A click from Georgetown and a click from Buda are an hour apart on I-35
Geography, not keywords, is the first split in an Austin roofing account.
Draw campaigns on the areas your crews genuinely cover, with their own budgets. Central Austin, the Williamson County corridor through Round Rock, Georgetown, Cedar Park and Leander, the Hays side through Kyle, Buda and San Marcos, and the western Hill Country runs out to Lakeway, Bee Cave and Dripping Springs are four different auctions with four different job mixes.
A single radius setting treats all of that as one market. It is not one market. Drive time changes what an appointment costs you in crew hours, and an estimator crossing town twice in a day is an estimator who saw fewer roofs.
Set location targeting to presence rather than interest, and exclude the areas you will not serve rather than hoping the bidding sorts it out. Reviewing the geographic report monthly and pruning is dull work that consistently pays.
Budget follows capacity. If the Williamson crew is booked three weeks out and the south side has openings, the spend should reflect that, not an even split written in January.
Hillside and steep-pitch work carries its own price, so stop averaging it in
Terrain west of the city produces jobs that cost more to win and are worth more to sign, and blended reporting hides both facts.
Steep pitch, multiple stories, cut lots over limestone with narrow approaches: those jobs need different crews, different staging and different quoting. They also attract different searches, mostly around metal, tile and non walkable roofs.
Separate them. A campaign or at least a clean label set covering the western submarkets lets you tolerate a higher click cost where the job value justifies it, and stops that cost from being spread across repair traffic.
Write ad copy that speaks to it. An ad that mentions steep and complex roofs will lose some clicks and keep the right ones, which is the entire point of paid search on a limited budget.
Track it separately from the first day. Without the split, an account that is quietly profitable on hillside replacements and quietly losing money on cheap repair clicks looks like an account that is doing fine.
Rooftop bars and roof racks will spend an Austin budget by Thursday
The negative keyword list matters more than the bids, and here it starts with words that have nothing to do with a roof over a house.
Rooftop bar, rooftop patio, rooftop pool and roof deck all carry the word roof and none of them are your customer. Roof rack, RV roof and metal building kits do the same damage.
Then cut the job seekers. Roofing jobs, roofing hiring, roofing salary and roofing crews needed will pull clicks all week, especially from a market with constant in-migration and people looking for work.
Cut the researchers who will never call: DIY roof repair, shingle prices per bundle, roofing calculator, roofing school. Some of them can be worth a content page and none of them are worth a bid.
Build the list before launch, then read the search terms report weekly for the first month and monthly after. Broad match without disciplined negatives is the fastest way to spend a roofing budget on nothing.
Solar removal and reset is demand almost no roofer is bidding for
Panels on a roof create a narrow, specific search that most local roofing accounts have never touched.
A city-owned utility has made rooftop solar a normal feature of this housing stock rather than an exotic one. That means a growing number of roofs due for replacement have an array sitting on them.
The homeowner searches something like solar panel removal for roof replacement. Very few roofing advertisers show up, because most accounts are built around three generic phrases and a radius.
Send that click to a page that explains the sequence, who handles what, and what generally has to be coordinated with the electrical work. Do not send it to a homepage. The specificity of the search is the reason it converts.
Keep it in its own ad group with its own copy. Volume will be modest. Job value and close rate on a search that intent-heavy usually make up for it, and you can measure that within a quarter rather than guessing.
Summer shortens the working day, so pace the spend the way you pace the crew
Sustained heat changes when crews work and when homeowners think about their roof, and a flat monthly budget ignores both.
Crews start early and lose the hottest part of the afternoon. If your production capacity drops in August, buying the same volume of appointments in August is buying a backlog and a pile of canceled estimates.
Homeowner behavior shifts too. Attic heat and cooling complaints send people looking in midsummer, while replacement research tends to cluster around the milder stretches when standing outside is tolerable.
Set ad schedules against the hours somebody actually answers. A lead that sits until the next morning is worth less than the same lead answered in ten minutes, and there is no bid adjustment that fixes an unanswered phone.
Pacing should be a monthly conversation about the production calendar, not an annual budget line. Slowing spend when the schedule is full and pushing it when a gap opens is the least glamorous lever in the account and one of the most valuable.
Judge the account on crew days sold, not on the count of forms in the dashboard
Lead volume is the easiest number to inflate and the least useful one to run a roofing company on.
Push conversions back from the CRM into the ad platform with values attached: appointment held, inspection completed, contract signed. Bidding optimizes toward whatever you feed it, and if you feed it form submissions it will find you people who like filling in forms.
Use call tracking properly. A large share of roofing inquiries arrive as calls, and an account that only counts web forms is reporting on a fraction of what it produced.
Report cost per signed contract by geography. The Hays County job and the Central Austin job can look identical on a cost per lead line and land in different places once drive time, close rate and job value are included.
Crew days sold is the honest summary. It captures whether the money bought work your business can actually deliver, which is the only question the owner was ever asking.
Questions we actually get
- What should we spend per month to make paid search work here?
- It depends on how much of the metro you are targeting and how competitive your job types are, and any agency giving you a figure before seeing your service area and average job value is guessing. The practical starting point is narrower: pick the geography your crews cover best, fund it well enough to gather data on signed jobs, and expand from evidence instead of buying the whole map at a thin budget.
- Should we run Local Services Ads as well as search ads?
- They occupy different real estate and answer different buying moments, so they are usually complementary rather than a choice. Local Services Ads require verification and generally reward responsiveness, which means they punish a company that does not answer the phone. Decide based on whether your intake can handle the calls before you decide on budget.
- Is it worth advertising for storm damage and insurance work?
- Tread carefully. Insurance language in ad copy and keywords carries real compliance and platform risk, and it attracts a caller who wants a ruling you cannot give. It is generally safer to advertise the inspection and the repair, keep claim language off the ads, and tell homeowners that coverage questions are for their insurer.
- How do we stop paying for clicks outside our service area?
- Set targeting to people present in the location rather than those merely interested in it, exclude the areas you will not drive to, and check the geographic report on a schedule. Radius targeting around a single address is the usual culprit, because a circle wide enough to reach Cedar Park also reaches a lot of places your dispatcher would rather not send a truck.
- How quickly should someone respond to a paid inquiry?
- As fast as your operation allows, and the difference between minutes and hours is usually larger than the difference between two ad headlines. Before increasing spend, it is worth measuring your answer rate by hour of day, because paying for calls nobody picks up is the most expensive mistake available in this channel.
