How to Start a Permanent Lighting Business

I launched 5280 Lights in the Denver metro in spring 2026. By October I'd collected $68,041.50 from 19 customers, held 15 five-star Google reviews, and built every piece of the system that feeds the business — website, search rankings, Google Business Profile, CRM, follow-up automation, aerial quoting, ads.

None of that is the hard part of this business. Hanging track is the hard part to learn. Everything else is the hard part to survive, and it's where most new installers stall out.

This guide is the second half — the part the supplier blogs skip because they're selling you a kit, and the part the franchises charge you for. I'm writing down what actually happened, including the months I wasted and the two tracking mistakes that cost me real data I can't get back.

What this business actually is

Permanent outdoor lighting is low-voltage LED fixtures seated in an aluminum channel mounted under the eave and along the rakeline. It stays up year-round. The homeowner runs it from an app — warm white for everyday use, color for holidays, team colors on game day.

The distinction that matters commercially: this is not a Christmas light business. Seasonal Christmas light installation is a ten-week sprint with a hang, a takedown, and storage. Permanent lighting is a single install, a bigger ticket, no takedown, and no storage liability. It also sells in April, which seasonal does not.

If you're coming from Christmas lights, gutters, siding, or exterior cleaning, you already have the ladder skills and the comfort working at height. That transfers. What doesn't transfer is the sales cycle: a $2,500–$5,000 ticket involves more hesitation, more spousal discussion, and more follow-up than a $400 seasonal hang. Plan your follow-up system accordingly — more on that below.

The part nobody warns you about

Here's the honest sequence of what sat between me and my first paying install. None of it is lighting work:

• Figuring out what to charge — per foot, per corner, per story — without guessing
• Getting a Google Business Profile verified, which took far longer than I budgeted
• Building a website that shows up for people searching in my city, not just a brochure
• Setting up financing so a $4,000 quote isn't an automatic no
• Learning how HOAs approve or kill a job before quoting it
• Sales tax, licensing, entity setup, and the EIN mistake I'll describe below
• Finding suppliers and track sources that don't disappear mid-season
• Running ads that generate calls instead of burning $200 a day
• Following up on quotes without relying on memory at 9pm

That list is the five months. If you're evaluating whether to start, budget for it honestly rather than assuming you'll be installing in three weeks.

Choosing your system and supplier

You're picking two things that are somewhat independent: the light and the channel it sits in.

What I run

5280 Lights offers two complete systems:

Govee Pro (H706) or Govee Prism (H707) paired with SnapTrack commercial aluminum channel. Govee has genuine consumer brand recognition — which, as I'll explain in the marketing section, turns out to be the single most valuable thing about it. SnapTrack is a commercial-grade extruded channel that holds up and can be color-matched.

Strata (shop.stratalights.com), which supplies both lights and channel as one system. I brand this offering "5280 Signature." Customer response has been strong.

The compatibility trap

Govee Prism is the H707 series. It is a different animal from the Pro (H706) and the Original (H705). Connectors and extensions do not interchange between them. Worse, most third-party "Govee compatible" parts on the market exclude Prism entirely — they're built for the Pro.

Verify H707-specific compatibility on every third-party part before you order. I've seen this bite people who assumed "Govee is Govee."

Prism parts run roughly $0.25–$0.45 per foot more than Pro. That's real but minor — it is not a reason to reprice a job.

Color matching

This is a legitimate upsell and a real differentiator in neighborhoods with strict trim standards. SnapTrack with Govee supports custom RAL color-matching across the full RAL Classic range — 200-plus codes, matched to the home's existing trim.

Strata handles this differently: a 14-color standard RAL range in stock at 3.3 ft, plus a 19-color made-to-order trim-match palette at 6.6 ft with a one-to-two-week lead time. Different structure, so quote the lead time accordingly.

One caution from experience: if a supplier doesn't publish physical swatches for its trim-match colors, buy the swatch kit before you promise a match to a customer. Screen hex values are approximations and a mismatched channel on a dark trim is a callback you'll eat.

A note on shipping

Track shipping has been immaterial for me — roughly $40 a shipment, which works out to about twelve cents a foot. Don't let anyone sell you on consolidating orders as a cost-saving strategy. Order what the job needs when the job is sold. Carrying inventory to save pennies a foot ties up cash you need elsewhere in year one.

