Four intents, four pages
Most HVAC keyword research produces a spreadsheet sorted by search volume, which is the least useful way to sort it. Volume tells you how many people type something. It tells you nothing about whether they will read your page, what they are worth, or whether the term can be won at all. Intent tells you all three.
The four intents, and how each behaves
Same trade, four completely different people.
Emergency searchers type a symptom rather than a service. They are on a phone, they decide in under a minute, and they are largely insensitive to price at two in the morning in August. They do not read. Position and a tappable number decide it.
Replacement searchers type a product and a cost question. They read for two or three weeks, gather several quotes, and involve a second decision maker. Content, pricing context and financing clarity decide it, and proximity barely registers.
Maintenance searchers are planning rather than reacting, which makes them the only group that reliably reads a page in a calm state. Commercial searchers are a different species again: a facilities manager on a long evaluation cycle who may be assembling a shortlist months before any tender exists.
Sorting by value rather than volume
The most searched term in your set is rarely the most valuable one.
A repair term with two thousand monthly searches and a six hundred dollar ticket and a replacement term with two hundred searches and a twelve thousand dollar ticket are not close. The replacement term is worth more per month even at a tenth of the volume, and it is usually less contested by the discount operators crowding the repair result.
The practical method is to multiply estimated volume by your own average ticket for that job type and by a realistic conversion rate, then rank the list by the result. It takes an afternoon and it routinely reorders a keyword list completely.
- Estimated monthly volume for the term
- Your own average ticket for that job type, not an industry figure
- A conservative conversion rate, and the same one across every term
- Multiply the three, then rank by the result rather than by volume
- Cross out anything you cannot service, however attractive the number
Symptom language beats service language
People describe the problem, not the solution.
Contractors name terms the way the trade names them. Customers name them the way the problem feels. Far more people type "ac not blowing cold" than "air conditioning repair services", and far more type "furnace making noise" than "heating system diagnostics".
This gap is where most of the easily winnable long tail sits. Symptom phrases are less contested because agencies working from a service list never think to target them, and they convert well because the searcher has just described their exact situation and found a page that matches it.
The caution is not to build a separate page for every symptom. A handful of symptom phrases belong on the main repair page, and the rest are better handled as sections within it than as thin pages competing with each other.
Terms worth deliberately ignoring
A keyword list is defined as much by what is left off it.
Competitor brand names are the first. Ranking for another contractor name is difficult, converts poorly, and produces a visitor who was looking for someone else. The effort is better spent almost anywhere.
Terms you cannot service are the second, and they are more tempting than they should be. A high volume term for work outside your area or outside your capability produces calls you have to turn down, and turned down calls generate the kind of review that costs you on every other term.
The third is anything where the result is dominated by national lead sellers and the intent is purely informational. Those pages can be worth writing as guides, but they should not sit in the commercial keyword set where their failure to convert will look like a program failure.
- Competitor brand names, which convert poorly and are hard to win
- Work outside your service area or capability
- Purely informational terms placed in a commercial set
- Very high volume generic terms where national aggregators dominate
How many keywords a contractor actually needs
The count follows from your service lines and areas, not from a package.
Multiply the service lines you sell by the areas you want to be visible in and the number grows quickly. A single trade shop covering repair, replacement and maintenance across two suburbs is already at a dozen meaningful terms before any symptom phrases are added.
A combined HVAC and plumbing operation covering four areas exhausts twenty keywords immediately, which is why keyword capacity rather than price is usually what decides the right plan. Buying a plan that does not fit means the program silently covers a fraction of your business while appearing to cover all of it.
Questions about keywords
What are the best SEO keywords for an HVAC company?
The ones that match work you actually want and can service, ranked by volume times your own average ticket rather than by volume alone. That calculation usually promotes replacement and commercial terms above the higher volume repair terms most lists lead with.
Should I target keywords with a city name in them?
Within your genuine service area, yes, and they are often less contested than the generic version. What does not work is building pages for cities you do not serve, which produces calls you turn down and reviews that damage the terms you do want.
How many keywords should one page target?
One primary term and its close synonyms. Phrases that mean the same thing to a searcher belong on the same page, because splitting them makes those pages compete with each other. Phrases with different intent belong on different pages entirely.
Are long tail keywords worth the effort?
Symptom phrases in particular are, because they are less contested and convert well. The mistake is building a separate thin page for each one. A handful belong on the main repair page and the rest work better as sections within it.
How often should the keyword list be revisited?
Once a season is enough for most contractors, and the shoulder months are the natural time. What changes is usually your own business rather than the search landscape: a new service line, a new area, or a shift in which work you want more of.
What about voice search and AI assistants?
They pull from the same underlying signals rather than from a separate index, so a page that answers a question clearly and a business profile that is accurate serve both. The practical adjustment is writing answers that stand alone in a sentence or two, which is good practice regardless of who is reading.
Should I target questions people ask rather than services?
On guides, yes. On commercial pages, only where the question carries buying intent. "What does a new furnace cost" is a replacement query wearing a question mark and belongs on a commercial page. "How does a heat pump work" is genuinely informational and belongs in a guide where it will not drag down conversion figures.
How do I estimate volume without paid tools?
Rank terms relative to each other rather than trying to pin absolute numbers. You know from running the business which work comes in more often, and search autocomplete shows you which phrasings are common. That is usually enough to order a list correctly, which is what matters.
Keep reading
We will build the list with you
Send us your domain and the work you want more of. We will come back with the terms you can realistically win, ranked against your own average ticket rather than by search volume.