Your Mount Juliet crew should not be paying for Murfreesboro clicks
Geography is the first and most expensive decision in a roofing account, and a radius drawn around the yard ignores the county lines that change everything else about the job.
Metro Nashville and Davidson County run as one consolidated government. Cross into Williamson, Rutherford, Sumner or Wilson and the rules, the housing stock and the buyer all change. A single campaign covering all of it is one budget serving five different sales conversations.
Split by county, not by radius. Each county gets its own campaign, its own budget, its own ad copy and its own landing page. Then you can see which one actually converts into work and move money accordingly, instead of averaging a good county and a bad one into one number that tells you nothing.
Set targeting to people in your locations rather than people interested in them, or you will pay for someone in another state reading about Middle Tennessee roofs.
Drive time is a real cost. If Mount Juliet is a stretch for your crews on a repair but fine on a full replacement, that is a targeting decision, not a dispatch problem to solve after the lead comes in.
Exclude the counties you do not serve outright. Money spent on a click you cannot dispatch is worse than money not spent, because it also occupies your intake staff.
The three phrases worth real money are leak, replacement cost and one covering type
Roofing keyword lists get bloated because the word roof appears in a hundred searches that have nothing to do with hiring a contractor.
Start narrow. Active leak language, replacement and tear off pricing language, and the one or two covering types you actually want to install. Those searches carry intent and a check behind them.
Separate repair intent from replacement intent into different campaigns. They convert at different speeds, at wildly different job values, and mixing them means the cheap, fast conversions eat the budget that was supposed to buy replacements.
Add the storm language as its own campaign that you can turn up and down rather than leaving it running at a trickle all year.
Match types matter more than most contractors are told. Broad match on a roofing term without a heavy negative list is a donation. Start on phrase and exact, read the search term report weekly for the first month, and build the negative list from what actually comes in.
Brand terms are cheap and worth owning, especially in a market where competitors will bid on your name.
Gutter guards, metal barn kits and job boards are where a roofing budget quietly drains
The search term report is the most useful document in a roofing account and almost nobody reads it after the first week.
The waste is predictable and specific. People looking for work rather than a roofer. Homeowners pricing gutter cleaning. Farm and pole barn metal panel shoppers. Students researching roof types. Suppliers. Anyone typing the word free.
Build the negative list as a living document. Every week, pull the terms that spent money and produced nothing, and add them. The list gets long, and a long negative list is a sign the account is being managed rather than parked.
Watch for the adjacent trades that share vocabulary: siding, gutters, skylights, chimney work. Some of those are jobs you want, and some are calls that waste an estimator's morning. Decide deliberately rather than by default.
Exclude your own service techs and office staff from seeing and clicking the ads, and add IP exclusions for the office. Small leak, easy fix.
A storm week will eat the month's budget by Wednesday unless the pacing says otherwise
Roofing demand in Middle Tennessee is not flat, and a flat monthly budget is a decision to be underfunded in the weeks that matter and overfunded in the weeks that do not.
Ice, wind and hail create a few days when everyone in the county is searching at once. Those are the days worth spending on, and a capped daily budget will simply stop showing your ads at the moment demand peaks.
Hold a reserve outside the routine monthly spend so there is something to deploy when weather actually creates demand, and give somebody the authority to release it without a meeting.
Pace against the crew calendar in the other direction too. Booked out four weeks on replacements is a reason to pull back on replacement bidding and let repair run, not a reason to keep buying leads you will make wait.
Dayparting is worth testing rather than assuming. Calls that come in when nobody can answer are the most expensive clicks in the account, so either staff the hours or stop bidding in them.
Winter slow weeks are not automatically a reason to go dark. They are usually a reason to shift budget toward repair and maintenance intent and away from replacement.
Short-term rental managers near Broadway convert on a date, not on a discount
A meaningful share of housing downtown and in the neighborhoods around it is operated as a business, and that buyer is bought differently from a homeowner.
The manager running units in The Gulch, on the East Nashville side, or in a small portfolio scattered across Davidson is buying turnaround, remote approval and a clean invoice. A free inspection offer means nothing to them.
Give them their own campaign and their own landing page. The offer is a date and a process: how fast someone can be on site, whether you can work without the owner present, whether you invoice per property, how vendor onboarding works.
Search volume for this is smaller than the homeowner side and the cost per click tends to be lower, because most roofing advertisers have never thought about it. Small campaign, disproportionate value, because portfolio buyers do not come one door at a time.
Do not publish any position on short-term rental rules in the ad or the landing page. Those rules are contested and change, and it is worth telling anyone who asks to confirm with Metro.
Send the signed jobs back to Google, or you optimize for whoever types fastest
Cost per lead is a vanity number in roofing, because the cheapest leads are almost always the ones that never turn into a roof.
The account should be optimizing toward booked and signed work, which means job outcomes have to make it back into the ad platform. Import offline conversions from the CRM: inspection booked, proposal issued, contract signed, with values attached.
Once the platform can see contract value it stops chasing the fifteen dollar click that produced a tire kicker in a county you barely serve.
Call tracking is not optional for a trade where most inquiries are phone calls. Track at the keyword level, record calls where permitted, and let intake review them.
Report by county. A blended cost per booked roof across five counties will hide the fact that one of them is carrying the rest.
We will not promise a lead volume or a cost per job. Anyone who does is quoting a number they made up. What we will commit to is that the reporting shows signed roofs and what they cost, not clicks.
Questions we actually get
- How much should a roofing company spend on Google Ads in Nashville?
- There is no honest universal number, and we will not invent one. The useful way to set it is backward from capacity: how many replacements can your crews actually build in a month, what is your close rate on estimates, and how many estimates does that require. Then fund the counties you most want to grow in and hold a reserve for storm weeks rather than spreading a flat monthly figure evenly.
- Should we run Local Services Ads as well as search ads?
- They are different buys and worth evaluating separately. Local Services Ads sit in their own unit and are charged per lead with a screening process, while search ads give you control over keywords, geography and landing pages. Many contractors run both and measure them independently. What matters is that they are not sharing a budget line, because you will never be able to tell which one produced the work.
- Why separate campaigns for repair and replacement?
- Because they behave nothing alike. Repair intent converts fast, cheap and small. Replacement intent converts slowly, expensively and large. In one campaign the fast cheap conversions win the algorithm's attention and quietly starve the replacement side, and the blended cost per lead will look great while your crews sit on small jobs.
- How quickly can paid search start producing calls?
- Ads can be live within days, and calls generally start once they are. What we will not do is promise a call volume or a timeline to profitability. The first few weeks are mostly about reading search terms, building negatives and confirming which counties actually convert, so budget expectations should account for that learning period.
- Do we need call tracking and CRM integration to run ads?
- Effectively, yes. Roofing inquiries arrive by phone, and without call tracking you cannot tell which keyword produced them. Without pushing job outcomes back from your CRM, the ad platform optimizes toward form fills rather than signed contracts. Both are setup work, and both are the difference between managing an account and guessing at one.