Licensing, entity, and the EIN mistake

I am not an attorney or an accountant, and licensing rules vary sharply by state and often by municipality. Verify everything below in your own jurisdiction before relying on it.

The low-voltage question

Permanent lighting systems typically run at 48V. In Colorado, the low-voltage side — Class 2/3 — is exempt from state electrical licensing under C.R.S. §12-115-120(3). That exemption is why a non-electrician can run this business here.

The boundary is the plug. Any 120V line-voltage connection requires a licensed electrician. If a job needs a new exterior outlet or any hardwired connection, that portion goes to a licensed electrician, full stop. Build that into your quote rather than discovering it on install day.

Other states draw this line differently. Some require a low-voltage license, some require a general contractor license above a dollar threshold, some require nothing. Find your state's answer before you quote your first job, not after.

Entity and EIN — learn from my mistake

Form the LLC properly and get an EIN for the LLC itself.

Here's what I did wrong. I had an EIN already — but it had been issued when I set up a DBA as a sole proprietor, years before the LLC existed. I assumed an EIN was an EIN. It isn't. That EIN was tied to me personally, not to the company.

The problem surfaced in an unexpected place: SMS compliance. When I registered for A2P 10DLC messaging, the brand verified under my personal legal name instead of the company name, because that's what the EIN said. The compliance scanner then failed me for not displaying the business name in consent text — a failure I couldn't fix by editing the website, because the root cause was in the EIN. Three rejected campaign filings later, I got an EIN for the LLC and refiled.

The same gap affects banking, contracts, and liability — not just texting. Raise it with an accountant early. It took me twenty minutes at IRS.gov to fix once I understood it, and weeks of confusion before that.

Insurance

Carry general liability before your first install. You're working at height on other people's homes and drilling into their fascia. This is not the place to be clever.

How to price the work

The structure I use

$18 per linear foot, single story. $21 per linear foot, two-story or complex.

That's the whole structure. It's deliberately simple, because complexity in pricing creates hesitation in buyers and arithmetic errors in you.

What pushes a job into the higher tier: - Two or more stories - Steep pitch requiring different ladder staging - Multiple gables, dormers, or stacked rooflines - Difficult access — landscaping, slopes, hardscape you can't set a ladder on - Unusual power routing

The rule I'd give any new installer

Never quote a per-foot or total price for a specific home before you've pulled measurements and assessed install complexity.

Ballpark ranges in conversation are fine and necessary — people need to know whether they're in a $2,000 or $5,000 conversation before they'll engage. But a specific number on a specific house, given before you've measured, is a number you'll either eat or walk back. Both are bad.

This is the single most common pricing mistake I see new installers make, usually because they're afraid of losing the lead by not answering immediately. You will lose more money to bad quotes than to slow ones.

Quoting without driving to the house

This is where most new installers burn their first month, and it's the most immediately fixable problem in the whole business.

Driving to every prospect's house to measure a roofline does not scale. At a 33% close rate, two out of three of those drives are unpaid.

Aerial measurement

I use two vendors for different situations:

FirstMate (app.1m8.ai) — delivers a zip containing structured 3D roof data. Expedited reports land in roughly three hours. This is what I standardize on, because the structured data file can be parsed programmatically.

InstantRoofer (instantroofer.com) — around $10, PDF output only. Cheap and fast, but a PDF is something a human has to read; it can't be fed into a tool.

If you intend to build or use any automated takeoff tooling, standardize on the vendor that gives you structured data. I built an elevation takeoff tool that parses FirstMate's file and outputs roofline footage broken out by elevation — eaves and rakes, front, left, right, rear. It turns a measurement report into a quote in under a minute.

Quote in three scopes, not one

This changed my close rate more than any other single thing.

Most homeowners want permanent lighting and are shocked by the full-perimeter price. If the only number you give them is the big one, you get a no. Give them a ladder instead:

On a recent job in Morrison, the structured options were:

• Front only — 74 ft, $1,480
• Front and back with a side jumper — 125 ft, $2,500
• Full perimeter — 237 ft, $4,720

Three real choices, same house. The middle option exists specifically to be chosen. A customer who would have said no to $4,720 says yes to $2,500, and a meaningful share of them come back for the rest the following year.

