Most smart lighting failures are not product failures. They're planning failures wearing a product's name tag. After eight years of commercial lighting retrofits and about $7,000 of my own avoidable mistakes, I'm confident in that statement.

I'm a lighting contractor who has handled retrofit projects for small hotels, offices, and retail chains across the Midwest since 2017. In that time, I've personally made—and documented—18 significant mistakes. Total wasted budget: roughly $7,000, plus two client relationships I had to rebuild. The frustrating part isn't that those mistakes were technically complex. It's that nearly all of them followed the same pattern: I treated smart lighting as a product to buy, not a system to plan.

This isn't a confession. It's the checklist I wish someone had handed me before the first crate arrived on site. If you're an electrician, a facilities manager, or a contractor planning a retrofit, I'd rather you learn from my invoice than your own.

How to Replace an LED Driver: Start With the Cavity, Not the Catalog

LED drivers are the least glamorous component in any LED installation, and that's exactly why they cause the most callbacks. When a client asks me how to replace an LED driver, my first question is no longer “What wattage?” It's “What's the cavity size, connector type, and dimming protocol?” I did not ask that question in September 2022.

The job was 60 recessed fixtures in a hotel corridor. The old drivers were humming, the hotel wanted them swapped quickly, and I was determined to move fast. I matched wattage, matched voltage, and confirmed dimming compatibility. Then I ordered 60 replacement drivers.

They didn't fit. Not electrically—physically. The new drivers were about 10 millimeters too long for the fixture cavity. All 60 units had to go back. Restocking fee: $900. Project delay: two weeks. The hotel manager's expression: memorable.

Everyone had warned me about this. Kevin, our senior installer, literally has a note above his bench that says “measure the cavity before you order.” I ignored it because I thought this particular job would be simple. That is one of the few times I'm glad to admit I was wrong.

(Note to self: “driver cavity dimension” is now at the top of every pre-order checklist. So is “connector style.” Kevin's note stays on the wall.)

I'm not an electrical engineer, so I can't speak to the finer points of driver topology or waveform design. What I can tell you from a contractor's perspective is that most LED driver replacements fail for boring reasons: wrong form factor, wrong connector, wrong dimmer compatibility. None of those show up on a voltage spec sheet. When you search for how to replace an LED driver, ignore any explanation that ends at “match the wattage.” Measure the physical space first. Then measure it again.

What a Feit Electric Smart LED Strip Light Extension Taught Me About Voltage Drop

A few months later I made the same kind of mistake with prettier packaging. A retail client wanted under-shelf lighting on a 40-foot display wall. We had already installed Feit Electric smart LED strip lights on the first eight feet, and they were excellent—consistent color, responsive control, no flicker. Phase two meant adding a Feit Electric smart LED strip light extension along the rest of the wall.

I calculated the total length, counted the connectors, and assumed the extension would behave exactly like the original segment. It did not. The farther the strip ran, the weaker the color became. When the client requested a color change, the extended section lagged, flickered, then settled on a visibly different shade. On a display wall painted in brand colors, that's not a small detail.

The problem wasn't the strip. It was voltage drop and signal integrity over a long run. I should have planned a separate power feed or a shorter segment. Instead, we re-ran the whole display and absorbed the labor cost. The store manager was patient. I was not.

That's when the pattern started to bother me. I wasn't buying bad products. I was buying products without planning the environment around them.

Feit Electric WiFi Bulbs Worked. The Building Didn't Agree.

Next example: a logistics office. Feit Electric WiFi bulbs in the main workspace were flawless—instant response, easy scheduling, no hub required. Then the annex began dropping devices every evening. My first instinct was to blame the bulbs. It was not the bulbs.

The annex sat behind a concrete block wall about 30 meters from the router. A Feit Electric WiFi bulb is still a WiFi device; it needs a usable signal. I had chosen a technology that was easy to specify, not one that was right for that building. The perimeter would have been better served with Zigbee outdoor sensors and mesh devices that relay through one another, instead of every device struggling to reach a single access point.

