Bid by borough, because a Manhattan click and a Staten Island click are different products
A single campaign covering the metro averages away the only distinction that matters to your margins.
Prices per click, competition and job size all change across the city. Manhattan attracts the heaviest bidding and often the largest projects. Queens and the outer boroughs are cheaper to reach, and the work skews toward two family houses and smaller co-ops. Nassau and Westchester behave like separate markets entirely.
Split them so the numbers stay readable. Separate campaigns, or at minimum separate ad groups with their own budgets, let you see that a Bay Ridge lead cost a fraction of a Chelsea lead and closed at a similar rate. Mixed together, that never surfaces.
Target the areas your crews can genuinely serve, and be honest about what serving means without a truck that can park. If a job in Riverdale means an hour each way plus a parking problem for a two-person crew, price that into whether you bid there at all.
Exclude what you will not travel to rather than hoping the platform reads your mind. Location targeting quietly includes people who merely show interest in an area, so check the setting that limits delivery to people actually in the places you chose.
A large share of the clicks in this city come from renters who will never own the wall
The single biggest source of waste in a New York remodeling account is traffic from people who do not control the apartment.
Renters, roommates hunting a sublet, tenants who want a landlord to fix a bathroom and supers pricing a repair all type phrases that look close to yours. So do students, job seekers and people looking for an apartment rather than a renovation.
Build the exclusion list before the first ad runs, then keep building it weekly from the search terms report. Words like rental, apartments for rent, landlord, super, cheap, DIY, jobs, hiring, salary, refacing and repair belong on it. So do the names of retailers and big box installers if you are not competing for that work.
Write the ads to repel as much as attract. Naming ownership, a scope and a price posture in the copy costs you clicks you did not want. Cheaper traffic is not the goal. Fewer wrong people is.
Review the terms report on a fixed day every week for the first two months. The list you build in that period does more for cost per job than any bid adjustment you will make all year.
Split the account by what the building will allow, not by room
A gut renovation in an elevator co-op and a bathroom refresh in a walk-up are different sales with different economics, so give them different campaigns.
Grouping by kitchen and bath is the obvious cut and the least useful one here. The real division runs along what the building permits and what access costs. A full apartment gut with an alteration agreement behind it is a long sale with a big number. A single bathroom in a two family house in Queens is a short sale with a small one.
Give each its own campaign, its own budget and its own landing page. Ad copy that speaks to board paperwork and freight scheduling will confuse someone who just wants a tub replaced, and copy about a quick bathroom refresh will lose the owner planning a whole floor.
Match the keywords to the sale. Gut renovation, apartment renovation and combining units belong in one place. Bathroom remodel, tub to shower and vanity replacement belong somewhere else, if you want that work at all.
Some shops should not bid on the small work. Run the math on what a two week bathroom actually earns after a stair carry and an afternoon lost to parking, and be willing to turn the campaign off.
New Yorkers search for apartment renovation, and your ads probably say home
Vocabulary is a targeting tool, and the words in your account were probably written for a market where everyone has a driveway.
Buyers here say apartment, unit, line, floor and building. They say co-op and condo. They say prewar. They rarely say home improvement, and almost nobody says single family.
Pull the search terms report and read the language coming in before rewriting a single keyword. The phrases people actually type will tell you which side of that vocabulary your buyers sit on, borough by borough.
Then mirror it. An ad headline that says apartment renovation in Manhattan gets read by a different person than one that says home remodeling services. Both may be true about your business. Only one sounds like it was written by somebody who works here.
The same applies to your qualifiers. Insured, licensed and experienced with building approvals carry weight in a market where an owner has been warned by three neighbors about a contractor who could not clear the process. Say it in the ad, then prove it on the page.
Pace the budget against a calendar that a board controls
Spending in a straight line makes no sense when a large share of your sold work waits on approval before it can start.
The gap between a signed proposal and a start date is real here, and it is not fully in your hands. Boards meet when they meet. Insurance certificates get rejected and reissued. Freight elevator time gets allocated. Money spent today can turn into work that begins a season later.
Plan spend against crew capacity months ahead rather than against this month's phone volume. A quiet week does not mean the account failed, and a busy week is often the worst time to raise budgets, because you are buying jobs you cannot schedule.
Build a simple rule: when approved and scheduled work reaches a set number of weeks, cut spend on the campaigns that produce the longest sales, not the shortest. Turning off the small bathroom campaign first is usually backward.
Keep some budget running through slow stretches. Owners plan renovations well before they act, and disappearing from search for a quarter costs you the pipeline that would have filled the next one.
Count the walkthroughs you were actually let into
Form fills flatter an account, and in this city a striking share of them never become a visit to the apartment.
A lead here can fail for reasons that have nothing to do with interest. The person is a renter. The building will not permit the scope. Nobody can meet you during the hours the building allows work. The apartment is a sublet and the owner lives in another country.
Track the stages that survive those failures. Inquiry, walkthrough booked, walkthrough held, proposal issued, job started. Cost per walkthrough held is the first number that means anything, and cost per started job is the one that pays the crew.
Feed real outcomes back into the platform. Importing signed values, or at minimum a qualified lead event, gives automated bidding something worth optimizing toward. Fed only form fills, it will faithfully buy you more of the wrong ones.
Give each campaign a fair window before judging it. With small monthly job counts, a single closed project can swing a month's numbers, so read three months together and look at the direction rather than the last thirty days.
Questions we actually get
- What should we spend to test paid search in New York?
- There is no honest universal number, and any agency quoting one has not looked at your market. Set a budget that can buy enough clicks in a single borough to produce a readable pattern within a quarter, rather than spreading a small amount across the whole metro. Testing one borough properly beats testing five badly.
- Are lead marketplaces worth using alongside our own campaigns?
- They fill a calendar quickly and they sell the same inquiry to several contractors, which sets the conversation on price. Some shops use them to keep a crew busy in a slow stretch. If you do, track those leads separately, because mixing their close rate into your own numbers will hide how your real campaigns are performing.
- Should we run ads on Meta as well as Google?
- Search catches people already looking. Social is better at showing finished work to owners who have not started yet, which suits a considered purchase like a kitchen. Treat them as different jobs and do not expect the same cost per inquiry from both.
- How do we stop paying for clicks from renters?
- Exclusions and ad copy, reviewed weekly. Add renter and repair language to the negative keyword list, name ownership and scope in the ad text, and put a qualifying question near the top of the form. Expect the list to keep growing for the first couple of months.
- Can you guarantee a cost per lead or a number of jobs?
- No, and we will not. Auction prices, competitor budgets and board timelines are outside anyone's control. What we commit to is structure, exclusions, tracking that follows an inquiry through to a started job, and honest reporting when something is not working.