Verify your own math

A caution learned the embarrassing way. My takeoff tool had a geometry bug that mis-assigned entry roofs and upper gables. On one house, the front-only measurement read 47 ft when the true figure was 74 ft. The customer caught it before my software did.

Whatever tool you use — mine, a vendor's, or a tape measure — sanity-check the output against the photo before it goes in a quote. A quote that's low by 36% is a job you install at a loss.

Financing

A meaningful fraction of your market wants this and cannot write a $4,000 check this month.

Without a financing option, those leads simply die. They don't say no — they say "let me think about it," and you never hear back. With one, a $4,000 quote becomes a monthly number the homeowner compares against their streaming bill.

Set this up before you start quoting, not after you've lost a dozen deals to silence. I use Acorn. Whatever you choose, have the application link ready to paste into a follow-up text the moment price is the objection.

HOAs

Learn how HOAs in your market handle exterior modifications before you quote in those neighborhoods.

Some require architectural review for anything attached to the exterior. Some have specific rules about color and operating hours. Some don't care at all. The failure mode is quoting a job, getting a deposit, and then having the homeowner denied — which is a refund, a wasted install slot, and a customer who tells their neighbors.

This is also a sales advantage once you know it. Being able to say "your HOA requires ARC approval, here's the form, and here's a color that's been approved on your street" is the kind of thing that closes jobs and that your competitors can't say.

The lead machine

This is the half of the business I spent five months building, and it's the half that determines whether you have a business or an expensive hobby.

Google Business Profile — start this first

Your GBP will drive more local buying intent than your website for the first year. Local buyers choose from the map pack.

Start verification immediately. It gates everything and it takes longer than you think. I began verification work in late April and wasn't verified until July 10 — about eleven weeks. I did not plan for that, and it meant my first two months of marketing had no local pack presence at all.

Nothing about that delay was unusual. Budget for it.

Reviews are the highest-value thing you can chase

My first review landed July 10. By October 1 I had 15, all five-star.

Those reviews do more for local ranking and for close rate than any amount of website optimization. Build a review request into your post-install process on day one — automated, not remembered. The request should go out while the customer is still looking at their lit-up house.

What actually worked in SEO — and what didn't

This is the section I'd most want a new installer to read, because the conventional advice is wrong for year one.

I could not outrank the established 500-review companies on the obvious local head terms. Not close. Five months in, "permanent christmas lights denver" sat around position 21, "lighting installation denver" around 19, "trimlight denver" around 28. My own city pages were in the twenties. You will not win those terms in year one against companies with a decade of reviews and backlinks. Anyone who tells you otherwise is selling.

What I won instead was page one for brand-comparison and brand-installer searches. Queries like "govee prism vs pro 2," "everlights vs govee," "govee light installers," "govee installation near me," "how much does it cost to install govee lights" — I rank in the top few positions for those.

That matters more than it sounds. Those are the searches people run immediately before deciding. Someone comparing Govee Prism against Pro is a buyer, not a browser. A single comparison page drives about half of all my organic clicks.

The strategy generalizes: in a crowded local market, don't fight for the head term. Take the searches that happen one step before the purchase decision — brand comparisons, cost questions, "X vs Y," "who installs X near me." They're less contested, they convert better, and they're winnable by a new site.

My site-wide average position moved from roughly 26 to 13 over five months, with 308 of 336 tracked queries moving up. Organic clicks went from about 41 a month over the summer to 281 in September. Some of that September number is seasonal — people shop for lighting in the fall. The ranking improvement isn't seasonal, and that's the number I'd watch if I were you.

AI search is already real

Since mid-August, when I started capturing how people found me, about one in ten of my leads came through ChatGPT or Gemini — some confirmed by referrer data, some self-reported. One of them is an installed, paid customer.

I did nothing specific to court that. My best guess is that the same comparison and buyer-guide pages that rank in Google are the pages AI assistants cite, because they directly answer the comparison questions people ask. I can't prove the causal link yet, but the practical advice is the same either way: write the comparison content.

Paid ads bridge the gap

SEO takes months. Ads work the week you turn them on, which is why they belong in your first ninety days even if you intend to be an organic business long-term.

