The evaporative to refrigerated conversion deserves its own campaign
A homeowner replacing a swamp cooler with central air is a different buyer, a different budget and a much slower decision than someone with no heat tonight.
Conversion searches start in spring and finish weeks later. The job needs an estimate visit, usually some electrical work, and often a conversation about the ductwork. Bidding it inside a general cooling campaign hands the money to whichever keyword clicks fastest, which is rarely the one that sells a system.
Give it a campaign, a landing page and a daily budget of its own, and give reporting a longer window, since a household can take weeks to sign. Judging that campaign on a seven day conversion window will make good spend look broken.
Emergency no-heat gets its own campaign with a tight schedule and phone-first ads. Keep the two apart so an October cold snap does not quietly eat the money set aside for April conversions.
A third, smaller bucket covers maintenance and indoor air, including humidification. Cheap clicks, modest volume, and it feeds the membership base that carries the slow weeks.
Golden looks close on a map, and the foothills route makes it a half day
Geography for a Front Range account is drawn on drive time and elevation, not on a circle around the shop.
A twenty mile radius from a Lakewood or Arvada address reaches into the foothills. The van climbs, the return trip is longer, and the same job pays worse. A radius does not know that, and neither does the person building the account from out of state.
Decide deliberately whether you work up the hill. If you do, price it and say so. If you do not, exclude it in location settings and say that on the site too, because turning down calls politely still costs staff time.
Set targeting to presence rather than interest. People researching Denver from somewhere else will click, and they will never be customers.
Denver is a city and county. Its neighbors are not. Aurora reaches across more than one county, so a county level target can pull in far more than intended. Build the map from the municipalities and ZIP codes you actually dispatch to, then check it against the job board every quarter.
Swamp cooler pads, filter sizes and rebate lookups will eat the budget
Most of what a broad HVAC keyword matches in this metro is somebody doing the work themselves.
The do-it-yourself layer is thick here. Cooler pads, belts, float valves, filter sizes, thermostat wiring, relighting a pilot. Every one of those is a paid click and none of them is a job. They belong on the negative list before the account goes live, not after the first invoice.
Program and paperwork searches are the second layer. Someone hunting for a utility rebate page or a program application is not ready to buy from you, and terms change anyway. Add them as negatives and let your written pages serve that reader for free.
Trade traffic is the third. Technicians and other contractors search setup and altitude questions, and job seekers search your competitors by name. Neither one books work.
Storm weeks bring a fourth. After hail moves through, the metro fills with roofing and insurance queries and with out of town crews bidding on anything nearby. Read the search terms report weekly through that stretch, then monthly once it settles.
A three day cold snap is your whole month, so hold cash for it
Budget spread evenly across a month is budget that misses the week the phone actually rings.
Cold arrives fast along the Front Range and leaves fast. The first hard snap after a warm stretch produces the no-heat calls, because equipment that has not run since spring gets tested all at once, everywhere, on the same night.
Keep a reserve. Run deliberately under budget in calm weeks so you can raise caps for a few days without needing anyone's approval. A shop that has to ask permission on a Tuesday morning has already lost the week.
Cap against capacity, not ambition. If the dispatch board is full through Thursday, more spend on emergency terms buys angry voicemails. Shift the money to replacement and conversion campaigns, which can wait.
Shoulder seasons are short. Spring conversion campaigns should be live before the first warm week, not assembled during it.
People here type swamp cooler, not evaporative, and your keywords should too
The vocabulary a Front Range homeowner uses is not the vocabulary printed on your service menu.
Homeowners say swamp cooler, refrigerated air, forced air, boiler and mini split. Your website may say evaporative cooling systems and ductless inverter technology. Build the keyword list from the words people type and save the trade language for the technician.
Use exact and phrase match for the terms that make money. Broad match belongs in an account with a mature negative list and real conversion history, and never in month one.
Ad copy should echo the query. An ad offering an evaporative cooling system assessment against a search for swamp cooler to ac cost gets scrolled past, and the click you do get is more expensive than it needed to be.
Read the search terms report as market research. The phrasing that converts in paid is the phrasing worth writing pages around, and the two channels should be trading notes.
Price a click against the last open slot on the board, not against a lead
The number worth reporting is what a booked job cost, and the ceiling on that number is your capacity.
Send outcomes back into the account rather than form fills. A booked job is one conversion with a value. A sold system is another. An account fed on form submissions will happily buy you more forms.
Report cost per booked job by season and by job type. Averaging a two hundred dollar diagnostic with a full changeout produces a figure that describes neither, and it hides the campaign that is actually paying for itself.
Watch what added spend does to the schedule. If more budget only stretches the arrival window, you are buying cancellations. Hire, raise price, or hold the budget where it is.
Nobody can promise a cost per booked job in advance, and any agency quoting one before seeing your close rate is inventing it. What can be committed to is honest measurement inside a season, and acting on what it shows.
Questions we actually get
- How much should a Denver HVAC company spend on Google Ads?
- There is no honest number to quote before seeing your close rate, average job value and dispatch capacity. The method is to start at a level you can afford to lose for one season, measure cost per booked job by campaign, and scale only the campaigns where that number sits comfortably under your gross profit per job. Capacity, not budget, is usually the real limit.
- Should we run Local Services Ads alongside search?
- Often worth testing, since the format rewards license and insurance paperwork you already keep and it sits above the search results. Treat it as a separate channel with separate accounting, not as a replacement for search campaigns, and keep an eye on lead quality since the intake questions are not yours to write.
- Do we bid on competitor names?
- Sometimes, and it depends on the competitor. Bidding against a national franchise with heavy brand demand can be cheap and productive. Bidding against a well liked local shop invites a spending fight that helps neither of you. Run it as a small, separately budgeted campaign so you can see what it costs and shut it off cleanly.
- How do we handle the foothills without wasting money?
- Make it a decision instead of an accident. If you serve up the hill, run it as its own campaign with its own bids so the extra drive time shows up in the numbers. If you do not, exclude those areas by municipality and ZIP rather than trusting a radius, and add a coverage line to your site so the calls stop arriving.
- What should we do with paid search during hail season?
- Watch the search terms report closely, because storm-driven queries and out of town operators change what your keywords match. Add negatives quickly. Keep ad copy on the equipment side of the conversation, point damage questions at a technician visit, and tell homeowners to speak with their own insurer rather than offering to deal with a claim on their behalf.
