Draw the map at the state line first and the Watterson second
A mileage radius around your office is the wrong shape for a market split by a river and a state line.
Radius targeting drawn around a pin in Jefferson County spills into Clark and Floyd County without asking. If you are not set up to work in Indiana, that is money leaving the account every day for inquiries you will decline.
Build two geographic groups from the start: the Kentucky side and the Indiana side, each with its own budget, its own ads and its own landing page. You can then turn one off in a minute, and the reporting tells you what each side actually costs.
Inside Kentucky, think in drive time. Work inside the Watterson is a different day than work out past the Gene Snyder, and Oldham or Bullitt County is different again. Set bid adjustments by area based on where jobs have been profitable, not on where you wish they were.
Use presence-based targeting rather than the setting that also includes people merely showing interest in the area. Otherwise you buy clicks from people researching a remodel in a city they are moving away from.
Countertop yards and cabinet showrooms are bidding on your click, and you should let them
A large share of remodeling search traffic wants a price on a product, not a contractor, and those clicks cost you the same as a real one.
Searches about quartz prices, cabinet refacing, vanity units and appliance fitting attract shoppers. Retailers and product sellers can afford that traffic because they sell the product. You cannot, because you sell a project that takes weeks.
Build the negative list before you build the campaign. Refacing, if you do not do it. Repair, parts, used, cheap, wholesale, DIY and how to. Careers, hiring and jobs. Apartment and commercial terms if you only do residential. Then read the search terms report every week for the first few months and keep pruning.
Local color creates its own noise here. A metro with a busy restaurant scene in NuLu and Butchertown produces commercial kitchen queries that look close enough to yours to trigger a broad match ad. Exclude them by hand.
Stay on phrase and exact match for core terms while the account is young. Broad match with a thin conversion history will spend a month teaching you things your negative list already knew.
Bid up the hours when second-shift households are awake
Air freight and logistics put a large part of this metro on schedules that do not match a nine to five office, and most remodeling ad accounts are scheduled as if they do.
Ad schedules are usually copied from the front door hours without anyone checking. In a market with heavy shift work, some of your best inquiries arrive at hours your competitors have switched off.
Once the account has run long enough to have real data, pull the hour of day and day of week report. Look at conversions, not clicks. Then adjust rather than guess.
Fix the answering before you raise the bids. Ads that run at six in the morning without anybody to reply just buy you a voicemail. If you cannot staff those hours, route them to a form or a text with a stated callback window you can keep.
Resist cutting hours too early. A few weeks of thin data will make a perfectly good Tuesday morning look dead. Give it a season.
A form fill is not a job, and in this metro some of them are the wrong state
Cost per lead flatters an account that is quietly buying inquiries you have to turn down.
Track the whole path: click, inquiry, qualified, booked measure, signed job. If the platform can accept offline conversion imports from your CRM, feed it the booked measures rather than the form fills.
Tag every inquiry with state, county and town the moment it lands. Without that tag your campaign reports cannot tell you that the Indiana budget is producing inquiries you cannot serve, or that Bullitt County looks cheap because half of it never gets booked.
Price the account on cost per booked measure. A kitchen is a considered purchase and the lag between click and contract is long, so judge campaigns in quarters, not weeks.
Nobody can promise a lead volume or a cost. What you can do is establish what a booked measure is worth to you based on your own close rate and average job, then decide what you are willing to pay for one.
Send Clarksville clicks to an Indiana page with an Indiana phone number
The landing page should settle in its first line the exact question that made the click uncertain.
One generic remodeling landing page waters down both sides of the metro. A homeowner in Clarksville is trying to find out whether you cross the river, and a headline naming Louisville tells them you probably do not.
Give the Indiana campaigns a page that names Southern Indiana towns, shows work finished there, and states your coverage plainly. Add a sentence noting that requirements generally differ by state, that you handle whichever applies, and that specifics are worth confirming with the local building department.
Use a separate tracking number for each side. It gives you clean attribution and it lets whoever answers know which state the caller is in before the conversation starts.
Put the town or state field at the top of the form, not buried under budget and timeline. It is the field that decides whether the inquiry is worth a truck.
A small budget buys one county, not all of Kentuckiana
Spread across a two-state metro, a modest monthly spend produces data too thin to learn anything from and too thin for automated bidding to work with.
Pick the tightest geography with the best margin and run there until it is reliably profitable. That might be a handful of Metro neighborhoods, or it might be one outlying county. Expand from a working account, not from an ambitious map.
Narrow the service line as well. A campaign trying to sell full kitchen remodels, bathroom conversions and small updates on one budget teaches you nothing about any of them.
Automated bid strategies want conversion volume. With a handful of conversions a month, tight targeting and a simpler strategy generally behave better while you build history, though it depends on your account and is worth revisiting once volume grows.
Pace against the calendar, not the month. When the crews are booked into the spring, pull spend back rather than stopping, because a dark account loses its history and takes time to rebuild.
Questions we actually get
- Should we advertise on the Indiana side at all if most of our work is in Jefferson County?
- Only if you are actually set up to work there. Registration, tax and insurance treatment generally differ by state, so confirm with your own counsel and the relevant authorities first. If you are set up, run it as a separate campaign with its own budget and landing page so you can measure it honestly and shut it off cleanly.
- What should we count as a conversion?
- Count the booked in-home measure, not the form fill. Form fills include the wrong state, the wrong scope and people pricing a countertop. If your CRM can send booked measures back to the ad platform as offline conversions, do that, because it changes what the bidding optimizes toward.
- Is Google Ads or Meta better for kitchen and bath work?
- They do different jobs. Search catches people who have already decided to do something. Meta is better suited to finished project photography reaching homeowners who have not started looking. Most remodelers should get search working first, since the intent is already there, and treat social as a second phase.
- How much of the budget goes to negative keywords work?
- It is time rather than money, and it is the highest-return time in the account. Read the search terms report weekly for the first few months and monthly after that. Product shoppers, DIY questions and job seekers will find your ads without help.
- Our phones are quiet in January. Should we stop spending?
- Going dark usually costs more than it saves, because the account loses its recent history and takes time to come back. Reducing spend and shifting the message toward planning for a spring start is generally the better move, since interior work sells well in a cold month and homeowners here are already thinking about it.
