Why your Facebook ads brought you price shoppers

Almost every contractor I talk to has run Facebook ads before. The story is close to identical every time.

An agency took the logo. Asked for some photos off the phone. Wrote something like Get Your Free Estimate Today. Pointed it at homeowners 35 to 65 within twenty-five miles. Ran it for four months.

The leads that came in wanted a price over the phone and vanished when they heard one. Half the numbers didn't pick up at all.

And the conclusion everybody draws from that is: Facebook doesn't work for my trade.

It's the wrong conclusion, and it's an expensive one to hold.

Targeting stopped being the hard part

Here's what changed and almost nobody told you.

Meta's algorithm handles targeting now. Advantage+ and broad targeting consistently beat hand-built interest stacks in home services accounts. You don't out-clever it by adding "interested in home improvement" as a filter — the system already knows who's renovating from behavior you can't see and wouldn't have guessed.

Which sounds like good news. It mostly is. But it has a consequence people miss:

If the machine does the targeting, targeting can't be your edge anymore.

The only variable left that meaningfully moves your cost per lead is the creative. What the ad actually is. And creative is the one thing most ad agencies don't make.

What "bad creative" actually means here

It doesn't mean ugly. Plenty of ads that flop look perfectly professional.

It means the ad didn't earn the two seconds it needed. On Meta, nobody is looking for you. They're looking at their nephew's baby photos, and your ad is an interruption. A search ad works even when it's boring, because the intent was already there. A Meta ad that's boring simply doesn't get watched, and an ad nobody watches can't be optimized into working. You can throw budget at it forever.

So what does earn the two seconds?

Something recognizably real. A crew that looks like a crew. A truck with a dent. A job site that isn't staged. People are extremely good at spotting stock footage, even when they can't articulate why something felt fake — and the instant it feels fake, they're gone.

Specificity. "Free estimate" is what every contractor in Pinellas County says. "Here's what we pulled out of a Clearwater kitchen line yesterday, don't put this down your drain" is what one plumber says.

A reason for the price. This is the part that decides whether you get price shoppers or buyers. If your ad shows the finished pool, you've given somebody exactly one thing to compare on, and it's cost. If it shows dig day, the forming, the plumbing rough-in, the tile going in — now they understand why one builder charges more than another, and the conversation stops being about the number.

That last one isn't a marketing trick. It's just true, and showing it is the whole job.

Price shoppers are a creative outcome, not a targeting outcome

This is the part worth sitting with.

You didn't attract price shoppers because the audience was wrong. You attracted them because the ad gave them nothing to evaluate except price. Show a finished product and a "free estimate" offer, and you have filtered for the person who cares only about the number. The ad worked exactly as built.

Change what the ad shows and you change who responds. Same audience. Same budget. Different people calling.

The two things to check before you blame the platform

Before deciding Facebook doesn't work for your trade, check these. In my experience they explain most dead accounts.

One: is the tracking actually correct? Pixel installed is not the same as Conversions API configured and firing on the right event. A shocking number of accounts are optimizing for landing page views or, worse, nothing at all. If Meta doesn't know what a good outcome looks like, it will happily buy you the cheapest clicks in Florida — and cheap clicks are exactly who called you.

Two: how many pieces of creative were actually running? If the answer is one video and three images, there was nothing to test. Creative testing isn't a tactic you bolt on; it only exists if you're producing enough to have something to test with. One hero video and a prayer isn't a campaign.

What I'd do instead

Half a day on your job site. Real footage — your crew, your trucks, the work in progress, not the after photo. Eight cuts out of that one shoot, with different hooks on the front. Run them against each other, kill the losers in the first week, put budget behind whatever survives.

Then fix the tracking so the algorithm knows what a booked job looks like, and put instant follow-up behind the form so a lead gets a text before they've closed the tab.

That's not clever. It's just the part most agencies skip, because it requires showing up with a camera.

---

If you've run ads before and they didn't work, I'd genuinely like to see the account. Usually the problem is visible inside ten minutes, and about half the time it isn't something you need to hire anybody to fix.