A radius around your yard buys clicks across water your trucks avoid
Radius targeting assumes distance and driving are the same thing, and in Tampa Bay they are not.
Draw a twenty-five mile circle around a yard in Brandon and it swallows Clearwater, Largo and Palm Harbor. Those addresses are reached by the Howard Frankland, the Gandy or the Courtney Campbell, and on a bad afternoon a trip that looks like twenty minutes is not.
Replace the circle with a location list. Target counties, cities and ZIP codes you have decided to serve, and remove the ones you have not. Precision here is worth more than reach because every unqualified click is paid for at construction keyword prices.
Set location options to people in your targeted places rather than people merely interested in them. Otherwise you pay for someone in another state reading about Tampa pools with no address here at all.
Then check the geographic report monthly. Cities you never intended to buy show up constantly, and a five minute exclusion is usually the highest return edit in the account.
Split the account at the Howard Frankland, not at the metro line
Hillsborough and Pinellas should be separate campaigns with separate budgets, because they will not perform alike for any one company.
Run the same keyword sets in two campaigns, one per county, with their own daily budgets, bids and ad copy. Copy that names St. Petersburg or Clearwater outperforms copy that says Tampa Bay to a searcher standing in Pinellas.
Separation is mostly an accounting decision. Pooled into one campaign, whichever side clicks cheaper absorbs the budget, and you never learn that the other side produces better jobs, or worse ones.
Set bids against real cost to serve. If a Pinellas job costs your estimator a half day of driving, either bid it lower, raise the minimum project size in the copy, or both. Equal bids across unequal drive times is a slow leak.
If you are new to the other county, treat it as a deliberate test with a fixed budget, a stop date and a decision rule written down before you start.
The phrase pool home belongs to real estate listings and will drain a build budget
In a market where a pool is a headline feature on listings, real estate search is the most expensive junk in a pool construction account.
Buyers across St. Petersburg and the Hillsborough suburbs search for homes with pools, pool homes for sale, and variations on both. Broad match will find every one of them and charge you construction rates for a click from someone shopping for a house.
Build the full negative list before you turn anything on: for sale, real estate, realtor, listings, rent, rental, cleaning, service, weekly, chemicals, above ground, inflatable, hot tub, spa parts, repair if you do not do repairs, jobs, hiring, salary, public pool, swim lessons, and pool table.
Read the search terms report weekly for the first six weeks, then monthly forever. Add negatives at both ad group and account level so one bad theme does not reappear in a new campaign.
Favor phrase and exact match for construction terms. Broad match is only defensible once you have a mature negative list and a bidding strategy with enough conversion history to steer it.
Bid down the hours when a Clearwater job is a two hour round trip
Ad schedules should reflect when you can answer the phone and when your estimator can realistically get there.
If the office starts early and your estimator is out by seven, a morning inquiry is worth more than an evening one. Set the schedule to match the hours a human answers, and lower bids for the hours nobody does.
Cross-bay timing deserves its own adjustment. A Pinellas inquiry arriving in the late afternoon usually cannot be seen the same week without wrecking a route, so either bid that window down or send those inquiries into a booking flow instead of a phone call.
Call-only ads and call extensions are pure waste outside answering hours. Nothing burns money faster than paying for a call that rings out, and the caller is on to the next builder in minutes.
Weekends need a decision, not a default. Homeowners browse on Sunday. Either staff it, or capture the inquiry with a form and a same-day automated reply that sets a real callback time.
Reaching the family that just closed on a house near MacDill
Search only finds people already looking, and a large share of this metro's best pool prospects have not started looking yet.
Relocation is constant here, and newcomers arrive with no contractor relationships at all. They are not typing pool builder yet. They are still unpacking, and they are reachable on Meta and YouTube long before they reach Google.
Target recent movers and newly sold areas where the platform allows it, and write to their actual situation: the first year in a house with a pool, what an aging shell means, what a renovation looks like from the owner's side.
YouTube suits this trade unusually well because excavation and shell work is genuinely watchable. Keep the ask small at that stage, a guide or a callback rather than a full quote request.
Fund it out of a separate budget line and judge it on a longer window. Left in the same pot, search will always look better on a monthly report and this work will get cut before it can pay.
Two counties produce two cost per job numbers, and the average hides money
A blended cost per lead across Tampa Bay tells you nothing about whether either side of the water is worth funding.
Track cost per booked site visit and cost per signed agreement, split by county. One side is usually subsidizing the other, and the blended figure is exactly the number that keeps that hidden.
Getting there requires discipline in intake. The person answering the phone has to record the city and county at first contact, and those values have to reach the same place your ad spend does, whether through offline conversion import or a clean CRM tag.
Cost per lead is close to meaningless in this trade. A lead can be a price shopper with no yard access or a full build with a spa and a deck, and averaging them produces a number that cannot inform a bid.
Judge over enough jobs to mean something rather than week to week. Volume here is small, monthly swings are mostly noise, and no honest account manager will promise a lead count or a cost.
Questions we actually get
- How much should a pool builder budget for Google Ads here?
- Enough to buy meaningful data in each county you run, which usually means funding one county properly before adding the second. Splitting a small budget across Hillsborough and Pinellas produces two campaigns with too little volume to judge. Set the number from your capacity and average project value, not from an industry benchmark.
- Should I bid on renovation terms or new construction terms?
- Separate them into different campaigns because the economics differ. Renovation and resurfacing volume is higher and the average job smaller. New build volume is thinner and the ticket larger. In one campaign the cheaper clicks will absorb the budget and you will never see the trade you actually made.
- Is Local Services Ads worth it for pool construction?
- Worth testing where the category is available to you, and worth verifying the eligibility rules for your license type before spending time on it. Treat it as a separate channel with its own cost per booked visit rather than as a replacement for search, and confirm current requirements directly with the platform.
- Why is my cost per lead low but my sales flat?
- Almost always a mix problem. Cheap clicks come from repair, service and real estate vocabulary, and those leads are not construction jobs. Look at cost per booked site visit instead, split by county, and you will usually find the cheap half of the account is producing nothing you can build.
- Can you promise a certain number of leads per month?
- No, and nobody should. Auction prices, competitor behavior and seasonal demand all move. What can be committed to is structure, exclusions, honest measurement by county, and a monthly reading of what the spend actually produced in booked visits.
