Two counties, two campaigns, two budgets that never borrow from each other
Hillsborough and Pinellas are separate auctions with separate travel realities, and averaging them produces a number that describes neither.
One campaign spanning the bay will drift toward whichever side is cheaper per click. Cheaper clicks are not better patients. If the Pinellas side clears at a lower cost and books people who cancel their second appointment, the account is quietly funding its own churn.
Split at the campaign level so budgets cannot bleed. Separate bid strategies, separate ad copy, separate landing pages, separate call tracking numbers. When you review performance, you are then comparing two markets rather than looking at one blended figure.
The exception is a genuinely metro-wide case type. Full-arch or complex surgical work travels in a way that hygiene never will. Give that its own campaign with its own geography, and keep it away from the everyday new-patient budget.
If you only run one location and only serve one county, do not build the second campaign at all. Restraint is cheaper than optimization.
Name the towns you will not serve, and exclude them on day one
An exclusion list is a service area statement, and writing it down forces a decision most practices avoid making.
Ad platforms will happily show you to anyone within a shape you drew casually. A circle centered on a South Tampa address covers a lot of water and a lot of Pinellas. Every click from a place you cannot realistically serve is money out.
Write the list by name. Include the towns you want: Brandon, Riverview, Westchase, New Tampa, Temple Terrace, wherever your book actually comes from. Exclude the rest explicitly rather than relying on a radius to do it for you.
Watch the location settings that quietly widen your reach. Targeting options that include people merely interested in your area will pull in searches from out of state. For a practice, presence targeting is almost always the correct choice, and it is worth confirming the setting has not reverted after a platform update.
Revisit the list quarterly against the schedule. If patients from Wesley Chapel are booking and keeping appointments, add it deliberately. If nobody from Clearwater has ever come twice, stop paying to reach them.
Brand terms are cheap, and the practice in Wesley Chapel is bidding on yours
Your own name is the least expensive inventory in the account, and leaving it uncovered invites competitors to buy your referrals.
Someone who heard about you from a coworker will search your practice name. If a competitor is bidding there and you are not, that referral is now a coin flip decided by ad position.
Brand campaigns look inefficient on paper because they convert well and cost little, which makes every other campaign look worse by comparison. Report brand separately so it does not inflate your overall numbers or hide weak non-brand performance.
Do not put a competitor's name in your ad copy. Using another practice's name inside an ad raises trademark exposure and is worth confirming with your own counsel before anyone tries it. Bidding on the term while writing generic copy is a different question and a common practice.
Keep brand budget small and effectively uncapped. It should never be the campaign that runs out of money at two in the afternoon.
Broad match will find you patients in Pasco whether you asked for them or not
Loose match types and automated campaign types will spend your geography and your intent for you unless you constrain them.
Broad match learns from conversions. If your conversion action is a form fill, it will find people who fill forms, which is not the same as people who show up with a molar problem. Feed it the wrong signal and it optimizes toward the wrong patient with impressive efficiency.
Send the appointment back into the platform, not the form. Call tracking that stops at the ring tells you nothing. Call tracking connected to whether the appointment was booked, and ideally whether it was kept, changes what the algorithm chases.
Automated campaign types are worth testing only after the basics are stable and the conversion data is clean. Running one from day one on a small dental budget usually means paying tuition for the platform's education.
Check the search terms report weekly for the first two months. Jobs, dental schools, free clinic queries and pure price shoppers all appear early. Each one you exclude buys back budget for the searches that matter.
A household that transfers out in two years has to pay you back in year one
The relocation churn around MacDill changes what a new patient can be worth, and therefore what you can afford to pay for one.
Standard dental lifetime value math assumes a long relationship: recalls, restorative work, family members added over time. Some of that holds here. Some of it does not, because a share of your new patients arrive knowing roughly when they are leaving.
Price accordingly rather than pessimistically. A household on a known timeline is often ready to complete treatment quickly instead of deferring it, which is a real advantage if your scheduling can absorb it. Short windows are not the same as low value.
Track cost per booked case against first-year production rather than an assumed decade of visits. If a campaign only pays back over ten years, it does not pay back at all for a meaningful slice of this market.
Nothing here justifies charging differently or treating anyone differently. It affects your bidding math and your scheduling priorities, not your fee schedule.
Send a Brandon click to a Brandon page, or you pay for that visit twice
Landing page and ad have to agree on geography and procedure, or the visitor restarts their search on your homepage.
A person searching for an emergency appointment in Brandon who lands on a general homepage has to work out, unaided, whether you are near them and whether you can see them today. Most will go back to the results and click a competitor instead.
Match at three levels: the place, the procedure and the promise. If the ad said same-day emergency, the page opens with the emergency phone line and today's availability. If the ad said implants, the page is about implants and nothing else.
Put the phone number where a thumb lands, and make it the primary action. Dental inquiries convert by phone more often than by form, and a page that buries the number under a hero image is optimizing for the wrong thing.
One page per campaign is a maintenance burden worth accepting. Fewer, better pages beat a dozen thin ones, so keep the campaign count honest and the pages real.
Questions we actually get
- What should we spend to start?
- Enough to gather readable data in one county rather than a thin amount spread across two. Splitting a small budget across Hillsborough and Pinellas usually produces two datasets too sparse to act on. Start where your chairs already fill, prove the cost per booked case, then expand.
- Do we need Local Services Ads as well as search ads?
- They occupy a different position and are bought differently, so they are worth evaluating separately. Verification requirements and availability change, so confirm current eligibility for your practice type directly with the platform before budgeting for it.
- Why is our cost per lead good but the schedule still soft?
- Usually because the conversion action stops before the appointment. Forms and calls are not bookings. Connect tracking to the appointment book, then look at how many of those bookings were kept. The gap between the two numbers is generally where the problem sits.
- Should we run ads for hygiene appointments?
- Rarely worth it directly. Hygiene fills through recall, the map pack and family members of existing patients. Paid budget usually returns more on the procedures with real case value, with hygiene arriving as a byproduct.
- How quickly can we tell whether a campaign works?
- Fast enough to catch obvious waste, slow enough that judging booked-case performance takes longer than most owners expect. Review search terms and exclusions within the first two weeks. Hold off on structural verdicts until you have enough booked appointments to mean something.
