A monthly budget cannot survive a Middle Tennessee freeze week
Winter demand arrives in bursts, and a flat daily cap turns your best week of the year into your most rationed one.
Hard freezes and ice are ordinary here, not exotic. When one lands, the auction gets crowded and the calls that come through are real work, not browsing.
An account running at a fixed daily cap will spend out by mid morning on those days and sit dark through the afternoon, which is exactly when the second wave of burst lines gets discovered.
Hold back a reserve during mild months so you can raise caps for the event and pull back afterward. Decide the trigger and the ceiling in advance so nobody is improvising at six in the morning.
Pace against capacity, not against ambition. Buying calls you cannot dispatch for three days produces refunds, cancellations and a review problem that outlasts the storm. If the board is full, throttle the campaign and say so on the phone rather than booking a job you will have to move.
A radius drawn from Downtown buys Broadway traffic you cannot bill
A circle on a map ignores what the drive actually costs your truck and who happens to be standing inside it.
Radius targeting centered on Downtown pulls in Broadway and the visitor core, where a large share of the people searching are not property owners and are not your customer.
Target by county and by named place instead. Davidson as one structure. Williamson with Franklin and Brentwood. Rutherford with Murfreesboro. Sumner with Hendersonville. Wilson with Mount Juliet. Each gets its own budget line and its own bid position.
Set location targeting to people physically present in your areas rather than people interested in them, or you will pay for a steady stream of clicks from people planning a move next year.
Then adjust by what the truck actually experiences. If a Bellevue job means a crosstown run at four in the afternoon and a Donelson job does not, the bids should not match. Drive time is a cost, and the account should carry it.
Rock excavation quotes justify bids a drain call never will
Group campaigns by ticket size, because an identical click price is a bargain in one campaign and a loss in another.
Sewer line and water service replacement in this area often means digging through shallow limestone, which is a longer sale, a bigger scope and a customer collecting bids. Those searches deserve a higher bid, a longer conversion window and a landing page that explains how rock affects the quote.
Emergency work behaves nothing like that. A leak, a stopped drain, no hot water. The searcher calls the second or third result and is done in ninety seconds. Tight bids, call assets, hours accuracy, and copy that says you are available now.
Keep those in separate campaigns with separate budgets. Mixed together, the emergency terms will consume the money and the replacement work will quietly never get funded.
Match the landing page to the campaign. Sending replacement searches to a general services page wastes the premium you just paid for the click.
Commercial and property management demand is a third structure. Operators buying on turnaround time, remote approval and clean invoicing are reachable through different keywords and respond to different copy than a homeowner does.
Apprenticeship, supply house and how to solder: the clicks you buy by accident
Broad and phrase match will spend your budget on people who were never going to hire anyone.
Build the negative list before you turn the account on, not after the first invoice. Job seekers, apprenticeship and salary searches. Trade school. Parts, fittings and supply house queries. Every variation of how to and do it yourself. Tool rental. Free.
Add the category negatives that fit your business. If you do not take home warranty dispatch, exclude it. If you do not do septic, exclude it. If you do not want tenant calls where the landlord makes the decision, that is a negative theme too.
Review the search terms report weekly for the first two months and monthly after that. The waste is never where you expected it to be, and a single unnoticed term can absorb a meaningful share of a small account.
Competitor names are a judgment call. They can work when your landing page gives a real reason to switch. They usually do not when the page just says you are local and trusted.
A Bellevue lead and a Murfreesboro lead cost the same and are not worth the same
Geography should be judged on booked revenue, not on the price of a form fill.
Two counties can produce identical cost per lead and completely different businesses. One books at a decent rate on real work. The other produces price shoppers, cancellations and a truck that spends an hour on the road each way.
Close the loop. Call tracking with outcome codes, and a habit of coding every inquiry: booked, quoted, no answer, out of area, not our work, price. Push booked outcomes back into the ad platform so bidding optimizes on something real.
Then read cost per booked job by county rather than by ZIP code. A single metro does not generate enough volume to make ZIP level decisions honest, and small samples will send you chasing noise.
When a county underperforms, the answer is usually one of three things: the bid is wrong, the coverage promise is wrong, or the landing page does not name the towns you actually serve. Check in that order before you cut the county entirely.
Ice on the roads should change what the ad promises that morning
Copy that promises fast arrival during an ice storm buys complaints instead of jobs.
During a weather event, swap the promise. Honest wait expectations, emergency triage, and clear guidance on what to do while waiting will convert better than a same day claim you cannot keep.
Have the alternate copy written and paused in the account before winter so it can be switched on in a few minutes rather than drafted in a crisis.
Use ad scheduling that reflects your dispatch reality rather than your storefront hours. If nobody answers after nine at night, running call ads at ten produces missed calls and paid clicks with no job attached.
For the property management and short-term rental side of the market, write to their actual constraint. They buy on turnaround time, remote approval and an invoice that clears their process. Speak to that in the ad and the landing page, and keep it separate from the homeowner campaigns.
Questions we actually get
- Should we run Local Services Ads, search ads, or both?
- They do different jobs. Profile based local ads tend to catch the immediate emergency searcher who wants the nearest available company. Search campaigns are where you can control messaging for planned replacement work, exclude the traffic you do not want, and send people to a page that explains a complicated job. Most plumbing accounts end up running both with separate budgets and separate expectations.
- What is a reasonable budget to start with?
- Start with enough to gather usable data in your primary county rather than spreading a small budget across five. One county funded properly will tell you what a booked job costs, and that number is what lets you decide whether to expand into Rutherford or Sumner. Any agency quoting an expected cost per lead before running your account is inventing it.
- How do we measure a booked job rather than a lead?
- Track calls and forms, then code every inquiry by outcome and reconcile against your dispatch software or job sheets. It takes a few minutes a day from whoever answers the phone. Without that habit, the ad platform optimizes toward whichever campaign produces the most cheap clicks that happen to ring, which is rarely the profitable one.
- Should we advertise during a freeze if we are already fully booked?
- Usually not at full spend. Buying calls you cannot serve creates cancellations and reviews that cost more than the jobs were worth. Throttle instead, keep the highest value replacement campaigns running, and set an honest expectation on the phone about when a truck can get there.
- Do property managers respond to the same ads as homeowners?
- No. Managers and short-term rental operators are buying turnaround time, remote approval and predictable invoicing, often across several properties at once. That needs its own campaign, its own keywords and a landing page that talks about scheduling and paperwork rather than about emergencies.
