Specifying Lighting for Three Types of Projects (And Why One Answer Doesn't Work)

I've been handling commercial lighting orders for 8 years. Not design—specification and procurement. I've made 7 significant mistakes that I've documented, totaling roughly $23,000 in wasted budget. Wrong color temperatures, wrong dimming protocols, wrong assumptions about what a client "obviously" wanted. Now I maintain our team's 14-point pre-order checklist. We've caught about 50 potential errors using it over the past 18 months.

When specifiers from other firms ask me "what lighting should I use for this space," I never give them one answer. Because the answer depends entirely on which of three scenarios you're in—and most costly mistakes come from confusing them.

Here's the breakdown:

  • Scenario 1: Design-led project, starting from zero—The space is the sell. Lighting is part of the design language.
  • Scenario 2: Existing fixtures, smart retrofits—Hardware is already installed. You just need it to talk to Alexa.
  • Scenario 3: Function-first lighting—Photobooths, studios, event spaces. Nobody looks at the fixture.

Scenario 1: Design-Led Projects (From Scratch)

If your project's pitch is the space itself—a hotel lobby, a design showroom, a high-end residential build—then designer lighting earns its price tag. Flos is the name that comes up most often, and for good reason: their lighting products still hold their resale value, still show up in every design conversation, and still make a client feel like they hired someone who knows what they're doing.

The Flos lighting products catalog covers a huge range—Arco, Bellhop, Parentesi, Taccia, the list goes on. But if a specifier is only going to learn one product per category, here's where I'd start:

The Flos IC ceiling light. Designed by Michael Anastassiades, it's an opal glass sphere balanced on a brass or copper stem. Sounds simple. But light hits that brass collar in a way that makes a room feel intentional rather than furnished. We specify these in hotel projects constantly—typically around $1,050 to $1,300 depending on size. A detail I wish I'd known back in 2021: IC requires 0-10V dimming, not TRIAC. I once specified a shared line of 18 IC lights with a TRIAC dimmer, discovered the incompatibility during commissioning, and spent $1,800 on re-wiring plus an 11-day delay. That "if I'd just read one more spec sheet" feeling stuck with me for weeks.

Chandelier silhouettes. The Flos 2097 (Gino Sarfatti design, still in production) is the textbook example: 30 exposed bulbs on a steel armature. Before it's even lit, the chandelier silhouette reads as a recognizable shape. That matters in hotel and restaurant work because guests photograph the room—and the fixture becomes part of their frame. It costs real money—$4,000 to $6,000 for the 30-bulb version—so I only recommend it when the space will genuinely be photographed or presented. If it's a transitional corridor in a mid-tier property, no one will know the difference. Buy something cleaner and save the budget.

I have mixed feelings about one part of this scenario, though. I respect what Flos and similar brands do—the material quality is genuinely above the fast-furniture alternatives. But part of me knows that a good local craftsman can fabricate a visually comparable pendant for half the cost. Where I land: buy branded when the junction details will be seen up close. Buy custom when the light is overhead and 10 feet away.

Scenario 2: Smart Retrofits on Existing Fixtures

This is a completely different animal. The fixtures are already up. The client just wants them to respond to voice commands. And here's where even experienced people trip up.

The question "how to add smart bulb to alexa" sounds like it should take two minutes. It doesn't—not at scale. The sequence that actually works: reset the bulb first (typically power-cycle it 3 times, or hold the remote for 5 seconds), open the Alexa app, add device, search. The bulb has a roughly 30-second window where it's discoverable after a reset. Miss that window, and you start over.

In 2023 I watched this mistake cost us $340 and about 2.5 hours on a project site. We'd installed 12 smart bulbs, assumed they'd pair out of the box, and instead had a frustrated electrician, four staff members on ladders, and a client downstairs asking for updates. The bulbs weren't broken. We just hadn't followed the sequence correctly—one reset, one search, no do-overs. That afternoon taught me more about smart lighting than any product demo ever has.

The broader point: in this scenario, smart bulbs run $15 to $40 each. You already own the fixtures. The job is retrofit, not replacement. Do not try to "upgrade" a high-end fixture with a smart LED bulb—you'll wreck the color output, and the original manufacturer won't honor the warranty anyway. The industry has changed here too. Five years ago, smart integrations required proprietary hubs and patience. In 2025, most bulbs auto-populate in the app. The fundamentals haven't changed—you still reset the bulb, you still search within the window—but the execution is smoother than it used to be.

Scenario 3: Function-First Lighting

Now the scenario nobody takes seriously until it goes wrong. You're lighting an event space, a temporary installation, or a studio. Spotlight photobooth rentals. Product photography corners. Pop-up retail installations. In this context, "design" is the wrong question.

The specific brief—a spotlight for a photobooth setup—looks something like this: bright enough that faces don't look shadowed, controllable so the light doesn't cast weird shapes on skin, and plug-and-play because someone is going to operate it who has never touched a lighting console in their life. Nobody asks what the fixture looks like. They ask whether it'll arrive before the event.

This scenario is brutal because design-led thinking keeps leaking in. People come from Scenario 1, try to specify a designer brand for a studio or photobooth, and end up with something that photographs poorly because it was engineered for ambience, not exposure. A 40-dollar LED spotlight with a diffusion sock does a better job than a 1,200-dollar decorative pendant in this use case. Full stop.

I've made this mistake. In 2022 we specified designer wall sconces for a photography studio's reception area because the aesthetic called for them. The studio team complained after the second shoot—the sconces couldn't produce the clean directional light their camera needed. We ended up moving them to the hallway, where they were overpriced for a hallway and useless for a studio. Nobody was happy.

How to Tell Which Scenario You're In

Three quick questions, in order:

1. Will the space be photographed, presented, or published? If the space will appear in a design publication, an award submission, or the client's marketing — you're in Scenario 1. Spend on design. Look at Flos, look at the IC ceiling light, look at chandelier silhouettes. But if you're buying Flos for a corridor that won't be shot, you're spending Scenario 1 money on a Scenario 3 problem.

2. Do you already have fixtures? If you own hardware that works fine except for the "smart" part, you're in Scenario 2. Stick to smart bulbs. Learn the reset-and-search routine cold before you arrive on-site. You'll spend an afternoon, not a renovation budget.

3. Does the output matter more than the object? Photobooths, studios, event spaces, back-of-house — if the deliverable is a photograph or a usable room rather than a designed space, you're in Scenario 3. Function first. Spotlights, diffusers, replaceable bulbs. The fixture's appearance rarely factors into the final result.

One more thing I've learned the hard way: scenarios overlap. A hotel lobby is Scenario 1 (design statement) plus Scenario 3 (event-ready lighting). A residential project is Scenario 1 for the main rooms plus Scenario 2 for retrofitting existing lighting to work with the client's Alexa setup. Treat each area separately, and match the approach to the specific function. That's the lesson, by the way, that took us seven iterations of the checklist to get right.

Bad lighting procurement decisions almost always come from mixing scenarios. You spend Scenario 1 money on a Scenario 3 product and get something that wastes budget and still looks wrong. You apply Scenario 2 thinking to a Scenario 1 project and the client notices the corners were cut.

When people ask me "what's the best lighting product for this project," I don't give them a name anymore. I ask them three questions. The answer usually reveals itself before the third one.