Facebook Advertising For Contractors: A Step-By-Step Guide

An ad cannot fix a weak offer. This step decides more of your results than everything below it combined.

"Quality work, free estimates" is not an offer. It's what the other 30 contractors in your market already say.

A real offer gives the homeowner a concrete reason to act now instead of someday. It removes a risk, names a number, or puts a deadline on something.

Good raw material for offers: financing terms your lender actually supports, a fast written quote, a named discount with an end date, a free add-on with a real dollar value, a warranty you already give but never advertise.

The rule: never advertise anything you wouldn't put in the contract. An inflated offer gets you cheap clicks and furious homeowners at the kitchen table.

This part is 20 minutes of clicking, and you only do it once.

• Business Page first. You probably have one. Make sure the phone number, service area and photos are current, because homeowners will check it before they call.
• Business portfolio second. Go to business.facebook.com and create one. It's the container that owns everything else.
• Ad account third. Inside the portfolio, create the ad account, set your time zone and currency, and add a payment method.

Do this yourself, in your own portfolio, even if you hire an agency later. The companies that lose their ad account history are the ones that let someone else own it.

One more thing: a brand-new ad account has no track record with Meta. Start with a modest budget and clean, honest ads, because fresh accounts that come out swinging with big spend and aggressive claims are the ones that get flagged.

When a homeowner taps your ad, they go somewhere. That somewhere decides your lead quality, so pick it now, not after launch.

You have two options, and the honest comparison looks like this.

Instant forms open right inside Facebook with the homeowner's name and email already filled in. They're easy to set up and the leads are cheaper, but two taps of commitment means a lot of people who barely remember submitting.

A landing page is a simple page you control: your offer, photos of your work, and a short form. It costs more per lead, and the leads are worth more, because the homeowner had to actually do something.

If a landing page isn't realistic for you this month, run instant forms, but add your own qualifying questions to the form and make the phone number field required. Autofilled contact info with no questions is how you end up agreeing with the Reddit threads below.

If you do use a landing page, install the Meta Pixel on it. It's a snippet of code that tells Meta which visitors became leads, so the system gets better at finding people like them.

Open Ads Manager, hit Create, and Meta asks what you want. This screen is where the boost-button crowd goes wrong.

Pick Leads. This one choice tells Meta's system to hunt for people who fill out forms, instead of people who like, comment and keep scrolling.

Then set the conversion location to match step 3: your website if you built a landing page, instant forms if you went that route.

Keep the structure stupid simple: one campaign, one ad set, one job type. You can get fancy after the first winner shows up.

Here's the part nobody wants to hear: the clever targeting you're imagining mostly doesn't exist anymore.

Meta removed most of the detailed options years ago, and its delivery system is now better at finding your homeowner than a stack of interest filters is.

So your targeting is three decisions:

• Location: a radius or set of cities matching where your crews actually drive. Don't pay to show ads in towns you'd turn down.
• Age: set a sensible minimum. A 21-year-old renter is not buying a kitchen remodel.
• Everything else: leave it broad, and let your ad do the filtering.

That last line is the trick. When your video opens on a sagging fence or a fogged-up window, homeowners with that problem stop and everyone else scrolls. The creative is the targeting now.

The budget question has a simple logic to it: spend enough to buy about one lead a day at your trade's going rate.

On our accounts, cost per lead ran $29.45 to $119.18 depending on the trade. Big-ticket trades sit at the top of that range, and that's fine, because the job is worth 100 times the lead.

Two rules that matter more than the exact number:

• Pick a number you can run for 30 days straight. A budget you panic-pause on day 9 buys you nothing but a confused algorithm.
• Don't spread it thin. $50 a day behind one campaign beats $10 a day behind five.

And if $1,000 to $1,500 a month of ad spend would genuinely hurt, fix that first. Ads amplify a business; they don't rescue one.

You do not need a production crew. The best-performing contractor ads we run look like they were shot between tasks on a job site, because they were.

What they all share is one discipline: the homeowner's problem is on screen in the first second. The rot, the fog, the crack, the leak. Their problem, not your logo.

Talk the way you talk at an estimate. If you wouldn't say "transform your outdoor living space" to a homeowner's face, don't put it in an ad.

Make three versions with different opening lines and let them compete. You'll be wrong about which one wins; everyone is. We've written hundreds of these, and real examples by trade are on our fence, deck and remodeling ad example pages.

No video yet? A sharp before-and-after photo of your own work with the offer over it is a legitimate first ad. Start there, film later.

Hit publish. Now comes the hardest step on this page: don't touch it.

Meta's system spends the first days in a learning phase, testing your ad against different pockets of people. Early numbers swing wildly, and edits restart the learning from zero.

Your only jobs in week one: confirm money is moving, confirm leads are landing where they should, and answer the ones that come in.

Judging an ad on two days and $40 of spend is the most common way contractors kill an ad that was about to work.

After a week, Ads Manager will show you a wall of numbers. You can ignore almost all of it.

Likes are not jobs. Reach is not jobs. Even clicks are not jobs. Three numbers matter, and each one answers a different question.

Cost per lead tells you which ads are working. It's the number you use week to week to decide what to kill and what to keep.

The rhythm, once you have data:

• Give each ad a fair run before judging: roughly three times your target cost per lead in spend.
• Kill the clear losers. No mercy, no "maybe it'll turn around."
• Keep the winner running and test one new variation against it. That loop, repeated, is the entire game.

What should your cost per lead target be? Anchor it to your ticket, not to someone else's trade. A $100 lead on an $8,000 job is a rounding error; the same lead on a $500 repair is a problem. Trade-by-trade breakdowns: fence, deck, remodeling, windows.

Cost per booked estimate tells you whether the leads are real. Take what you spent and divide it by the estimates that actually landed on your calendar.

Say you spend $1,000 and get 10 leads. If 5 of them book, each estimate cost you $200. If only 2 book, the same ads cost you $500 an estimate, and the problem is lead quality or follow-up, not the ad.

ROI tells you whether any of it is making money. Add up the jobs you signed from the ads and divide by what you spent on them.

If the ads bring back at least 2 to 3 times what you spend on them, you can run them profitably. Most of our clients do at least 5x.

That's why a higher cost per lead isn't automatically a problem. A $150 lead that books and signs beats a $40 lead that never picks up the phone.

To track it, tag every lead in your CRM or a simple spreadsheet with where it came from. When a job signs, you'll know it was Facebook.

Here's an uncomfortable truth from inside the accounts: half of "bad lead quality" is really slow follow-up.

A homeowner who submits a form is sitting on the couch thinking about their project right now. Call in the first few minutes and you're the helpful contractor who's on it. Call tomorrow and you're an interruption from a company they don't remember.

The minimum system, no software required:

• Call the moment the lead lands. No answer? Text right away: "Hi, it's [you] from [company], just got your fence quote request. When's a good time for a quick call?"
• Try again that evening, and once the next day. Most contractors quit after one attempt, which is why the persistent ones win.
• Book the estimate on the call. A named day and time, not "I'll get back to you with some options."

Across our contractor accounts, clients report booking 50 to 80% of the leads we generate. The clients at the top of that range are, without exception, the fastest to the phone.