Count the ads on your own service query
Open a private browsing window, set the location to your service area, and search the term you would most like to own — ac repair tampa, emergency plumber st petersburg, roof replacement clearwater. Then count the paid results and read who owns them.
For most home service, health and professional categories in this market, the answer is uncomfortable. The text ad block frequently contains two or three national marketplaces — Angi, Thumbtack, Yelp, Porch, Networx, and in property-adjacent categories Zillow — before a business with a Tampa Bay address appears at all.
That changes what your bidding is actually doing. Most owners set bids by picturing the two or three companies in their ZIP code they think of as rivals. Those companies are in the auction, but they are usually not the ones setting the price.
Why an aggregator can pay more per click than you can
You buy a click in order to win one job. That puts a hard ceiling on what a click can rationally be worth: gross profit per job, multiplied by the share of clicks that become enquiries, multiplied by the share of enquiries you close. Every competent bid strategy is an attempt to stay under that ceiling.
A lead marketplace is buying something else. It buys a contact and sells that contact to several service providers in the same category. Its revenue per acquired lead is a multiple of yours before any of the secondary monetisation — the follow-up email programme, the retargeting, the second and third category it can sell the same household later.
Three further asymmetries compound it:
- No capacity constraint. No crew, no trucks, no schedule. A marketplace can keep buying volume in Tampa Bay long after every local operator in the category is booked out three weeks.
- Portfolio-level optimisation. They are bidding across hundreds of metros and dozens of categories, so a market that underperforms is subsidised rather than switched off. You have one market and one P&L.
- They can afford to be wrong. A misrouted lead costs them very little. A misrouted click costs you the whole click.
The part that makes this worth a business owner’s attention: the lead they bought with the click that beat your bid may be sold back to you. Marketplace lead fees run from the mid-teens for small jobs into three figures for high-ticket trades, and plenty of contractors are simultaneously paying Google to compete with the platform they are buying leads from.
What that does to Tampa Bay CPCs by category
Auction price is set by the highest sustainable bid, not by the average one. One national bidder with a resale-based ceiling raises the floor for every local advertiser on that keyword, permanently, because they are not going to run out of budget in the way a local competitor does.
Tampa–St. Petersburg is a large metro with dense service demand and sharp seasonal spikes, which puts it well above the mid-size Florida markets and somewhat below the South Florida metros. What we typically see, and what you should verify against your own account rather than take from an article:
- General home services and trades: roughly $8 to $25 on core category terms.
- Emergency HVAC and plumbing at peak demand: frequently $25 to $60, occasionally higher during heat or storm-driven spikes, when marketplace resale value is also rising and their bids climb with it.
- Restoration, legal and injury: multiples of the above, and the categories where aggregator and lead-gen pressure is most severe.
- Med spa, dental and elective health: mid-range on service terms, considerably higher on procedure terms with financing intent.
Pull the top of the range from Keyword Planner for your own category, not the middle, and compare it with the top-of-page bid estimate in your account. If those two numbers are drifting upward year over year while your conversion rate is flat, you are watching auction pressure rather than a campaign problem.
Where aggregators cannot follow you
Marketplaces build coverage programmatically at the city and ZIP level. That is what makes them fast and comprehensive, and it is also the seam in their model. They are structurally weak wherever relevance requires knowing something specific.
Neighbourhood and landmark modifiers. Ac repair seminole heights, plumber old northeast, water heater replacement near westshore. Volume is a fraction of the head term, the cost per click is routinely half, and a directory page listing forty contractors across the metro cannot be as relevant to that query as a page that names the neighbourhood and shows work done there. The same neighbourhood vocabulary that earns organic visibility in St. Petersburg works in the auction for the same underlying reason.
Brand plus service. Your own name attached to a service term converts at rates nothing else matches and costs very little. Worth holding, worth testing for incrementality rather than assuming, since a share of those clicks would have arrived free through organic.
Genuinely long-tail job descriptions. Cast iron repipe st petersburg, ten ton rooftop unit replacement, wind mitigation inspection before closing. Specific enough that no templated directory page can be a credible landing experience.
Local Services Ads. The block above text ads requires licence, insurance and background verification tied to the business performing the work, which is why the marketplaces are not in it. If you qualify in your category, this is the one placement where the structural advantage runs the other way.
None of these individually replaces head-term volume. Collectively, for a business serving Tampa and the surrounding suburbs, they usually add up to a better account than the one being outbid on the obvious terms.
Negative keywords that strip directory shoppers
Broad match paired with automated bidding will find directory-shopping and research queries unless you fence them out, and those queries are exactly the ones the marketplaces are best positioned to win.
Five lists worth building at the account level as shared negatives:
- Marketplace and directory brand names. Someone searching a directory by name wants the directory. You will pay for that click and lose the comparison on a page you do not control.
- Comparison and listing intent. Top 10, list of, rankings, compare, directory, vs. Commercial-sounding, almost never a caller.
