Where a roofer's next ad dollar goes after the LSA move

Mid-September is when roofing owners settle next year’s money. Spring’s hail claims are mostly paid, coastal crews are watching the hurricane peak go by, and the March budget has to be agreed before the slow months arrive. This year there is a new wrinkle: in August, Google began moving Local Services Ads into Google Ads, and roofing was named in the first group to go.

Here is the decision laid out with published numbers: what each channel costs, what changed, and where we would put the next dollar.

What a roofing lead costs in the published numbers

The most useful benchmark for roofers is LocaliQ’s 2025 report on home services search ads, built from 3,211 campaigns between April 2024 and March 2025. Its Roofing & Gutters category stands out for the wrong reasons:

• Cost per click: $10.70, the third highest of the 16 home service categories.
• Conversion rate: 3.70%, the second lowest. Fewer than four in a hundred clicks became a lead.
• Cost per lead: $228.15, the highest of all 16, against $90.92 for home services as a whole.

WordStream by LocaliQ’s 2026 benchmarks, covering April 2025 to March 2026, put a click in the broader Home & Home Improvement category at $8.33, and a lead across every industry combined at $66.69. Its 2026 benchmarks for Facebook ads, published this week, show leads from lead-form campaigns in Home & Home Improvement at a median of $42.95, with a 5.32% conversion rate.

The low conversion rate is the part to take seriously. Plenty of roofing searches are research: a homeowner pricing a job months out, checking a contractor after a knock on the door, or looking up what hail damage looks like. At the median, every extra lead it takes to sign one roof adds another $228.15 to what that roof cost you to win.

What changed in Local Services Ads this summer

Google’s own migration notes say the first phase began in August 2026 for select home service advertisers, roofing among them, and that Local Services Ads become Performance Max campaigns with pay-per-lead goals. In plain terms:

• What stays. You still pay only for valid leads: calls, messages and, for businesses here and in Canada, bookings. The Google Verified badge carries over, and so do the Search and Maps placements.
• What goes. Manual bidding, such as setting a maximum cost per lead, is no longer supported. Insurance and license reverification are no longer required.
• What is already gone. Google no longer credits leads marked “job type not serviced” or “geo not serviced”.

Put those together and the control has shifted. With no maximum bid to lean on, what you feed the campaign does the steering: the job types you switch on, the service area you draw, the weekly budget, and how fast someone answers. Google’s ranking for these ads weighs profile quality, which it defines as your rating, number of reviews, average response time, use of high-quality images and the verification checks you have completed. A roofer with 12 reviews and a voicemail greeting loses placements, and the leads that come with them, to a better-reviewed competitor that picks up.

That is why reviews belong in an ad budget: BrightLocal’s 2026 Local Consumer Review Survey found 47% of consumers will not use a business with fewer than 20 reviews.

Shared leads: judge them on signed roofs

No independent benchmark covers what shared roofing leads cost or how often they close, so here your own records are the only numbers that count. Lead marketplaces can work for roofers, usually for repair volume while a company builds its own reviews. The test is simple and takes one full season:

• Tag every marketplace lead in your CRM the day it arrives.
• Record how quickly someone called back, because a shared lead is often a race.
• Count the inspections, the estimates and the signed contracts it produced.
• Compare cost per signed roof with Local Services Ads and search, across at least one storm month and one quiet month.

If the marketplace wins on signed roofs, keep it. If it only wins on cost per lead, it is not winning.

Insurance work and retail work need separate math

Harvard’s Joint Center for Housing Studies found that homeowner insurance settlements funded 18% of 2023 spending on exterior replacements such as roofing. The Insurance Information Institute, using ISO data, puts the average homeowners claim for wind and hail at $14,747 over 2019 to 2023, outside Texas and Alaska, with 2.80 such claims a year for every 100 insured homes.

Those jobs behave differently from retail ones. They arrive in bursts after storms, and the Texas Department of Insurance tells homeowners that replacement cost claims are paid in two checks, the second after work starts. In Colorado a homeowner may rescind a roofing contract within 72 hours of a written claim denial. A channel judged on a storm month will look cheap; judged in October, the same channel may look expensive. Blend insurance and retail jobs into one cost-per-lead figure and you cannot tell which channel is paying for the trucks.

Where we would put the next dollar

For an owner-run roofer with a couple of crews, this is the order we usually argue for, first to last:

• First, the profile and reviews the paid channels lean on. BrightLocal’s July 2026 survey of 1,227 consumers found 52% began their latest local search on Google Search and 9% on Google Maps. Photos of your own roofs, a review request after every job and a reply to each review cost little and lift both the map and Local Services Ads ranking.
• Then Local Services Ads, once reviews are solid and calls are answered live.
• Then search ads on leak and storm-damage terms in the weeks before and during your storm window, and replacement terms in the shoulder months. In LocaliQ’s data these leads cost more than in any other home service, so they need tight negative keywords and the closest control.
• After that, Facebook ads for before-and-after proof and follow-up inside a storm footprint. The leads cost less and need a faster callback.
• Last, shared leads, kept only where your own records show signed roofs.

What to do this month

• Export the last twelve months of leads and mark each one with source, job type, insurance or retail, and whether it signed.
• Work out cost per signed roof for each channel, once for storm months and once for quiet ones.
• If your Local Services Ads account has moved into Google Ads, review job types and service areas now that wrong-area leads go uncredited.
• Keep review requests going through the winter, so the reviews homeowners read in April are recent.
• Set next spring’s budget month by month against your own storm calendar rather than as an even split.
• Decide who answers an evening call in a storm week, before the next storm decides for you.

No benchmark can tell you what your next roof will cost to win, but a season of your own tagged leads can. Our page for roofing companies sets out how we plan that work, and our Google Ads and Local Services Ads work shows how we run the paid side alongside review requests.