Two auctions in one county: the single sliding door and the building that buys forty
Retail demand and building demand should never share a campaign, a budget or a target, because one qualified inquiry from each is worth a completely different amount to you.
Split the account at the top. One side chases homeowner intent across single family stock in areas like Doral, Kendall and Hialeah. The other chases association and management intent, which is thin in volume and heavy in value.
On the retail side, structure by product line and by job size. Whole house replacement, entry and patio doors, and single opening work behave differently and should not compete for the same budget. Whole house work is generally where the margin is, so let it take priority when the daily budget is tight.
On the building side, expect very low search volume. That is not a failure signal. A handful of clicks a week from a management company is a legitimate outcome, and it means exact match, tight ad groups and manual oversight rather than broad match with an automated bid strategy chasing volume.
Feed both sides signed contract data rather than form fills. Importing offline conversions is the difference between a bidding system optimizing toward cheap inquiries and one optimizing toward jobs you actually sold.
Your negative keyword list starts with car tint and ends with screen repair
In this county the word window pulls in an enormous amount of traffic that has nothing to do with replacing an opening, and unpoliced broad match will find all of it.
Automotive window tinting is a large market here and it competes for the same words. Add tint, tinting, ceramic tint, car, auto, vehicle and windshield early. Add the operating system too, because searches for computer windows leak into badly built accounts constantly.
Then exclude the service work you do not want. Screen repair, screen replacement, glass repair, broken pane, foggy glass, seal repair, track repair and roller replacement all attract callers with a small urgent problem and no replacement budget. If you genuinely sell service work, give it its own campaign with its own economics rather than letting it consume replacement budget.
Exclude the research and cost tail if your budget is small. Terms around free, cheap, DIY, grants, careers, jobs and salary rarely become contracts. Terms around cost and price sometimes do, so test rather than assume.
Review the search terms report weekly for the first stretch and monthly after that. In a market this crowded, the negative list is a permanent job, not a setup task.
Bid on the buildings, not the borders, because geography in Miami-Dade is vertical
A single tower can hold more households than several residential blocks, so location targeting drawn as neat municipal shapes misses how demand is actually stacked here.
Radius targeting around dense residential corridors in Brickell, Edgewater and along the Miami Beach barrier island reaches a concentration of potential openings that a broad county wide setting will dilute. Build those as their own geographic segments with their own bid adjustments.
Set adjustments on the drive as well as the demand. A measure in Hialeah and a measure in Miami Beach do not cost your team the same amount of time, and if your crews are stretched, bidding harder where you can service efficiently protects margin better than chasing county wide coverage.
Exclude what you will not travel to. Paying for clicks from areas you decline to service, or from renters in buildings where the association controls the glazing, is a quiet and constant waste.
Use location of interest carefully. Someone searching from out of state about a Miami condominium may be an owner planning work remotely, which is a real buyer in this market, so do not blanket exclude people physically outside the county without checking the data first.
Spanish campaigns are a separate auction that fails on the phone, not in the ad
Running Spanish creative is straightforward, and every failure we would expect to see happens after the click rather than before it.
Build Spanish campaigns from native keyword research, not from a translated copy of the English set. The phrasing people use for impact windows and doors in Spanish does not map word for word, and the auction is generally less crowded than the English head terms, which is exactly why it deserves its own budget line.
Point them at Spanish landing pages with matching URLs. Sending Spanish ad traffic to an English page, or to a page with a translate widget, tends to produce clicks with no inquiries and looks like a keyword problem when it is a landing page problem.
Staff it before you scale it. If a caller reaches an English only phone, you have paid for a lead and handed it to the next dealer. Call handling, follow up messages and the quote document all need to work in the caller's language.
Measure the two languages separately. Blended reporting hides both the cheaper acquisition on one side and the intake breakdown on the other.
Professional-network and paid social targeting reach the community association manager
The person who can put you in front of a board rarely runs a search for a window contractor, so search alone cannot reach the highest value buyer in this county.
Professional network targeting by job title and industry reaches community association managers, property managers, building engineers and board members in a way keyword targeting cannot. Volume is small and cost per click is generally higher than search. Judge it on conversations opened with buildings, not on click cost.
The creative has to match that reader. A packet of documentation, a short explanation of how you sequence an occupied building, and an offer to walk a property are more persuasive than a discount. Save promotional creative for the homeowner side.
Paid social on the retail side does a different job. It reaches the owner who has been putting the work off, using install footage and honest before and after material rather than a countdown offer. Check how each platform classifies your ads before you build audiences, since some categories carry targeting restrictions, and confirm the current rules with the platform.
Retarget separately by side of the business. A board member who read your association page and a homeowner who abandoned a quote form should not receive the same message.
Cost per signed opening, and a spend curve shaped by board meetings rather than months
Cost per lead flatters a channel that produces cheap inquiries, and evenly spread monthly budgets ignore how decisions here are actually timed.
Track cost per booked measure, cost per signed contract and cost per opening sold. The last one is the honest number in a trade where one approval can be dozens of units and another job is a single door. Two campaigns with identical cost per lead can be a factor apart on cost per opening.
Boards typically meet on a monthly cycle and move through proposals, engineer input and votes over a longer period. Building side spend should be steady and patient rather than pushed hard in short bursts, and the follow up sequence has to survive several months without going cold.
There is a related timing signal worth understanding. Buildings in Miami-Dade become subject to the county's 40 year recertification program at 40 years and every 10 years afterward under Section 8-11(f) of the county code, which puts boards in front of engineers and capital planning. That report concerns structural and electrical safety, so never market as though you perform it, but it is a period when envelope budgets get discussed. Worth confirming details with the county before you reference it publicly.
On the retail side, pace against your install and permit throughput. Buying more measures than your crews can convert into scheduled work creates a backlog, a slower response and a lower close rate, which raises your real cost per contract even while the reported cost per lead looks fine.
Questions we actually get
- What should a Miami window dealer expect to pay per click?
- We will not quote a figure, because a credible one comes from your own account and your own geography, not from an industry average. What we can say is that this is the most competitive of the three South Florida metros and the head terms attract national operators, so the useful question is not the click price but how many clicks it takes to reach a signed contract in each campaign.
- Should paid search and paid social run at the same time on a limited budget?
- If the budget is genuinely tight, start with high intent search on the retail side, because it harvests people already shopping. Add the building side targeting once you can afford to wait a quarter for results, since that channel builds a pipeline rather than filling next week's calendar.
- Do I need a separate Spanish campaign, or can I just add Spanish keywords?
- Separate campaign, separate landing pages, separate reporting. Mixing languages in one campaign makes the budget uncontrollable and hides which side is performing. It also tempts you to send Spanish clicks to an English page, which is the most common way this spend gets wasted.
- How do I stop paying for service calls when I only want replacement work?
- A disciplined negative list is most of it: repair, broken, foggy, seal, screen, track, roller, single pane. Ad copy that says whole home or full building replacement helps filter as well, and your intake script should route genuine service inquiries out quickly rather than spending a morning on them.
- How should I attribute a condominium job that took five months to close?
- Record the building name and the management company on the first inquiry and keep the record at building level rather than lead level. Import the signed contract back into the ad platform as an offline conversion when it closes. Attribution will still be imperfect over that timescale, and it is better to accept that openly than to pretend a last click model is telling you the truth.
