Bid Staten Island on its own, because a crew crossing the Verrazzano loses a morning
Geography here is priced in travel hours and tolls, so one metro radius will always overpay for the jobs furthest from your yard.
Draw the account by borough, not by a circle around your shop. Manhattan, Brooklyn, Queens, the Bronx and Staten Island are separate markets with separate travel costs, separate building stock and separate competitors. Nassau and Westchester deserve their own line again.
Set bids by what the job costs you to reach. A two opening measure in Astoria and the same measure in Riverdale consume very different amounts of a crew's day. Paying the same click price for both quietly funds the worse of the two.
Watch the location settings. Default targeting sells impressions to people merely interested in an area, which in a city this dense means paying for renters browsing, out of state owners and students. Restrict to presence in the targeted area and review the location report monthly.
If you genuinely will not cross a river for a small job, exclude the area rather than bidding it low. Low bids still buy clicks at three in the afternoon when the auction is thin.
Co-op and condo work deserves its own campaign, not a line item in the homeowner one
Building work and apartment work are two businesses with different buyers, different copy and different close cycles.
A managing agent and a shareholder respond to opposite arguments. The agent needs insurance limits, occupied unit staging, protection of common areas and a schedule that survives freight elevator booking. The shareholder needs to know how long their bedroom is open to the weather.
Split them at the campaign level so budgets do not compete. Building side search terms are lower volume and higher value, and they will lose every internal auction against homeowner terms if you leave them in one bucket.
Write ad copy that names the process. Alteration agreement experience, certificates of insurance, phased work in occupied buildings. Generic efficiency language reads like a national brand and gets ignored by anyone who has managed a building.
Expect a long gap between click and contract on the building side. Approval calendars are not yours. Report those campaigns on a rolling quarter or the monthly view will make you kill something that is working.
Half the "window repair Brooklyn" clicks belong to a landlord you cannot bill
Repair intent, glass only work and screen jobs will consume a replacement budget before Thursday if you let them.
Old stock generates repair demand at scale. Broken pane, sash cord, screen replacement, window will not lock, glass replacement near me. Some of that is a small paid job you actually want. Most of it is a superintendent looking for a fifty dollar fix.
Build the negative list from your own search terms report, not from a template. Start with repair, glass only, screen, storm window repair, parts, DIY sizing, plus terms for auto glass and storefront work, which show up constantly in a city where every ground floor is a business.
Add the buyer types you do not serve. New construction, general contractor bids, rental turnover, tenant complaints. Each one looks like your keyword and closes at a rate that will not pay for the click.
Review the report weekly for the first month, then monthly. Search terms drift, and in this market they drift toward commercial and rental work you did not intend to buy.
Landmarked block, prewar walk-up and new tower are three bids hiding behind one keyword
The same phrase means three different jobs depending on the building, and only one of them is a straightforward retrofit.
A window replacement in a new Long Island City condo is a scheduled elevator delivery and a clean opening. The same job in a six story Bushwick walk-up is a stair carry, a sidewalk permit conversation and a longer day. On a landmarked block the specification question comes first and is worth confirming with the local building department.
Ad groups should reflect that. Keywords, copy and landing pages built for prewar and brownstone work, separate ones for elevator buildings, separate ones for one and two family homes in the outer boroughs.
Copy that names the building type outperforms copy that names the manufacturer. An owner in a walk-up wants to know you have carried units up five flights before, not which brand you carry.
Send each ad group to a page that continues its argument. Dropping a brownstone searcher on a generic quote form wastes the money you spent teaching them you understood the building.
Set your daily cap by freight elevator availability, not by the weather forecast
Spend should track what your crews can actually install, and installation capacity here is gated by building access.
Demand in this trade climbs with cold weather, and so does everyone else's bidding. Spending harder in a week when your install calendar is already full buys quotes that go stale before you can schedule them.
Pace against the install calendar. If freight access in the buildings you have signed pushes work three weeks out, keep the account steady rather than chasing a spike you cannot serve.
Watch the daypart data before trusting instinct. Building side inquiries tend to arrive during business hours from an office. Homeowner inquiries arrive at night from a phone. Those are two different clocks and they can be budgeted separately.
Keep a reserve. Money held back for the weeks when a competitor pauses is worth more than the same money spent evenly across a month.
Price the click in stair-carry hours and cost per booked measure, not per lead
Cost per lead in this city flatters an account that is buying the wrong jobs.
Track the chain: click, inquiry, qualified inquiry, booked measure, held measure, signed contract. Most window accounts stop at inquiry, which is the point where the numbers look best and mean least.
A booked measure in Manhattan with no elevator and no loading zone is not the same asset as a booked measure at a Forest Hills co-op with freight scheduled. Tag inquiries by building type at intake so the ad account eventually learns which ones close.
Feed conversions back. Import the booked measure, and the signed contract where your CRM supports it, so bidding optimizes toward jobs instead of form fills. Without that, the platform will happily find you the cheapest possible clicks in the cheapest possible boroughs.
Set the target from your own math: average job value, close rate on measures, and what a crew day costs when the job includes carrying units up four flights. Every published benchmark you will find was calculated somewhere with a driveway.
Questions we actually get
- How should we split budget between Manhattan and the outer boroughs?
- Split by where you can profitably work rather than by population. Price each borough against the crew time a job there consumes, including travel, parking and stair carry, then fund the ones that clear your margin. Many dealers find one or two boroughs carry the account and the rest are worth a small, capped presence for larger jobs only.
- Are Local Services Ads worth running alongside search?
- They are worth testing, and they compete for the same phone as your search ads, so read them together rather than in separate reports. Eligibility and screening requirements vary and are worth confirming directly with the platform and with your insurer regarding coverage documentation.
- What negative keywords matter most for a New York window and door dealer?
- Repair and glass only terms first, since old stock generates them constantly. Then screen work, storm window repair and parts. Then auto glass and storefront terms, which appear more here than in most markets. Then buyer types you do not serve: new construction, contractor bids, rental turnover and tenant complaints. Build the list from your own search terms report weekly at first.
- Should paid social run alongside search for this trade?
- It can work for the homeowner side, where the decision is postponed rather than urgent, and for building side awareness among managing agents. Treat it as demand creation with a longer measurement window, keep it on its own budget, and do not judge it on last click attribution against search.
- How do we account for co-op approval delays in reporting?
- Report building side campaigns on a rolling quarter and track the stage each opportunity sits in, including waiting on board approval. A monthly cost per contract view will look terrible on work that is progressing normally, and cutting a campaign on that view is the most common mistake dealers make here.
