A Kentucky-only license makes half the metro a waste of budget
Before anybody touches a keyword list, decide which side of the Ohio River you are equipped to serve.
Louisville is in Kentucky. Jeffersonville, New Albany and Clarksville are in Indiana, minutes away, in the same metro and the same dinner-table conversation. Licensing, registration and taxes differ by state, and it is worth confirming your standing with the local building department, your insurer or your own counsel.
Google does not know any of that. A twenty-mile radius from a yard near the Watterson spends real money in Clark and Floyd counties whether or not you can take the work.
Run Kentucky and Indiana as separate campaigns with separate budgets, separate ad copy and separate landing pages. Then you can see what each side actually costs and shut one off without touching the other.
Use presence targeting rather than presence or interest, and read the location report every month. Radius settings drift, and platform defaults do not care about state lines.
In a shift-work town, half your cheap clicks are people looking for work
The negative keyword list moves more money here than the bids do, and the largest single drain is employment traffic.
A large logistics and air-freight base means a lot of people in this metro are searching for work at every hour, and roofing crews are hiring. Those clicks look cheap and convert to nothing.
Start negatives with jobs, hiring, careers, salary, apprentice, training, resume. Then add DIY, supply, wholesale, distributor, gutter guard, metal building kit, barn, solar panel, shed and tarp.
Read the search terms report weekly for the first month, then monthly forever. Broad match and smart bidding will keep finding new ways to spend on the wrong intent.
Competitor terms are a decision, not a default. Buy them only if you have a landing page that makes an argument rather than a comparison chart nobody believes.
Third shift changes when your ad money should be on the table
A metro that runs around the clock searches around the clock, and most roofing ad schedules assume a nine to five homeowner.
Somebody coming off a night shift at seven in the morning researches the roof then. Somebody heading into one at eleven at night does the same. After-hours availability is a genuine differentiator here rather than a slogan.
Do not simply switch to twenty-four hour delivery. Match the schedule to who answers. If a person picks up at six in the morning, bid then. If an answering service takes it, decide in advance what that service is allowed to promise.
Call-only ads and call extensions belong in the hours you actually answer. Outside those hours, push to a form that asks for a photo and states a callback window.
An ad that runs at three in the morning into a voicemail box is a donation.
January sells repairs and June sells replacements, and one budget cannot do both
Real winters split roofing demand into two products that need separate money and separate pacing.
Wind, ice and freeze-thaw drive urgent repair searches in the cold months. Full replacement research clusters in the warmer stretch, when a homeowner has time to get three bids.
A flat monthly budget spread evenly across the year serves neither. Pace against the crew calendar and the season, and decide ahead of time what happens to the budget during a wind event.
Keep repair campaigns cheap to run and easy to switch off. Give the patient money to replacement, where the sales cycle is longer and the value is higher.
Write the storm-week plan before the storm. Who lifts the budget, who answers, what you promise on scheduling, and when you stop advertising because the calendar is full.
Name Jeffersonville in the ad, not Kentuckiana
Regional shorthand reads as a Louisville company hedging, and the specific place name answers the only question the searcher has.
A homeowner in New Albany already suspects you are a Kentucky outfit that will quote high and show up late. Naming their city in the headline settles it faster than any trust badge.
Run at least one landing page per state. The Indiana page names Indiana cities, describes Indiana work, and states what you are set up to do there without asserting requirements you have not confirmed.
If you split out Oldham or Bullitt County campaigns, they get their own pages too. Same phone treatment, same form, same geography language from ad to page.
Consistency is not a style preference. A click that lands on a page contradicting the ad is a click you paid for twice.
Drive time to Oldham County belongs in the cost per job math
Cost per lead flatters the wrong campaigns; cost per booked job, with the drive priced in, tells you what to keep.
Push booked jobs and contract values back from your CRM into the ad platform. Without that, the system optimizes toward whoever fills out forms fastest, which is not the same as whoever signs.
Then split the number by geography. A signed roof in eastern Jefferson County near the Gene Snyder and one an hour out do not consume the same crew hours, even at the same price.
Report cost per completed inspection and cost per signed job, by campaign and by state. Two columns, always.
Expect surprises. Some expensive clicks will turn out to be your most profitable geography, and some cheap ones will lose money quietly for a year. Only the booked-job number exposes it.
Questions we actually get
- What should we expect to spend to get started?
- It depends on your service radius, your average job value and how many campaigns the geography forces you to split. What we will not do is quote a cost per click or a lead number for this market, because we have no verified figure for it and neither does anybody who tells you one.
- Should we run Local Services Ads too?
- Worth evaluating early, because those units sit above search ads and are priced per lead rather than per click. They also come with their own verification requirements that generally differ by state, so confirm what applies on each side of the river before you build around them.
- How do we handle a wind or ice event?
- Decide the plan in advance: a budget ceiling you are willing to lift to, who answers the phone during the surge, and the point at which you pause because the calendar is full. Reacting on the day usually means the budget is gone before noon and the calls go unanswered.
- Can you target only homeowners?
- Not directly and not reliably. What you can do is exclude renters-heavy intent, exclude employment and DIY terms, ask the qualifying question on the form, and use the landing page copy to set expectations about the kind of work you take.
- Should the ads mention insurance claims?
- Keep claim language out of the ad copy and off the landing page beyond describing your own process. Do not promise coverage or an outcome. Tell homeowners to confirm their policy with their insurer or their own counsel, and keep the offer focused on the inspection.
