Split the account at the Perimeter before you split it by pool type
Geography is the first budget decision in this metro because it decides what a job costs you to deliver, not just to win.
Most pool accounts start life as one campaign with a radius drawn around the yard. A radius treats a Dunwoody homeowner and a far Gwinnett homeowner as equivalent, which they are not once trucks, crews and a full day of I-285 are involved.
Build separate campaigns by drive time zone instead. A core zone you serve every day, a secondary zone you serve on set days, and an outer zone you either bid low on or exclude outright. Each gets its own budget and its own bids, which is the only way to pay more for the work you actually want.
Set targeting to people present in a location rather than people merely interested in it, and add explicit exclusions for the areas you will not drive to. Otherwise the platform will happily spend your money reaching someone shopping for a pool two counties past your limit.
The vocabulary is a bonus. Ad copy and landing pages that say inside the Perimeter, or name Sandy Springs, Marietta and Alpharetta, read as local to a homeowner who has been comparing three national-looking sites.
Where an Atlanta pool budget drains: aquatic centers, above ground kits and job seekers
Pool is one of the least specific words in paid search, and an unmanaged account funds every meaning of it.
The traffic you do not want arrives constantly: public and community aquatic centers, swim lessons, lifeguard and construction job hunting, above ground kits and big box liners, chemical and supply shopping, pool tables, and homeowners looking for a weekly cleaning route.
Start with a negative list built from all of that, then keep reading the search terms report every week for the first stretch. The terms that waste money in a Cobb County account are not identical to the ones wasting money in a DeKalb account, and no starter list is complete.
Match types matter more than most accounts admit. Broad match with a thin negative list will find every unrelated meaning of the word. Start tight, expand deliberately, and let the search terms report tell you what to add.
Every dollar recovered here funds a click from someone who actually wants a pool built. Cleanup is unglamorous, and it is usually the single highest return week of work in a new pool account.
Renovation clicks are cheaper, which is exactly why they need a fenced budget
Left in the same campaign, low cost renovation traffic will quietly consume the money meant for new construction.
Resurfacing, equipment replacement, deck work, tile and coping, converting an old pool to a modern system: those searches are steadier and generally cheaper to buy than new construction, and the auction will send you all of them if you let it.
Fence the money. Separate campaigns, separate daily budgets, separate landing pages, separate conversion actions. Otherwise a single automated bidding strategy optimizes toward the cheap conversions and starves the expensive ones, and the reports look great while the build calendar empties.
Renovation demand also skews toward intown and inner ring suburbs where the pools are simply older. Decatur, Brookhaven and parts of Buckhead behave differently in this account than a newer subdivision in Johns Creek does, and geographic bid adjustments should reflect that.
Both lines are worth having. A renovation job keeps a crew working during a soft month and often turns into a full remodel later, but it should not be funded by accident.
Winter is when Atlanta pools get sold, so that is when the money has to be there
A real winter separates the season when pools are bought from the season when they are built, and spend should follow the buying.
Homeowners here research in the cold months, take quotes in late winter and early spring, and want water by summer. Accounts paced to spend evenly across twelve months underfund the weeks that decide the year.
Plan for two curves rather than one. A heavy push ahead of construction season, and a lighter fall push aimed at renovation, equipment and closing work when the calendar has gaps.
Competitors help by disappearing. Plenty of pool advertisers switch off when the weather turns, which generally makes the winter auction less crowded exactly when the serious shoppers are in it.
Pace against crew capacity too. Selling more digs than you can schedule pushes delivery into the following season and turns a good month of advertising into a year of difficult phone calls.
Ad schedules and call routing built around the afternoon crawl on the Perimeter
Buying a click at six in the evening and letting it hit voicemail is the most expensive thing an account does quietly.
Homeowners search when they are home or stuck in traffic, which in this metro means evenings and weekends. Offices close at five. The mismatch shows up as impressions, clicks and no conversations.
Fix answering before you touch bid adjustments. Route calls to a mobile phone after hours, use a service that takes a real intake, or at minimum send an automatic text back with a booking link. Then look at the hourly data, because it means something different once calls are being answered.
Call tracking is not optional in this trade. A pool inquiry is a phone call far more often than a form, and an account that only counts form fills is optimizing against half its own results.
Weekend behavior is worth its own read. A Saturday morning searcher standing in the backyard is often the best inquiry of the week, and many accounts have never checked whether anyone is picking up.
Count excavations on the schedule, not form fills in the inbox
Cost per lead is a comfortable number that tells a pool builder almost nothing about whether the advertising worked.
The chain runs inquiry, answered call, booked site visit, design agreement, dig scheduled. Every step loses volume, and the losses are not spread evenly across campaigns or across the metro.
Feed the later steps back into the platform. Importing offline conversions for booked consultations and signed agreements lets automated bidding aim at outcomes that pay, rather than at whichever campaign produces the cheapest click.
Report by zone, not as a metro average. A blended cost per booked job hides the fact that one geography is carrying the account and another is being subsidized by it. Fulton, Cobb and Gwinnett can behave like three different markets in the same spreadsheet.
Give each change enough time to mean something. A pool account rarely produces enough weekly volume for fast conclusions, so change one structural thing at a time and read it across a longer window than feels comfortable.
Questions we actually get
- How much should a pool builder budget to start with paid search?
- There is no honest single figure, and any agency quoting one without seeing your market and job mix is guessing. The useful way to set it is backward: what a design agreement is worth to you, roughly how many inquiries your team converts, and how many digs you can actually schedule. Fund one geography properly rather than spreading a thin budget across the whole metro, where it will produce data too sparse to act on.
- Google Ads, Local Services Ads, or Meta?
- Search generally comes first because it captures people already looking. Local Services Ads can sit alongside it and are worth testing where they are available for your category, since they charge per lead and sit above the standard results. Meta reaches homeowners who are not searching yet and tends to work better for renovation and for showing finished work than for pulling immediate new construction inquiries. Start with search, add the others once measurement is clean.
- Should we bid on competitor names?
- It is allowed to bid on competitor terms, but claims in your ad text about another company or about credentials are a different matter, and the platform policies and your own legal exposure both apply. Competitor traffic also converts poorly for high consideration work. Spend the money on problem and project searches first, and treat competitor bidding as a small experiment with its own budget.
- Should we keep ads running in winter?
- Generally yes, and often harder than in summer. In a two-season climate, winter is when homeowners research and get quotes for spring construction, and part of the field switches off, which usually makes the auction less crowded. The shape of the year matters more than the total: money ahead of construction season, a lighter fall push for renovation and equipment work.
- Our cost per lead looks good but we are not booking jobs. What is wrong?
- Usually one of three things. The account is counting cheap, low intent conversions such as renovation or service inquiries in a new construction budget. The geography is buying inquiries outside a serviceable drive time. Or calls are not being answered when homeowners actually call, which is evenings and weekends. Track the chain through to booked site visits and signed agreements, and the leak generally identifies itself.
