Contractor keyword research has a reputation for being complicated, and that reputation is mostly manufactured by people selling $99-a-month tools. For a trade business serving a handful of towns, the entire job takes about twenty minutes, costs nothing, and produces a list of maybe thirty phrases you will use for the next two years.
We are S4, two builders out of Rigby who now build websites for contractors across Eastern Idaho. Here is exactly how we do it, in the order we do it, with nothing you have to pay for.
Minutes 1–5: your invoices
Start here, not in a tool. Open your last two years of invoices or your job list and write down what you actually got paid to do, in the words a customer would use.
Not “residential construction services.” Water heater replacement. Shop slab. Tub-to-shower conversion. Gutter replacement. Panel upgrade. Whatever is literally on the invoices.
Then do the same for towns. Where did you actually bill? For most Eastern Idaho shops this comes out to three to six towns with real volume — usually some combination of Idaho Falls, Rigby, Rexburg, Ammon, Shelley, Ucon, Iona, or Blackfoot — plus a long tail of one-offs.
You now have the two lists everything else builds on: what you do, and where. This step alone gets you eighty percent of the way, and it is the step everybody skips because it does not feel like keyword research.
Minutes 6–10: Google autocomplete
Open a browser in a private window so your own history does not skew things. Start typing your service plus your town and watch what Google offers to finish.
Type “bathroom remodel Idaho” and let it suggest. Type “concrete contractor Rex” and let it suggest. Type “how much does a” plus your service. Then try the alphabet trick — type your service and then “a,” then “b,” and so on — to shake loose phrasings you would not have guessed.
These suggestions are drawn from what people actually search. It is the single best free signal there is, and it is right there. Write down anything that matches work you do, in the exact wording Google showed you. The wording matters — if people say “cement” when you say “concrete,” you want both on the page.
Minutes 11–15: People Also Ask
Search a few of your main service-plus-town phrases and scroll to the People Also Ask box. Expand a couple of the questions and it will generate more.
This is your question list, and it feeds two things: the FAQ section on each service page, and the headings on those pages. Writing a heading as the actual question — “How much does a bathroom remodel cost in Idaho Falls?” — and answering it in the next line is one of the highest-leverage formatting choices you can make, both for Google and for the AI summaries that increasingly sit above the results.
Ten to fifteen real questions is plenty. Write them down verbatim.
Minutes 16–20: your own Google Business Profile data
This is the source most contractors do not know they have, and it is the most valuable one on the list, because it is not an estimate — it is what people typed before they landed on you.
In your Google Business Profile, open the performance section and find the report of search terms people used to find your business. Google documents where to find it in Business Profile Help.
You will usually find two things. First, branded searches — people typing your company name, which mostly means word of mouth is working. Second, and more useful, the non-branded phrases: the actual service-and-town combinations that surfaced you. Those are proven demand in your exact market. Every one of them deserves a page or a section.
If you also have Google Search Console on your site, do the same thing there and pull the queries you already get impressions for. Free, and it is real data about your market rather than a national average.
Turning the list into pages
You should now have thirty to sixty phrases in one document. Sort them into three buckets and the site map writes itself.
- Service pages. One page per service you actually sell, using the customer’s word for it. If people search “tub to shower conversion” and you call it “bathroom modification,” the page uses their words.
- Town pages. One per town where you have done real work — and only those. A town page with no real job behind it is a thin page, and duplicated town pages are named directly in Google’s spam policies as scaled content abuse.
- Questions. These become FAQ entries and headings on the pages above. Most of them should not get their own page; they should live where the buying decision happens.
Then prioritize brutally. Build the pages for your highest-revenue service in your home town first. Not the phrase with the most searches — the one attached to the work you most want more of.
Why search volume misleads you here
This is the part that matters most in a small market, and it is where paid tools do real damage.
Punch “bathroom remodel Ucon” into a keyword tool and it will report zero or ten searches a month. The obvious conclusion is to skip it and target something bigger. That conclusion is wrong, for two reasons.
First, the tools’ estimates are unreliable at small volumes. They are modeled numbers, and in towns of a few thousand people the model is mostly noise. “Zero” frequently means “below our reporting threshold,” not “nobody searches this.”
Second, ten searches a month for a $25,000 bathroom remodel is a fantastic keyword. If you convert one of those ten, the page paid for itself many times over. Meanwhile the high-volume phrase the tool pushes you toward — “bathroom remodel Idaho” — pulls in people in Boise and Coeur d’Alene you cannot serve, and you will be competing against companies with far more resources for the privilege.
In a market like Eastern Idaho, intent and proximity beat volume every time. A phrase with obvious buying intent and a town name in it is worth more than a phrase with fifty times the traffic and no location.
What to skip
- Paid keyword tools, at this stage. Free sources cover a local trade business completely. Revisit if you ever expand to a genuinely different market.
- Chasing statewide or national phrases. “Best contractor Idaho” is unwinnable and useless if you won it.
- Building pages for services you do not want more of. If you hate small punch-list jobs, do not build a page optimized for them. You will get exactly what you ask for.
- Keyword-stuffing. Once in the heading, once or twice naturally in the body. Nine repetitions reads like a robot and does not help.
- Redoing this quarterly. Twenty minutes once, then a refresh once a year or when you add a service or a town. It is not a subscription.
- Building thirty pages this month. Two good pages beat thirty thin ones. Work the list down over a few months.
FAQ
Do I need a paid keyword tool as a contractor?
No. Autocomplete, People Also Ask, your Google Business Profile search terms, and your own invoices cover a local trade business. Paid tools become useful at multi-market scale.
How many keywords should I target?
Roughly one primary phrase per page, and most contractors need somewhere between fifteen and thirty pages total. If your list is at two hundred, you are collecting rather than deciding.
What if the tool says my town has zero search volume?
Build the page anyway if you have done real work there. Small-town volume estimates are unreliable, and a handful of high-intent searches for a five-figure job is a good trade.
Should I target “near me” phrases?
Not directly — you cannot optimize for someone else’s location. Google resolves “near me” using proximity and your Business Profile, so the way to win those is a complete profile and real town pages.
How often should I redo this?
Once a year, or whenever you add a service line or start working a new town. It is not ongoing work, whatever a monthly retainer implies.
What is the single most valuable keyword source?
Your Google Business Profile search-terms report. It is the only one that tells you what real customers in your real market typed right before they found you.
If you would rather skip the twenty minutes and have builders who know these towns map it out for you, here is how we work. How we work →
S4 Media 
Leave a comment