production batch sits somewhere inside that range. Lumen output is easy to verify — IES LM-79-19 exists precisely to make that measurement repeatable — but the number on the PDF is a claim, not a promise about your order.
IES LM-79-19 specifies how to measure the photometric and electrical characteristics of LED products. It does not require that every production unit match a published "typical" value.
What most buyers focus on is the column that's easiest to fill in: lumens per watt, unit price, CRI. What they miss is the column that decides whether the project succeeds — the documentation trail and the lead-time reality behind the number.
Three things I've learned the hard way:
1. "Standard lead time" is a queue-management number, not a promise about your order. When a distributor quotes 4-6 weeks, that window often includes buffer they use to keep their own production calendar healthy. Your order may take three weeks or nine. That's not dishonest — the standard window just isn't calibrated to your deadline. It's calibrated to theirs.
2. Certification documents show up late, and late is the same as missing. I've had projects where the fixtures were in the building and the install crew was standing around because the UL listing paperwork for the specific driver configuration hadn't come through. The fixture was compliant. The proof wasn't there yet. On a fit-out with a fixed handover date, "compliant but undocumented" is a stop-work order.
3. A great sample does not predict a great batch. This is the one that still gets me. We once approved a designer pendant — the kind of statement piece that lifts an entire atrium, a large moooi paper chandelier — based on a single gorgeous unit. The production run had visible color-temperature variance between fixtures. On a small pendant, nobody notices. Across a lobby, four meters of paper chandelier reading three slightly different whites, that variance looks like damage.
That last point matters more in designer lighting than in commodity fixtures, which is why a house like moooi pours so much energy into controlling the gap between prototype and production line. The form is the function. Finish and color consistency are part of the spec, not a bonus on top of it.
What the deeper failure actually costs
Let me put real numbers on it, because "get the spec right" doesn't move anybody.
On the hotel lobby job: roughly $9,000 in expedited freight, plus about $14,000 in labor to redo the ceiling install on 140 of the 340 fixtures. That's the direct cost. The number that actually mattered was in the main contractor's contract — a $2,500-per-day liquidated damages clause for late handover. Four days of delay would have been $10,000. We ate it by working nights, so it didn't hit the clause. That time.
The pattern I see is consistent across projects:
- Rework labor usually runs 1.5-2.5x the original install cost on the affected fixtures, because you're paying for removal, disposal, and a second install.
- Expedited freight on replacement units runs 3-8x standard freight, and that's before customs or brokerage charges.
- Schedule float disappears fast. A project with two weeks of buffer can absorb one mistake. It cannot absorb two.
And the cost that never shows up on an invoice: the next time you spec lighting, you over-spec. You pad lead times, you add contingencies, you pay for redundancy you don't need. That's the long-term tax of a failed project — it teaches you to distrust the whole category.
The process gap that caused it
We didn't have a formal photometric verification step. That's the honest version. We had a sample approval and a spec review, but nothing that said "before release to production, confirm that the published file for this exact SKU and driver matches the fixture we're buying."
The third time a beam angle came back wrong, I built the checklist I should have built after the first time. Five questions. They won't guarantee a perfect project — nothing does — but they filter out the suppliers who are going to waste your schedule.
Five questions that actually predict delivery
Ask these before you sign. Ask them about your worst-case fixture, not your easiest one. They apply whether you're buying downlights, working through track lighting specifications in a distributor catalog, or specifying a custom piece.
- Can you send the IES file for this exact SKU and driver combination, dated within the last 12 months? Not a typical file. Not a family file. This SKU. If they hesitate, that's a red flag, and it's a deal-breaker.
- What's the measured color consistency bin, and how tight is it? Ask for SDCM or MacAdam ellipse data. "3000K" by itself tells you nothing. A supplier who knows their process will answer this in seconds.
- When, specifically, will the compliance documentation arrive relative to the goods? Before, with, or after? Put it in the purchase order.
- If your first production run fails your own QC, what's the recovery plan — and what does it cost me in time? "We'll make it right" is not an answer. A schedule is.
- Who does the punch list, you or me? On complex designer fixtures, someone has to own the final visual inspection. That someone is either a line item in the quote or it's your site team at 11 p.m.
Pricing will vary. Ballpark, expect the suppliers who can answer all five to sit 10-20% above the ones who can't. That premium isn't for the fixture. It's for the certainty that what's at the dock is what you specified, on the day the schedule needs it.
Even after we confirmed the second optics supplier could hit the date, I kept second-guessing. What if their beam was off too? I didn't relax until 11 p.m. the night before handover, when the last fixture tested clean. That's what certainty costs you — and honestly, it's cheaper than the alternative.
Bottom line: in lighting, the cheapest quote and the most expensive quote are often the same fixture. The difference is what's included. And what you're really buying when you pay more is the answer to question four.
If you're specifying statement pieces — a chandelier chosen for how it reads in daylight, a paper pendant that has to look identical across the whole run — the conversation shifts a little, because form is part of the deliverable. But the evaluation logic doesn't change. You still want the supplier who can tell you what will arrive, and when, and back it up.
Looking back at that March morning, the fix wasn't a better spreadsheet. It was one more question, asked when it would have cost nothing.