Two warnings from my own account:

Set up conversion tracking correctly before you spend a dollar, and verify it fires. Mine was broken for months. My website pushed one event name and my tag manager was listening for a different one that didn't exist. Every lead I generated was invisible to Google, which meant the algorithm was optimizing against no signal at all.

Create a phone call conversion action. I didn't have one. In a business where a large share of buyers call rather than fill in a form, that omission made my reported cost-per-lead dramatically worse than reality — I was measuring a fraction of my actual leads and dividing full spend by it.

Watch your search terms report weekly and add negatives. In one five-month window I put roughly $249 into eleven search terms that produced nothing — mostly competitor brand names and off-season queries.

Capture lead source on every single form, from day one

My largest regret, and the cheapest thing on this list to get right.

I didn't put a "How did you hear about us?" field on every form until August. The result: of my first 19 paying customers, I can only attribute 9 of them to a source. Ten customers, representing the majority of my revenue, are permanently unattributed. I will never know what worked.

Make it required. Include an "AI assistant" option — you'll want it sooner than you think.

Follow-up has to be systematic

A $3,000 quote does not close on the first touch. It closes on the fourth, sometimes three weeks later, usually after the homeowner has talked to their spouse and gotten one other quote.

If your follow-up depends on you remembering at 9pm, you will lose deals you already paid to acquire. Put quotes into a CRM with a stage, and build an automated sequence behind each stage. Mine runs a series of emails and texts over roughly two weeks after a quote goes out, with real install photos rather than plain text.

15 of the 45 leads I got between May and August became paying customers — a 33% close rate. That number is a direct function of follow-up, not of sales talent.

What the first five months actually look like

Real dates from my own build, so you can calibrate:

Two things stand out when I look at that timeline.

First, revenue started before the marketing worked. My first payment landed in late June, well before rankings or GBP verification. Early jobs come from your network, from neighbors who see the first install, and from ads. Don't wait for SEO to start selling.

Second, the infrastructure mistakes cluster at the beginning and cost you for months. Conversion tracking broken until September. Lead source uncaptured until August. Those aren't glamorous problems, and fixing them doesn't feel like progress, but they're the difference between a business you can steer and one you're guessing at.

Frequently asked questions

Do I need an electrician's license? For the low-voltage portion, usually not — in Colorado, 48V Class 2/3 work is exempt under C.R.S. §12-115-120(3). Any 120V line-voltage connection does require a licensed electrician. Rules vary significantly by state; verify yours before quoting.

How much does it cost to start? The real costs are insurance, your first material order, a measurement vendor, and marketing. Material is ordered per job once sold, so inventory isn't the constraint. The expense most people underestimate is the marketing runway: you need enough budget to run ads for the months before organic search starts producing.

Is this better than seasonal Christmas lights? Different business. Bigger ticket, no takedown, no storage, and it sells year-round. The tradeoff is a longer sales cycle and more follow-up per deal. Many operators run both — permanent lighting through the year, seasonal in Q4.

How long before SEO produces leads? Months, not weeks. My rankings moved meaningfully between month three and month five, and the organic click curve only got steep in month five. Run ads to bridge it. Anyone promising faster organic results in a competitive metro is guessing.

What should I do first? Start Google Business Profile verification. It's free, it gates your local visibility, and it takes the longest. Everything else can be done in parallel; that one can't be rushed.

Can I do this as a second business? Yes, and many do — it pairs naturally with gutters, siding, exterior cleaning, and seasonal lighting. The install work is schedulable. The part that doesn't fit around a day job is the follow-up, which is exactly why it should be automated rather than manual.

Where this leaves you

Everything above is doable. I did it with no lighting background, and the install work was never the bottleneck. The bottleneck was the five months of building the machine that feeds the install work — pricing structure, quoting workflow, financing, GBP, reviews, content, ads, CRM, follow-up.

If you want to build it yourself, this guide is an honest map of the route. Take the GBP verification seriously, capture lead source from day one, verify your conversion tracking before you spend, and don't fight the head terms in year one.

If you'd rather not spend those five months, that's what I do now — I build the same system for other installers, on their accounts, in their name. I'm an installer who happens to be good at this half of the business, not an agency that picked up lighting as a vertical.

Either way, the thing I'd most want you to take from this: the lighting is the easy part. Plan for the rest of it.