Legrand Smart Lighting: A 6-Step Pre-Order Checklist From Someone Who's Paid for the Mistakes

Why This Checklist Exists

I've been in the electrical trade since 2017 — first as an apprentice, now as project lead for a small contractor that does lighting and low-voltage work. In that time I've made and documented enough specification mistakes to total roughly $4,800 in wasted budget. Not design theory. Reorders, restocking fees, service callbacks, and one very memorable weekend re-pairing a client's entire lighting system.

This checklist is for the people who spec and install Legrand smart lighting: electrical contractors, specifiers, and GCs working with real deadlines. It covers the legrand light switch family decision, zigbee keypad setup, stair lights, hub choices, and the eyeball light vs downlight question. Six steps, ten minutes of reading. If it saves you one reorder, it's paid for itself.

Step 1: Decide the Switch Family Before Anything Else

The legrand light switch family question looks cosmetic, but it's structural. Radiant and adorne are the two collections, and they are not interchangeable.

Radiant is the workhorse: toggle and rocker switches, dimmers, occupancy and vacancy sensors, smart switches, and the broadest SKU range. It's what I recommend for most commercial and multifamily work — clean, familiar, and flexible. adorne is the design-forward line: slim profiles, nearly flush to the wall, soft-touch operation. It looks more like millwork than electrical trim. Both are good products. The problem starts when you mix them, or when you pick a family first and discover later that the smart device you need doesn't exist in that line.

I made this mistake in 2018. I speced Radiant dimmers in the bedrooms and adorne switches in the hallway because the client loved the showroom display. Nobody verified that every smart SKU was available in both families — or that the trim sizes and box requirements matched. They didn't. Eight items had to be reordered, the job stalled for two weeks, and I ate $890 in restocking. That was a cheap lesson compared to what a full-house spec error could have cost.

Checklist item: confirm every smart SKU on your order — dimmer, switch, keypad, sensor — exists in the family you chose. Verify current part numbers on legrand.us before you send the PO. The catalog changes; last year's part number is not a reliable memory.

Step 2: Map the Zigbee Keypad to Human Behavior

The legrand zigbee keypad is a simple device: one, two, or four buttons, each triggering a scene. Simple hardware. The complexity is in the mapping. And I have the scar to prove it.

In 2020, I configured a four-button keypad at a client's entrance as “Scene 1”, “Scene 2”, “Scene 3”, and “All Off.” Technically correct. Completely useless. The nanny pressed buttons at random and asked me which one was “the TV one.” We ended up printing a label and taping it under the keypad. It stayed there for two years.

The fix is to design scenes around what the user actually does. Walk in the door at 7pm? That's “Evening,” not “Scene 2 with dimmers at 35%.” Leaving the house? “Away” is more useful than a list of zones you turned off. The scene names should make sense to someone who has seen the app for exactly thirty seconds.

Checklist item: write the button labels before you order the keypad. If you can't describe each button in one phrase a guest would understand, you're not ready to configure it. Also plan radio coverage — the keypad is a Zigbee end device, so it needs a hub or coordinator within range (step five). That constraint affects placement, and placement happens during rough-in, not after drywall.

Step 3: Plan Stair Lights as a System

Legrand stair lights are small LED step lights. They look forgiving. They are not.

Three mistakes, two of which I made myself.

The first is spacing. In 2021 I counted steps instead of measuring each run, and a stairwell with a landing is not a single run. Uniform spacing on a non-uniform stair produces pools of light that look like a runway at night. Measure the actual runs, mark the mounting surfaces on the plan, and confirm the spacing against the trim's beam spread before you order.

The second is control. Stair lights tied to a dumb switch get left on all night, or they sit dark until someone stumbles. The installs that actually work tie stair lights into an occupancy sensor — the Legrand RRW600 occupancy/vacancy sensor is the one I use in Zigbee setups — or into a keypad scene. Lights come on when someone approaches, go off when nobody's there.

The third mistake is the expensive one: local code. In many jurisdictions, stairway lighting counts as egress and has specific requirements — possibly including no full dimming or no full automation. I assumed “stair lights are decorative” once. The inspector didn't. The redo came out of my pocket and pushed handover back a week. Check the local code, and if it's ambiguous, email the AHJ. That email takes five minutes.

Checklist item: count the runs, not the steps. Confirm the control path — sensor, scene, or both. And confirm egress requirements before you dim anything.

Step 4: Use the Eyeball Light vs Downlight Decision Correctly

In 2019 I wrote eyeball trims into nearly every spec. On paper they do everything a downlight does — plus they swivel. Why would you ever choose a fixed trim? Then I stood in a hallway full of eyeball trims that nobody had ever swiveled, collecting dust and catching light at odd angles. “On paper” is not a ceiling.

