The State of Roofing Advertising 2026
We checked 834 roofing companies across 17 US metros and counted every live ad they run. More than half advertise, almost all of it on Google, and the two hardest hail markets do not have a single live Meta ad between them. Here is the full picture.
Built from Muffin Intel, our in-house ad-intelligence tool. Each roofing company was resolved by its website domain, matched to its Google advertiser ID, and its ads counted by that ID across Google Ads Transparency Center and Meta Ad Library. n = 834 roofing companies across 17 US metros. Google counts read to a depth of 200 ads per advertiser, so the largest advertisers read as 200 and true totals run higher. Snapshot dated July 29, 2026.
Roofing is supposed to be the most cutthroat ad market in local services. Storm chasers, door knockers, and a bidding war on every hail cloud.
We wanted the actual numbers, so we pointed Muffin Intel at 834 roofing companies across 17 US metros, from Dallas to Minneapolis to Tampa. Each company was resolved by its website domain, matched to its Google advertiser ID, and every ad it runs was counted. Not estimated. Counted.
The roofers turned out to be far more serious about advertising than the med spas we studied in June. But they are serious about exactly one channel, and the gap they have all left open is the most interesting thing in the data.
More than half of roofers advertise. This is not a quiet market
Of 834 roofing companies, 447 are running ads right now. That is 54 percent, more than three times the 16.5 percent we measured among US med spas. Roofing is a category where the average operator has already decided that ads pay.
The rate is not uniform. Houston leads at 35 of 50 companies advertising, 70 percent. Atlanta trails at 15 of 50, 30 percent. Every other metro sits somewhere between, and even the quietest market here is busier than the loudest med spa market we have measured.
Google carries 32 times the ad volume of Meta
Here is where the market shows its hand. Across all 17 metros we counted 4,572 live Google ads against 141 live Meta ads. Thirty two times the volume on search. By advertiser it is just as lopsided: 409 companies run Google ads, 75 run Meta ads, and only 37 run both.
The logic is sound as far as it goes. Nobody scrolls Instagram wondering whether to fix a leaking roof. The demand arrives with the weather, the homeowner types roof repair near me, and the search auction is where that moment gets bought. Roofers have collectively concluded that Google is the channel, and on intent capture they are right.
141 live Meta ads across 17 metros is eight per city. An entire channel, effectively empty, in a category built on before and after photos.
The harder the storms hit, the more the money moves to search
Sort the metros by storm profile and a pattern appears. Oklahoma City and Tulsa, the two hail and tornado markets, run 847 live Google ads between them and exactly zero live Meta ads. Not few. Zero. When demand arrives by hailstone, every dollar goes to the moment the homeowner searches.
Now look at the calm end. Seattle and Stamford, the two low-storm markets in the study, hold 38 of the 141 live Meta ads, 27 percent of all Meta activity from 10 percent of the companies. Where storms do not hand roofers ready-made demand, a few of them have started creating it in the feed. The rest of the country has not.
Roofing ads run long. Very long
Among the 431 advertisers with dated Google ads, the average longest-running single ad has been live for 548 days. 250 of them, 58 percent, have an ad that has run unchanged for at least a year. 124 have one past the two year mark. The record holder, a Kansas City roofer, has an ad that has been live for 1,739 days. Almost five years.
Some of that is deliberate. An evergreen search ad on roof repair keywords can keep earning as long as the phone keeps ringing. But an ad running unchanged since 2021 is more often an ad nobody has looked at since 2021. In a market where 54 percent of competitors are live, the roofer who actually reads the auction, refreshes copy, and tests against those five-year-old ads is competing against inertia, not effort.
The Meta gap is the strategic finding
Roofers have solved intent capture. What almost none of them have touched is demand creation, and roofing is unusually well built for it. The product is visual. The before and after is dramatic. The purchase is high ticket. The homeowner who sees a roofer's work in the feed for six months is the homeowner who calls that roofer first when the hail comes, before the search auction ever starts.
The numbers say this lane is open almost everywhere. Eight live Meta ads per metro. Zero in Oklahoma City and Tulsa. Only 37 companies in the entire study run both channels. For a roofer in a storm market, a steady Meta presence is not a way to outbid 409 Google advertisers. It is a way to be remembered before the bidding starts, at a moment when nobody else in your city is trying.
The playbook the data supports has three parts. Hold your ground on search, because that is where the storm demand lands. Build the Meta layer your competitors have left empty, with real project photos, not stock. And refresh what you run, because the average competitor's best ad is a year and a half old.
How we know this
This report is built from Muffin Intel, the ad-intelligence engine we run in-house. Each of the 834 roofing companies was resolved by its website domain, matched to its Google advertiser ID, and its ads counted by that ID, with Meta activity confirmed through Meta Ad Library. The figures are counts of real ad activity as of July 29, 2026. Google counts read to a depth of 200 ads per advertiser, so the biggest advertisers read as 200 and the true totals run higher, not lower.
We update this dataset continuously, so the numbers move. The pattern does not.
Want your spa's slice of this?