Every fixture you spec is a business card. The moment a client walks into a space and the lighting feels off—whether it's a flickering landscape light or a chandelier that sits awkwardly against the ceiling—they're not thinking about the fixture. They're thinking about you, and whether they should have hired someone else. I learned this the expensive way, and I'm documenting it so you might not have to.
I've been handling commercial and residential lighting orders for nine years now, and I've personally made (and documented) 13 significant specification mistakes—totaling roughly $18,000 in wasted budget and rework. That's a conservative number. I keep a checklist now that's caught 47 potential errors in the past 18 months, but it exists because I made every mistake on it at least once. So if you're a contractor, a designer, or a facility manager who thinks lighting specs are just details, this is for you.
The Transformer That Burned Me (Literally)
Back in September 2021, I took on a landscape lighting project for a client who'd specifically asked for a system that required a Kichler 60-watt transformer. The spec was in the contract. I had it in writing. And then I swapped it for a different brand's transformer to save $42.
Here's the thing: the numbers said it was fine. The specs were similar on paper—comparable wattage, similar dimensions, same IP rating (or so I thought). My gut said stick with the Kichler. I've seen transformers fail before. But I went with the spreadsheet because the client was watching costs, and I wanted to look like the guy who could shave the budget. (Surprise, surprise—that didn't work.)
By month nine, the transformer started buzzing. By month eleven, it was dead. One section of the client's property went dark, including the area with their outdoor sports setup—they'd had me install some basic sports lighting on the same circuit, and it all ran at 80% brightness for three months before failing completely. The $42 I saved turned into a $1,850 redo, a full weekend of running cables, and a client who now questions every single line item on every invoice.
To be fair, the other transformer wasn't garbage. It was probably a fine product for a different application. But that's exactly the point: "probably fine" isn't a lighting spec. When you swap a component that was specified because it integrates with a system—the Kichler transformer has documented compatibility with Kichler photocells and LED drivers—you're not saving $42. You're gambling with the client's trust. And the house always wins.
The Library Chandelier That Won the Space (but I Made It Hard)
Beautiful jobs expose sloppy processes. I learned that in April 2023 on a library renovation.
The client wanted something sculptural, warm, and substantial in the main reading room—11-foot ceilings, dark walnut shelving, brass accents everywhere. We landed on the Kichler Eris 8 Light Chandelier (model 44276PN). I'd spec'd it on two previous projects and it's a genuinely stunning piece: polished nickel finish, that clean modern form, just the right presence without overwhelming the room. The client loved the sample photos. I loved that it was coming to a library, because library chandeliers are a category that rewards restraint, and the Eris is all restraint.
The mistake wasn't the fixture. The fixture was flawless. The mistake was my preparation. I'd assumed the ceiling structure in a 1930s building would be solid up there (which, honestly, was not a safe assumption in an old building with plaster ceilings). The Eris is a substantial chandelier—it requires proper mounting support, and the existing junction box wasn't rated for the weight. I caught it before mounting, thankfully, but "caught it" meant a call to the client, a trip to source an adjustable ceiling canopy and support bracket, and a full day's delay on a tight schedule.
Here's the part I didn't expect: the client's perception of the project changed. Not massively, but enough. The delay made them nervous. They started double-checking our line items. Even after the Eris went up and looked absolutely perfect—and it did—there was a lingering hesitation in the room. The chandelier itself had done its job. My process had created a crack in trust that the product couldn't fully cover.
That's what I mean when I say quality is bigger than the product. The product quality was there. The process quality wasn't, and the client felt it.
The "Simple" Fixture Replacement That Taught Me the Hardest Lesson
If you search "how to replace a light fixture" online, you'll get 40 million results, and most of them will tell you it's a simple job. Turn off the breaker, remove the old fixture, connect the wires, mount the new one. Done. I believed that for years—until it bit me.
We'd won a small contract to replace outdated fixtures in a community center's corridor. Eighteen fixtures, one-for-one, straightforward swap. I sent a junior guy with instructions and told him to call if anything looked off. He called. The new fixtures were a different shape, which meant the mounting bar didn't align with the existing junction boxes. And because the wiring in the building was older, the ground wire situation was messier than expected. What should have been a two-hour job stretched into a full day. The building manager walked past twice, saw the ladders and the puzzled faces, and quietly asked if "everything was under control."
The embarrassing part: I'd been doing this for eight years at that point. I should have known the checklist by heart. Check the mounting bar pattern before you go. Check the junction box rating. Check whether the old wiring has a ground. Check the weight of the new fixture versus the box. "How to replace a light fixture" isn't academic—it's a series of small decisions, and each one either builds the client's confidence or chips away at it.
Per the National Electrical Code (NEC, Article 314), junction boxes must be accessible and have adequate volume for the wiring—and that's before accounting for a heavy fixture's support requirements. Code isn't just bureaucracy. It's the accumulated lessons of everyone who wired things wrong before you. When you skip the checks, you're basically telling the client that their safety isn't worth 15 extra minutes of verification.
But What If the Client Doesn't Care About the Details?
"My clients don't know what a transformer is. They don't care whether the fixture has a UL listing. They just want it to look good and work."
I get the pushback here, honestly. A lot of contractors will say that. To be fair, it's partially true. The client won't compliment you on your transformer selection. They won't mention the NEC compliance of your junction boxes at the final walkthrough. But they will notice the darkness when a transformer fails. They will notice the delay when the mounting hardware doesn't match. They will notice the buzzing sound coming from a cheap power supply. And when they notice those things, they will remember the name on the invoice more vividly than any compliment they paid you in the first month.
That's not just my opinion—it's also just how memory works. Experienced poor outcomes are stickier than pleasant first impressions. I've had clients who loved a completed project for two years, then something failed, and suddenly the entire relationship was reframed as "the firm that installed something that broke." It's not fair, but it's real. In my experience, the math on quality is simple: the upfront cost of doing it right is almost always less than the cost of doing it twice.
The Checklist That Saves Me Now
After the transformer failure, I started documenting every mistake in a shared document. After the library chandelier, I added a pre-install verification step. After the community center, I added a "simple replacement" checklist that's absurdly detailed on purpose. It's caught 47 potential errors in the past 18 months—the most satisfying one was catching a mis-spec'd dimmable power supply on a $3,200 lighting package. The driver would have been incompatible with the controls we'd quoted, and we caught it before ordering. There's something genuinely satisfying about flipping the problem from "why did this fail?" to "we caught that before it shipped." That's the closest thing to job security I've found in this industry.
As of early 2025, the non-negotiables haven't changed much: verify fixture weight against the junction box rating, confirm transformer specs match the LED load, check photocell compatibility, verify dimming standards match the control system, and never swap a specified component without checking integration with the rest of the system. I also keep an eye on environmental claims now—if a client asks whether a fixture is "eco-friendly," I point to Energy Star certification rather than making loose verbal claims. Per the FTC Green Guides (ftc.gov), environmental claims need substantiation, and that matters for both compliance and credibility. Things evolve, of course. LED technology changes fast, and codes get revised. But the principle doesn't change: the quality of what you deliver is the quality of what you specify, verified, and installed—and it all lands on the client's perception of your name.
Part of me still wishes I could hand clients budget options and let them make the call. Another part knows that clients are hiring me partly because they don't want to make those calls. They want me to have already made them, correctly, a thousand times. That's the service we're actually selling. It has to show up in the products, the process, and the details—or it doesn't show up at all. Nobody's going to see the transformer. But everyone will see the darkness when it fails. And they'll associate it with you.
