Beltway drive time, not a mileage radius, should draw your targeting map
A circle on a map ignores the river, the bridges and an hour of morning traffic.
Draw a twenty mile radius from a shop in Tysons and it swallows half the District, which your truck may or may not reach by eight in the morning depending on the day. Draw the same circle from Capitol Hill and it collects parts of Prince George's County you may not want.
Build targeting from explicit place lists instead. Name the cities, counties and ZIP codes you actually serve and exclude the rest. It takes longer to set up and it stops the single largest source of waste in a remodeling account here.
Check the location option that decides whether you reach people present in an area or merely interested in it. The default catches researchers who are nowhere near the house, which matters in a metro full of people planning moves from elsewhere.
Once you have close rate by area, adjust bids toward the places where jobs actually sign. Click volume from a dense area is not the same as revenue from it.
Give the District, suburban Maryland and Northern Virginia their own budgets
Three jurisdictions with separate licensing, permitting and taxation should not share one campaign and one reporting line.
A blended metro campaign hides which market is paying for itself. Separate campaigns for the District, for Montgomery County and its neighbors, and for Arlington, Alexandria and Fairfax County make the differences visible within a few weeks.
Separation also protects you operationally. If your registrations cover some jurisdictions and not others, and requirements do generally differ across the three, a blended campaign will generate inquiries you cannot legally or practically serve. Confirm what applies to you with each jurisdiction and your own counsel, then target accordingly.
Average project value and competitive pressure vary across the metro. Splitting the account lets you pay more where the jobs are bigger without dragging the whole budget up.
Practical benefit that owners appreciate most: when the spring calendar fills, you can pause one market and keep another running rather than blunting everything at once.
Federal and commercial fit-out clicks look like yours and never are
This metro produces a constant stream of office, agency and contractor renovation searches that will quietly drain a residential budget.
Terms like office renovation, tenant fit out, solicitation, RFP and government contract sit close enough to your vocabulary that broad match will find them. So will commercial kitchen searches from restaurant operators who need hood work you do not do.
Add the usual residential leaks on top: cabinet refacing, appliance repair, flat pack cabinet assembly, countertop shopping, and every how to question from somebody planning a weekend project.
Pull the search terms report weekly for the first two months and daily during the first week. The negative keyword list is where the account gets built, and it is never finished.
Multifamily turn work belongs on the list too unless you want it. Property management companies in Silver Spring and Navy Yard search in language that overlaps yours, and their unit turns price nothing like a custom kitchen.
The ad should name the jurisdiction, because half the doubt is whether you serve it
A homeowner in Falls Church reading a metro-wide ad has to guess whether you cross the river.
Put the jurisdiction in the headline and repeat it on the landing page. Somebody searching from Bethesda who sees Bethesda in the ad is more likely to click and far more likely to be a real inquiry.
Say where you do not work as well. A line that rules out an area costs you a click and saves an hour of unpaid consultation. In this metro, that trade is worth making every time.
Use location assets and call assets so the ad carries a phone number and a service area without the searcher hunting for it. A phone call at ten in the morning is worth more than a form fill at midnight.
Keep credential language careful. Stating that you are licensed and insured for the jurisdictions you serve is fine. Implying blanket coverage of a metro with three separate licensing regimes is not, and it is worth confirming your own status before it goes in an ad.
A snow week empties the phones, so carry budget into the thaw
Real winters produce weeks when nobody books a measure and the clicks still cost full price.
Snow response, a cold snap and the holiday stretch all suppress booking behavior in the same way. Automated bidding does not know that and will keep buying clicks from people who are not going to answer the phone back.
Pull spend down during a storm week and push it back after. The demand does not vanish, it shifts by a few days, and the accounts that hold budget for the thaw buy the same jobs cheaper.
Pace against your production calendar rather than the month. If the crew is committed through April, spending March budget at full tilt buys inquiries you will annoy by making them wait.
Keep a reserve for the holes. Every remodeler gets a cancellation. A campaign you can turn up on two days notice is worth more than one running flat all year.
Tag every closed job by jurisdiction before you judge a campaign
Cost per lead flatters whichever market produces the cheapest useless inquiries.
Record source, jurisdiction, project type, whether the site visit happened and whether an agreement was signed. Without those five fields, the account is being managed on a number that does not pay anyone.
Feed the outcomes back. Importing offline conversions so the platform learns from booked jobs rather than form fills changes what it goes looking for, though it needs enough volume to be useful.
Judge slowly. A kitchen is a considered purchase with a long cycle, and a handful of closed jobs is not a sample. Making structural changes to an account every two weeks generally destroys more than it fixes.
No one can promise you a cost per booked job in advance. What can be promised is that the number will be visible, tracked consistently, and reviewed against actual signed work rather than against a dashboard of clicks.
Questions we actually get
- How should we split the budget across the District, Maryland and Virginia?
- Start roughly even across the markets you genuinely serve, then reallocate on closed jobs after you have enough of them to mean something. Splitting by jurisdiction from day one is the important part. The exact ratio should come from your own data, not from a benchmark, since job values and competition differ across the metro.
- Is Google enough, or do we need Meta too?
- Search captures people already looking, which is where a remodeler should start. Meta can work for a firm with genuinely strong finished project photography, since a kitchen is a visual purchase and the metro has plenty of homeowners who have not started searching yet. Treat it as a second phase, funded once search is producing booked work.
- What negative keywords matter most in this market?
- Commercial and federal fit-out language is the local specialty: office renovation, tenant fit out, RFP, solicitation, contract award. Beyond that, the usual residential leaks apply, including refacing, appliance repair, countertop retail, flat pack assembly and every how to query. Review the search terms report weekly rather than trusting a starter list.
- Should ads mention licensing?
- Only what you can support. Requirements differ across the District, Montgomery County and Northern Virginia, so a general claim about metro-wide licensing is risky. Naming the jurisdictions you are registered in is stronger anyway, and worth confirming with each jurisdiction before it goes live.
- How long before we can tell whether paid search is working?
- Long enough to accumulate real closed jobs, which for a considered renovation purchase means several months rather than several weeks. Early signals worth watching are search term quality, call quality and how many inquiries come from jurisdictions you actually serve. Cost per booked job only stabilizes once there are enough booked jobs to average.