That mistake cost us a full day of diagnosis, a second visit, and a network upgrade that belonged in the original quote. I had to explain to a client why their “smart” building went dark every night, and the honest answer was that I skipped the RF planning step. The products never failed. I did.

The GU10 Spotlight Upgrade That Wasn't an Upgrade

Not every expensive mistake involves a driver or a wireless protocol. Sometimes it's just a GU10 spotlight and a contractor's ego.

A client asked for dimmable GU10 spotlights above the front desk of a boutique hotel. They wanted a soft, wide wash of light across the check-in area. I decided to impress them by ordering a narrower beam angle—24 degrees instead of 40 degrees. It sounded more precise. It looked like an interrogation room.

All 36 fixtures came down and were replaced. Product cost plus labor: roughly $600. The client asked, “Why didn't you just give us what we asked for?” That question still stings because it's the right question.

Now, every GU10 spotlight order in our office includes beam angle, color temperature, and dimmer compatibility on the spec sheet—even when the client didn't specify them. If the client did specify a beam angle, I follow it. “Fancy” is not a lighting specification. Many LED packages also carry a Lighting Facts label that lists lumens, wattage, and CRI, which is handy for comparing products. It just won't tell you how the light will feel in the space; that part is still your job.

So Is Smart Lighting Just Overhyped?

At this point, you might expect me to conclude that smart lighting isn't ready for prime time. I hear that from fellow contractors all the time. I used to say it myself.

Then I looked back at the actual failures. A driver dimension I didn't measure. A voltage drop I didn't calculate. A WiFi range study I never did. A beam angle I overruled. In none of those cases did the equipment malfunction. The Feit Electric smart LED strip light extension performed within its design limits. The WiFi bulbs responded perfectly once the signal was strong. The GU10 spotlights rendered beautifully—at the wrong angle, creating the wrong effect.

The common denominator was my process, or more precisely, the lack of one. Saying “smart lighting isn't ready” when I did not plan the system around it is like blaming a car for running out of fuel because I never checked the gauge. The products did what the spec sheet said. The system around them was never designed.

That does not mean every room needs a network. A storage corridor or a mechanical room with one switch and a standard LED fixture does not need WiFi, Zigbee, or a smartphone app. I still install plenty of conventional controls, and I'm happy to do it. Digital efficiency is about choosing the right tool for the job, not decorating every ceiling with a connected thing.

But when the goal is energy reporting, time-of-day scheduling, scene control, or integration with other building systems, smart lighting delivers real benefits. One hospitality client cut after-hours lighting energy by more than a third once we stopped relying on staff to remember switches and started scheduling scenes. That saving did not come from the bulb. It came from the system working as a system.

The Real Efficiency Gain Was Boring

After the driver incident in September 2022, I built a pre-quote checklist. It's not clever. It asks boring questions: What is the driver cavity size? What is the actual cable run length? Do we have RF coverage where the wireless products are going? Did the client specify a beam angle, and are we planning to change it without telling them?

The results showed up in 18 months: zero driver-related callbacks on 47 completed installations, change orders down by roughly a third, and net margin up by about 8 percent. We achieved more efficiency by slowing down the quoting process than by speeding up the installs.

I do not mention those numbers to brag. I mention them because they contradict the usual story. The largest efficiency gains in lighting are not hidden in a bulb's chipset. They live in the decisions made before anyone opens a box.

A quick caveat so you know the scope of my experience: I've worked on about 150 mid-size commercial jobs in the Midwest—mostly hotels, offices, and retail fit-outs. If you're handling industrial-scale lighting or high-end architectural projects, your requirements will differ, and you should absolutely involve an engineer who works at that level. But I suspect this principle transfers: understand the system before you buy the parts.

Smart lighting is not a shopping list. It's a system. Treat it like a system and you'll get the reliability and efficiency it promises. Treat it like a product and you'll pay for the lesson the same way I did, with a restocking fee, a rework day, and a client asking why you didn't plan better.

Measure the cavity. Map the network. Respect the client's beam angle. Plan first, product second.