- Employment. Jobs, hiring, salary, apprentice, careers, union. In the trades this is a surprisingly large share of wasted spend.
- DIY and research. How to, diy, tutorial, video, parts, manual, replace myself.
- Adjacent services you do not sell. Every trade has neighbours. If you do not do ductwork, or repairs, or commercial, say so in negatives rather than in a phone call three minutes long.
Review the search terms report weekly for the first two months and monthly after that. Add negatives at the tightest level that works — phrase match at the ad group, exact at the account level for anything you might legitimately want elsewhere.
One clarification worth keeping straight: negatives cut wasted spend immediately, and they lower your price only indirectly and later, through the quality mechanism below.
Quality Score as the only lever that lowers your price
In an auction you cannot win on budget, there is exactly one lever that changes the price rather than the quantity.
Ad Rank combines your bid with expected click-through rate, ad relevance and landing page experience. Two advertisers in the same position can pay materially different amounts per click because of that second half. This is not a loyalty discount — it is the auction pricing relevance, and it is the only discount available to a local advertiser.
It is also where the aggregators are beatable. They have enormous click-through data and generic destination pages. On a neighbourhood-specific or job-specific query, a tightly built ad group pointing at a page that answers exactly that query outscores a directory listing page on relevance and landing page experience, which is how a smaller bid holds a position.
In practice that means single-intent ad groups rather than one ad group per service, the query echoed in the headline, and one page per ad group. Broad ad groups covering six services with one shared page are the most common reason a well-funded account pays above the market rate.
Landing pages built for one query, not one service
Most paid traffic in this market lands on a general service page or, worse, the homepage. That page has to serve five different intents, so it matches none of them, and the auction charges you for the mismatch twice — once in Quality Score and again in the conversion rate.
A page built for one query does a small number of things: it repeats the query in the headline, proves you work in that specific area with named jobs and photographs, puts licence and insurance detail where a nervous buyer looks for it, states a response time you actually meet, and puts a tappable phone number above the fold on mobile. Everything else is optional. We have written separately about the mistakes that kill conversion on these pages, and the paid version is the unforgiving case, because you are paying for every visitor who bounces.
The economics justify the effort. A page that lifts conversion from 3% to 6% halves your cost per lead without touching a bid, and it improves the quality component that determines the bid you needed in the first place. That is why landing page design belongs in a paid search budget rather than beside it.
When to stop bidding and put the money into local SEO
There are categories in Tampa Bay where the arithmetic does not work at any level of execution quality, and recognising one early saves a great deal of money.
Three signals, all of which should be judged after sixty to ninety days of competent management, not sooner:
- Cost per acquisition exceeds gross profit per job on your narrowed, best-performing terms — not on the account average, which hides the good campaigns inside the bad ones.
- Budget cannot reach conversion volume. Below roughly fifteen conversions per campaign per month, automated bidding never leaves a learning state and you pay premium prices for the privilege.
- Impression share lost to rank is dominated by aggregators at every hour of the day. Dayparting into the windows where they bid softer is worth testing; if there are no such windows, there is no position to buy.
Where those hold, the money moves to the channel where accumulated authority competes better than a bid does. We laid out the full arithmetic for that decision in Google Ads vs. SEO, and the ordering matters more than the label — a few months of paid search is often the cheapest way to learn which terms convert before committing to local SEO content around them.
A worked budget for a Tampa Bay service business
Take $3,000 a month in media spend for a residential trade, excluding management and page build. A defensible split, with illustrative numbers you should replace with your own:
Local Services Ads — $900. Priced per lead rather than per click, commonly $40 to $90 in Tampa Bay trades. Call it a dozen leads at a lower close rate than your website produces, because the qualification burden sits with you. This is the one placement the marketplaces are not in.
Core service terms, tight radius — $1,200. At $18 a click that is around 66 clicks. At 8% on a matched landing page, five enquiries; at a 40% close rate, two jobs. On $900 gross profit per job, $1,800 returned on $1,200 spent. Positive, thin, and the first bucket to cut if page quality slips.
Neighbourhood and long-tail terms — $600. At $7 a click that is 85 clicks — more traffic than the core bucket bought for half the money — and specificity usually lifts conversion. At 10% and the same close rate, roughly three jobs from $600. This is the bucket most accounts underfund because the keyword volumes look unimpressive in a planning tool.
Brand defence — $300. Cheap, high-converting, and worth periodically testing by pausing it in part of the service area and comparing total conversions rather than paid ones.
The point of laying it out this way is that the account average would report a mediocre campaign. Read by bucket, it says something specific: the money is in the segment the aggregators cannot reach, and the head-term bucket exists mainly to feed volume and learning. Most accounts we review have that ratio inverted.
If you want to know who is actually in your auction and where the reachable pockets are in your category, that analysis is the first deliverable in our PPC management work, and we will tell you plainly if the answer is that the budget belongs somewhere else. Get in touch with your category and service area and we will run it.