Crossing the Mississippi is a licensing decision before it is a targeting decision
Turning on Illinois in a location setting takes four seconds and commits you to a different state's rules.
The Metro East is a genuine part of this market. Belleville and Edwardsville have households with failing furnaces and money to spend. They are also in another state, under another licensing regime.
Before you buy a click in Madison or St. Clair County, confirm with your own counsel or the relevant state authority that you can legally perform and permit the work you are advertising. Advertising into a jurisdiction you cannot serve is expensive and it is a complaint waiting to happen.
If you can serve it, run it as its own campaign with its own budget and its own ad copy. Illinois buyers do not respond to Missouri geography, and blending them into one campaign hides which side of the river is profitable.
If you cannot serve it, exclude it explicitly rather than trusting a radius to stop at the water. Radius targeting reaches across the river without asking.
People searching a high Ameren bill are not buying a furnace
The exclusion list is where most of the money is saved, and most HVAC accounts have one built for a different climate.
Utility bill searches, outage checks and thermostat programming questions all pull HVAC ads and none of them book a truck. So do filter sizes, DIY repair walkthroughs and parts lookups.
Build the negative list from your own search terms report, weekly at first, and read it with the phone log open next to it. The terms that produced calls nobody could sell are more informative than the terms that produced nothing.
Home warranty traffic deserves specific attention. Someone whose warranty company assigns the contractor is not choosing you, whatever they click. Decide deliberately whether you want that work at all, then set the account to match the decision.
Rental and landlord terms are a judgment call rather than an automatic exclusion. A landlord with several brick four-families around Tower Grove is a portfolio, not a nuisance, if your pricing and scheduling can handle the work.
Split the account by what heats the house, not by season alone
A boiler call and a furnace call come from different houses, cost different amounts and close at different rates, and one campaign cannot price both.
Hydronic work in older city brick is a specialty. Fewer shops take it, the searcher has fewer options and the ad copy that wins is completely different from a furnace repair ad.
Give boiler and radiator terms their own campaign or at least their own ad group with its own landing page. Sending that searcher to a generic heating page wastes a click you paid a premium for.
Do the same for ductless. Someone asking how to cool a third floor in a house with no ductwork is a considered purchase, not an emergency, and the ad should say something about the house rather than about fast service.
Keep emergency no-heat separate from all of it. Different hours, different bids, different copy, different measure of success.
During the first hard freeze, buy less volume at a higher price
When the phone is already ringing off the hook, more budget buys you calls you cannot answer.
The morning after the first real freeze, every HVAC shop in the metro is buried. Auction prices climb, the phone is jammed, and additional spend converts into missed calls and canceled bookings.
The better move is usually the opposite of what the dashboard suggests. Raise bids so you stay visible for the highest value terms, and pull total budget back to what dispatch can genuinely take that day.
Then spend the recovered budget in the weeks the market is quiet, when a replacement shopper has time to talk and a competitor is not being called first. Even budget across twelve months matches nothing about this business.
Set a rule in advance about what happens during an ice event, because nobody makes a calm pacing decision at seven in the morning with forty voicemails waiting.
Name the suburb in the ad, because every competitor is bidding on the word St. Louis
Ad copy that names where the reader lives outperforms copy that names a metro they only half identify with.
A homeowner in Florissant, Kirkwood or St. Charles does not think of themselves as living in St. Louis in any practical sense. The city is an independent city, legally separate from the county, and residents hold that distinction more firmly than outsiders expect.
Use location insertion carefully and back it with honest coverage. If you do not go to Chesterfield on a Sunday, do not run copy that implies you do, because the call still costs you money and the booking fails at the arrival window.
Give each geographic campaign a landing page that matches. A Metro East ad landing on a Missouri page loses the reader in the first line.
Say the constraint out loud in the copy where it helps. Naming your actual dispatch area filters out the calls you would have turned down anyway, and a slightly lower click volume with a higher booking rate is the trade you want.
One conversion action cannot stand for a diagnostic and a changeout
If every call counts the same, the platform will optimize toward the cheapest and least valuable half of your business.
Google spends your money in pursuit of whatever you told it to count. Count form fills and it will find people who fill in forms. Count calls over a set duration and it will find talkers.
Feed it outcomes instead. Pass back which calls booked, and where you can, what they were worth. Even a rough split between a service visit and a system replacement changes how the account allocates spend.
That requires somebody to reconcile the dispatch board against the ad account on a regular schedule. It is unglamorous work and it is the difference between reporting and managing.
End every report on the number that matters: what a booked job cost, by campaign and by geography. Cost per lead can fall while the business gets worse, and it often does.
Questions we actually get
- How much should a St. Louis HVAC company budget for Google Ads?
- Work back from capacity rather than from a benchmark. Decide how many additional jobs you can genuinely dispatch in a week, estimate what a booked job is worth to you, and set spend against that. Any number quoted without knowing your close rate and your average ticket is a guess dressed as advice.
- Should we only run ads in the peak heating and cooling months?
- Usually not. Peak weeks are the most expensive and the most crowded, and they are also the weeks your phone is already full. Many shops do better keeping steady presence in the shoulder seasons, when replacement shoppers have time to talk and the auction is calmer.
- Is it worth bidding on competitor names?
- Sometimes, and it is rarely the first thing to fix. It tends to be defensible when your own account is already clean and you have a real reason for someone to switch. Bidding a competitor's name while your negative list is empty is paying for the privilege of arguing.
- Do Local Services Ads make sense for an HVAC shop here?
- They can, particularly for urgent repair demand, and they reward the license and insurance paperwork you already keep. Treat them as a separate line item with their own measurement, not as a replacement for search, and check the leads carefully in the first weeks because the qualification varies.
- Our cost per lead dropped but revenue is flat. What happened?
- Almost always the account learned to buy cheaper, lower value calls. Diagnostics and parts questions convert as leads and never become installs. Fix it by passing booked job data back into the account so it optimizes toward what actually invoices, then judge the account on cost per booked job by campaign.
