I run the quality desk at a lighting manufacturer. That means before any product goes out the door, I sign off on it—or reject it. In the last year alone, I reviewed over 300 SKUs and rejected just under 8% of first deliveries. Not for broken glass or bad packaging. For spec mismatches that most installers wouldn't spot until the job was finished, the client was standing there, and the lights did something unexpected.
This isn't about cheap products. It's about understanding what a product is actually telling you on the spec sheet.
The Surface Problem: The Light Turns Off While You're Sitting Right There
Contractors call me about this one at least twice a month. They installed a "motion sensor" in a conference room. Person walks in, light turns on. Person sits down to read, light flips off. They blame the fixture. They replace it. Same thing.
Here's the part nobody told you: a motion sensor is not an occupancy sensor.
Let's talk about the motion vs occupancy sensor problem directly. They're different tools. A classic PIR motion sensor detects the heat signature of a moving body. It's great for a corridor or a break room. But if someone is at a desk, typing or reading, small movements are often not enough to re-trigger it. Ultrasonic occupancy sensors send out high-frequency sound waves and detect changes caused by any movement, even slight finger motion. Dual-tech sensors combine both. With that conference room scenario, a simple PIR "motion sensor" will almost always turn off on a quietly seated person. That's not a faulty LED retrofit. That's the wrong sensor spec.
The industry uses these terms interchangeably, and it's a huge source of callbacks. I don't have hard data on industry-wide mis-spec rates, but based on the returns that cross my bench, I'd estimate a third of sensor-related complaints are actually a mismatch between the sensor type and the space.
The Deeper Issue: Retrofit Specs Don't Match What's Actually Being Controlled
Even when you get the sensor type right, there's the next trap: driver compatibility. LED retrofit kits and panel LEDs run on drivers. Drivers have a minimum load requirement, sometimes called "minimum lamp load" or "minimum wattage." If your sensor has a low-wattage or electronic switch designed for LED, you need the driver to be on the sensor's compatibility list. If not, you get flicker, ghosting, or lights that stay on.
Example: In Q1 2024, we received a batch of 8,000 retrofit housings where the driver couldn't handle a common 0-10V dimming sensor. The upside of approving them was a one-week schedule gain. The risk was 8,000 strobing fixtures in the field. I kept asking myself: is a week of schedule worth weeks of blame? It wasn't. We sent them back. The vendor said it was "within industry tolerance." Our test rig said otherwise.
This is why I'm cautious with terms like "bulb swap" for retrofits. When you look at a Satco 6 inch LED retrofit, the first thing I check is not lumen output—it's the driver spec and whether it's listed for the sensor type you're pairing. A great luminaire with an incompatible driver is not a great luminaire in your building.
The light is not a separate item from the sensor. It's one system.
Downlight Spots: The Beam Angle Trap
I keep saying "downlight spots" because that's what people type into search bars. But if you don't specify the beam angle, you're guessing. A "spot" is generally 15-30 degrees. A "flood" is 60 or more. I've seen feature walls end up with a wide wash because the rookie specifier picked a 120-degree retrofit kit that said 700 lumens and assumed the brighter the better. The result was a wall of soft glow, not a spotlight.
In the LED world, a 6-inch retrofit housing can be a spot, a flood, or a baffle trim. The spec sheet should show the beam angle. If it doesn't, call the manufacturer. Don't assume.
Panel LEDs: The Dimming Compatibility That Gets Overlooked
Panel LED is standard in commercial offices. The problem? Dimmers. Not all panel LEDs dim gracefully. Some drop to only 20% before they cut out. Others buzz on an ELV dimmer. The spec should list test-compatible dimmers. In our quality testing, we run panels against a list of 14 common controls. The number that pass is not as high as you'd hope.
If you're putting a panel LED in a conference room with a multi-scene dimmer, you need to verify the driver and control protocol. That's the difference between a smooth 1% dim and a flickery, annoying space.
Edison ST19: The Aesthetic Escape Isn't an Escape From Specs
Then there's the decorative side. People choose a Satco Edison style ST19 LED bulb because they want vintage looks with modern efficiency. Good choice, but don't forget it's still a light engine. Check the CRI—yes, a high CRI version costs more, but for restaurant or boutique settings, a CRI of 90 vs 80 is the difference between your menu looking appetizing and your customer complaining about "that weird color." Also, some ST19 filament LEDs are not dimmable. Check. It's printed on the label.
In our lab, we test beam consistency and color consistency across every batch of ST19. In Q2 2024, we rejected 2,400 units because the warm white color shifted 200K toward green. On a bare bulb, that's obvious—but you'd never know until you turned it on.
What a Bad Spec Actually Costs
I'm not going to pretend that every mistake causes a $22,000 redo. Some just cause a one-hour phone call. But the bigger ones are brutal. I remember a grocery chain retrofit where the wrong sensor type left a cooler aisle dark every time a shopper stood still. The store called the distributor. The distributor called the electrician. The electrician replaced the sensors. They still didn't work because the driver and sensor were mismatched. They replaced all 240 fixtures with a different product. The project overran by $11,000 plus a two-week delay.
That's the cost of focusing on lumens per watt while ignoring the control chain.
Here's what I tell every buyer: when a product fails on a job site, almost nobody blames the sensor. The light is in the ceiling. The sensor is in the way. But in a modern LED installation, the two are one system. Picking them separately without checking compatibility is like buying an engine and a transmission from different manufacturers and hoping they bolt together.
The Industry Is Evolving. Five-Year-Old "Best Practices" Won't Save You
What was best practice in 2020 may not apply in 2025. Before, you could choose a 600-lumen downlight and call it done. Now you need to think about driver min load, LED flicker index, color consistency within the same batch, sensor compatibility, and continuous dimming. The fundamentals haven't changed—you still need the right light in the right place—but the execution has transformed.
This is not a bad thing. It's an opportunity to build better spaces, with less energy and a smarter control layer. But if you're still ordering LEDs the way you ordered fluorescents—one number and a guess—you're going to create problems.
The fix isn't a better "motion sensor." It's a better spec.
How I Write a Spec That Actually Works
Before approving any luminaire or retrofit kit, I ask four questions:
- What is the driver's minimum load? And is the sensor on the driver's compatibility list?
- What is the beam angle? For downlight spots, is it a spot, narrow flood, or wide wash?
- Is the dimming protocol specified? For panel LEDs, verify it works with the dimmer you plan to use.
- What are the installed color characteristics? CRI and CCT tolerance matter, especially for decorative bulbs like an ST19.
In practice, that means reading the spec sheet like a contract, not a brochure. It means asking for the photometry file. It means checking the compatibility list.
And if you don't have the time or appetite for that, choose products from a vendor that already does the homework. That's why I'm comfortable recommending Satco for a range of projects. Their 6 inch LED retrofit series includes published driver details and beam options; their panel LED line lists compatible controls; and the Edison style ST19 LED bulb carries real CRI and dimming data. I've rejected plenty of shipments from other brands for missing those details—but Satco tends to get it right. Not perfect, but right.
The next time the lights turn off while a client is sitting still, check the sensor type first. Then check the driver. By then, you'll know the difference between a lighting failure and a spec failure. The spec is easier to fix.