Roofing advertising in 2026: every channel compared, with the real math.

1. Google Local Services Ads — the strongest single roofing channel

Best for: repair and emergency demand — homeowners searching with a problem right now.

LSA sits above regular ads at the top of Google with the “Google Guaranteed” badge, and you pay per call, not per click. Roofing calls in Miami-Dade, Broward, and Palm Beach typically run $90–$180 each, exclusive to you — and junk calls (solicitors, wrong service, out of area) can be disputed for credit. One observed Broward roofing account averaged $52 per charged call in its first month ($156 spend, 3 leads). The catch: rank isn’t bought, it’s earned — review count, review velocity, and answer speed decide who Google shows first, and screening (license, insurance, background check) takes 1–3 weeks before your first ad runs. Full pricing breakdown: what Local Services Ads cost.

Trade-off: capped by search volume — LSA can’t make more homeowners need roofs; it only wins the ones already searching.

2. Google Search ads (PPC)

Best for: replacement and insurance-claim searches — the five-figure tickets.

Search ads charge per click and give you what LSA doesn’t: keyword control and a landing page that pre-sells the estimate. Replacement and insurance-claim clicks cost more than repair clicks, but they feed tickets 10–15× larger — a page that walks a homeowner through deductibles and documentation converts searchers LSA never sees. Run repair terms through LSA and reserve Search budget for the high-ticket intent. Setup quality decides everything: broad-match keywords on a generic homepage is how roofers conclude “Google Ads doesn’t work.” Details: Google Ads for roofers.

Trade-off: clicks aren’t calls — without a tuned landing page and call tracking, Search burns budget invisibly.

3. Meta ads with real filmed video

Best for: demand creation — reaching homeowners before they search, at the lowest cost per lead.

Facebook and Instagram reach homeowners who have a 12-year-old roof and no urgency yet. That makes Meta leads cheaper — $25–$60 each in South Florida — but slower to close. Creative is the whole game: in one documented two-contractor test, the roofer running real edited job-site video generated leads roughly 15× cheaper than a competitor spending more on AI-generated creative (the full case study). Homeowners scroll past stock footage; they stop for a real crew tearing off a real Miami roof.

Trade-off: needs real footage and follow-up discipline — Meta leads that wait a day in an inbox are dead.

4. SEO + AI-assistant visibility

Best for: the long game — leads that eventually cost effectively $0 each.

City and service pages that rank organically keep producing after the ad spend stops — the only channel where cost per lead falls every quarter. It takes months to build, which is why it’s a terrible first channel and a mandatory third one. The 2026 twist: homeowners and commercial buyers now ask ChatGPT and Perplexity who to hire, and AI assistants recommend companies with citable, current pages — in August 2026, AI-assistant referrals closed at roughly 40–50% for us, about double typical quoted leads. Playbooks: SEO leads for roofers and AI visibility for roofers.

Trade-off: months of patience before the first organic call — start it while paid channels carry the pipeline.

5. Shared-lead marketplaces (Angi, HomeAdvisor)

Best for: plugging an occasional slow week — nothing more.

The marketplaces sell speed: profile today, leads tomorrow. But the same homeowner request is typically sold to 3–5 roofers at once, so the attractive sticker price hides a race — answer in under a minute or you paid for a competitor’s job. Divide by actual booked jobs and shared leads usually cost more than exclusive ones. And nothing compounds: cancel, and there’s no site, no reviews, no ranking left behind. The full breakdown: buying leads vs. generating your own.

Trade-off: renting leads forever at a rising price, while building someone else’s brand.

6. Door-knocking, canvassing & referrals

Best for: storm response and brand-new companies with more time than budget.

The oldest roofing channels still work — a good canvasser after a hail event books inspections, and referrals close better than any paid lead. The limits are structural: canvassing doesn’t scale past the crews you can hire and manage, Florida homeowners are increasingly wary of post-storm door-knockers, and referrals arrive on the customer’s schedule, not yours. Treat these as a floor, not a growth plan — and put a review-request system behind every finished job so word of mouth compounds into Google rankings too.

Trade-off: doesn’t scale, and compliance around storm canvassing keeps tightening.

Exclusive channels

At $90–$180 per exclusive LSA lead and a 1-in-5 close, that’s roughly $450–$900 in ad spend per booked job — against an $800–$2,500 repair or a five-figure replacement. Meta at $25–$60 per lead closes slower (longer nurture), but even at 1-in-8 it lands near $200–$480 per job.

Shared leads

A $60 shared lead sounds like a bargain until the race: closing 1-in-10 while four competitors dial the same homeowner puts you at $600 per booked job — more than exclusive LSA — with zero reviews, rankings, or brand built along the way. The sticker price is the bait; the close rate is the bill.

Working backwards

Want 4 extra jobs a month? At 1-in-5 that’s ~20 leads ≈ $1,800–$3,600/month in LSA spend at Miami prices. Most South Florida roofers start at $2,000–$3,000/month total ad spend. Run your own numbers in the contractor ad-spend calculator or the full benchmarks table.

Storm season moves every price

June through November, every roofer in the county bids at once and per-call prices surge with demand. The companies paying the least in September set up in March: screening passed, reviews stacked, campaigns tuned before the first named storm. Advertising planned around hurricane season beats advertising reacting to it — every year.

Half the market shops in Spanish

English-only campaigns are invisible to a huge share of Miami-Dade homeowners. The search terms differ, the objections differ, and translated captions aren’t bilingual marketing — campaigns have to be written and filmed natively in both languages. Whatever channel you pick, run it in both languages or concede half the map.

Insurance advertising is a compliance question

Florida’s claim environment makes “free roof” bait ads a double loser: junk leads now, regulator attention later. Honest claim-help positioning — inspections, documentation help, straight talk about deductibles — converts the homeowners you actually want and keeps your license clean. Any channel amplifies your message; make sure it’s a legal one.

Reviews are the hidden ad budget

In LSA’s auction, review count, velocity, and answer speed outrank raw budget — a 150-review roofer with 90-second response beats a bigger spender. Modeled at the same ~$3,000 budget, a strong profile lands near ~$55 blended per lead vs ~$111 for a weak one (the worked example). Every channel gets cheaper as reviews grow.

Days 1–21: paperwork & plumbing

Apply for Google Guaranteed screening on day one — license, insurance, background check take 1–3 weeks and nothing runs until they clear. While waiting: call tracking on every number, missed-call text-back live, review requests wired into job close-out, and the Google Business Profile completed section by section. None of this costs media budget, and all of it makes every later dollar cheaper.

Days 21–60: first calls & first data

LSA goes live and typically rings within days. Every call gets answered fast — rank rewards it — and every junk call gets disputed for credit. By week six there’s real data: cost per call, repair vs. replacement mix, which zip codes convert. Search ads launch now on the terms the data says are worth a dedicated landing page.

Days 60–90: compound channels start

Filming begins on active job sites — tear-offs, crews, finished roofs — feeding Meta ads at $25–$60 per lead. The first SEO city/service pages publish, starting the long clock toward near-$0 leads. By day 90 the blended numbers exist in writing: spend, calls, cost per call, estimates — and the budget conversation becomes math instead of opinion.