Here's the eyeball vs downlight breakdown I use now, room by room:

Eyeball light: the housing stays flush; the inner trim pivots so you can aim the beam. Use it when there's a surface that needs direct light — a kitchen island, a reading nook, a piece of art, a work counter. The cost is mechanical: a swivel joint on a ceiling is a thing that moves, accumulates dust, and can drift out of its aimed position over time.

Downlight: fixed beam, straight down, no moving parts. It does one job cleanly, it's usually cheaper per trim, and it disappears into the ceiling. For hallways, stairwells, closets, and general ambient coverage, it's the better answer.

The mistake is choosing one approach for the entire project. The correct sequence is to walk the plan zone by zone: does this location need a directional beam? Mark it as eyeball. If you're not actively aiming light at something, choose a downlight. And keep trims consistent within a visible ceiling plane — mixing eyeball and downlight trims in the same open ceiling usually looks accidental, not intentional.

One more thing on this: if you're dimming either, verify compatibility. Flicker at low levels is almost always a trim/driver/dimmer mismatch, not a flaw in the product line. Legrand's Halo trims, in general, are matched with the corresponding dimmers, but the moment you swap in another manufacturer's bulb or driver, you become the compatibility lab. Nobody budgets for that.

Checklist item: on the floor plan, mark every ceiling location that needs directional light. If it's not marked as an eyeball, it's a downlight.

Step 5: Choose the Zigbee Coordinator Path Early

Legrand smart lighting runs on Zigbee 3.0, and something has to be the coordinator. This decision affects how every device pairs, how keypad scenes are configured, and how firmware updates happen. And it's usually made last, after the walls are closed.

Three paths I've used:

Path one: Legrand's own hub. The supported route — plug it in, pair the devices, set the scenes, done. For the majority of my clients, this is the right call. Easy to maintain, no tinkering required.

Path two: Home Assistant with a USB Zigbee coordinator. More control, more flexibility, and more ownership. I've done this for homeowners who genuinely enjoy smart-home tech. It works well when someone is willing to maintain it.

Path three: OpenWRT with a Zigbee coordinator. Yes, the openwrt zigbee setup is real. If you run an OpenWRT router, you can attach a compatible USB Zigbee stick and run a controller like Home Assistant on the router itself. It's a legitimately elegant setup for a network-savvy user.

I got this wrong in 2022. I set up the OpenWRT path for a client because I thought it was a no-brainer — technically elegant, one less hub, more direct control. The client didn't want elegant. They wanted their switch to turn on the light at 6pm. When a router firmware update dropped the Zigbee adapter config, I got a 9pm call and spent a weekend rebuilding scenes from memory. My mistake wasn't the technology. It was choosing the technology based on who I am, not who maintains the system after handover.

And before anyone writes the comment: there are projects where a conventional wired relay panel is the right call. Steel-framed buildings, mission-critical circuits, clients who want zero app dependency. I spec that when it fits. But if you're going wireless on Legrand's Zigbee system, treat the coordinator as infrastructure — decide it at the design stage, not during commissioning. Another consideration: some clients care about keeping their system in an open ecosystem rather than relying on a single vendor app. That preference is valid, and it's a good reason to consider the Home Assistant/OpenWRT path. Just make sure the person who'll maintain it knows what they're signing up for.

Checklist item: write down who maintains the system after handover. If the answer is a normal homeowner, the decision is already made. The hub stays.

Step 6: Write Down the Scene Map

The most important document on a smart lighting project is not the one-line diagram. It's the scene map. I learned that in September 2024, when a client reset their smart-home account and every scene, keypad assignment, and sensor automation returned to factory default. I had to re-commission the whole house from memory. Two days. Not billable. A humbling experience.

The fix is a ten-minute deliverable:

• A keypad layout map — button, scene name, what it affects, in the client's own words.
• A scene schedule — which zones, which dim levels, which automations.
• The coordinator details — hub model, firmware version, login owner, spare parts list.

Print it, put it in the panel or client folder, keep a copy in your job file. In my company, this is now a required deliverable on every smart lighting install.

Checklist item: if the scene map doesn't exist in writing, the project isn't finished.

The Mistakes That Didn't Fit in the Steps

A few closing notes from the field.

Verify part numbers against the current catalog. Legrand revises SKUs, and a product you specified last year may be superseded. A trim mismatch on the job site is usually a version-control problem, not a shipping error.

Check stock before committing to a schedule. The supplier with the best price can't ship what they don't have. Ask about stock on every SKU, not just the headline items — and get the answer in writing.

And don't decide the easy things early and the hard things late. Every mistake in this article came from a decision made in the wrong order. The switch family, the keypad mapping, the stair-light spacing, the coordinator path, the trim choice — these are all design decisions. Make them before you write the PO, and the hardware does what good hardware does: it disappears into the background.

Then the only question anyone asks is “wait, who makes these switches?”

And you'll know the answer.