The Short Version
Sticker price on an LED tube is roughly 30% of what it actually costs you. The other 70% lives in driver failures, warranty fights, and the labor to swap them out mid-project. That's the number from our own tracking system after auditing about $180,000 in cumulative lighting purchases across six years. If your sourcing decision is still "which quote is lower," you're optimizing a variable that doesn't matter as much as you think.
Everything below is what I actually check now, and the case that changed how I buy.
Context: Why This Is Worth Listening To
I'm a procurement manager at a 220-person facilities services company. Our lighting and electrical category runs about $65,000 annually. I've been in this role six years, negotiated with 15+ vendors, and — this is the part that matters — I log everything. Every invoice. Every warranty claim. Every "this failed again" email thread.
If I remember correctly, we're past 340 line items in the lighting category since 2019. If you'd asked me in year one whether that tracking was worth the effort, I'd have said no. I built the system after the third time a "cheap" order came back and bit us. Should've built it after the first.
The Order That Rewrote Our Sourcing Rules
Q2 2023. I needed 500 4-foot LED tubes for a retrofit.
Vendor A — a Sylvania distributor we'd bought from for years — quoted $6.40/unit. Vendor B quoted $4.10. Same color temperature on paper. Same lumen output. On a 500-unit order, that's a $1,150 difference. My gut said stick with A. The spreadsheet said take B. I went with the spreadsheet. Seemed like a no-brainer.
Eleven months later, 62 tubes had failed. Not the LEDs — the integrated drivers. The datasheet said 25,000 hours. What it didn't say was that the LM-80 test data behind that claim was collected at a case temperature that had nothing to do with how those tubes actually sit inside a closed-back recessed fixture.
Parts and labor ran about $1,400. Two weeks of project delay. Emergency after-hours swap-outs added roughly $2,500 more. Net result: the $1,150 we "saved" cost us about $2,750 in real money, plus a very uncomfortable call with the client.
The most frustrating part of that whole experience: the vendor's response was "spec sheet says 25,000 hours." Technically true. Practically useless.
What I Check Now, In Order
1. LM-80 data, reported with the test temperature
Anyone can print "50,000 hours" on a box. What actually matters is the LM-80 report — from an accredited lab, following the IES LM-80 standard — and the case temperature (Ts) at which that data was collected. TM-21 projections are only meaningful when you know the LM-80 conditions they're based on.
Per IES LM-80 and TM-21 standards, lumen maintenance claims should always be reported alongside the test case temperature. A 50,000-hour claim measured at Ts=25°C tells you almost nothing about a driver running at 75°C inside an insulated ceiling.
When I went back and asked both vendors for this, Vendor A's contact had Sylvania LED driver reports already on file and sent them within 20 minutes. Vendor B sent a two-year-old PDF for a different product line. That comparison was worth more than the $1,150 we "saved."
2. Compliance, in writing, before ordering
Anything going on a commercial spec now gets checked for:
- UL or ETL listing — actual listing, not "pending"
- DLC qualification, if there's any utility rebate attached
- RoHS compliance statement
- FCC Part 15 for anything with a driver
We lost a $2,100 rebate last year because the fixtures weren't on the DLC qualified products list. Nobody checked. That one still stings when I look at the spreadsheet.
3. Driver replaceability — the spec nobody asks about
When a driver dies — and it usually dies before the LED does — what happens next? Can you swap it, or do you scrap the whole fixture?
For our internal retrofit program, we've standardized on Sylvania LED drivers for most interior applications. Not because they're the lowest line item. Because when one fails, I have a replacement on the shelf and the swap takes 10 minutes. That's worth more than a $2 difference per unit up front.
The same logic applies to any track lighting wholesale order, or to fixtures from any light fixture manufacturer. If the driver isn't a standard form factor and isn't field-replaceable, you're buying future disposal costs disguised as a product.
4. Case temperature — not ambient temperature
I wish someone had told me this in year one. Driver life depends on case temperature (Tc), not the temperature of the room. Two identical fixtures in the same office can have case temps 30°C apart depending on insulation and airflow. If a supplier can't tell you what Tc their driver was tested at, you're comparing apples to nothing.
Same rule applies when you're buying Sylvania LED bulbs or stocking house-brand tube replacements — the bulb may be fine, but the driver behind it determines how long anything actually lasts.
Where This Breaks Down
Honest limits:
This TCO approach works at volume — call it 200+ units per order, or a rolling replacement program. If you're a small distributor buying 20 tubes to fill a shelf, running down LM-80 reports and DLC listings probably costs more in admin time than the failure risk is worth. Pick a reasonable supplier and move on.
My numbers are also from a facilities services context. Industrial high-bay and cold-storage lighting have different failure profiles — the driver is still the weak point, but the temperature stress swings the other direction.
And the whole system assumes your supplier will answer spec questions before you buy. Some won't. That's information too. If a vendor won't answer pre-sale technical questions, they're certainly not going to answer post-sale warranty questions. I've learned that lesson more than once.
Bottom line: track your actual failure rate for 12 months. Then work backward and figure out what you should've bought. My guess is you'll find the "premium" option was the budget option all along — you just couldn't see it on the quote.