Keyword research for a national brand is a research project. For a business serving one town it is closer to an afternoon of listening, and most of the sources are free.
The goal is not a spreadsheet. It is a list of twenty phrases that matter and a decision about which pages own them.
Start with what you already know
Before any tool, write down what customers actually say. Not your industry terms, theirs.
Three sources you already have:
Your phone. What do people ask for when they call? “Do you do walk in showers” is a keyword. “Bathroom renovation solutions” is not.
Your inbox. Search your sent mail for the questions you answer repeatedly. Those are pages.
Your staff. Whoever answers the phone knows what people call things better than you do, because they hear it forty times a week.
This list is usually more accurate than anything a tool gives you, because it comes from your actual market rather than a national average.
Then use the free tools
Google’s own suggestions. Start typing your service into the search box and read what it offers. Those are real searches. Do it with your town name, without it, and with “near me.”
People also ask. The expanding questions in the results. Each one is a heading you could write, and clicking one produces more.
Related searches at the bottom of the results page.
Search Console, if your site has been up a while. Under performance, you can see the actual phrases people used before landing on your site. This is the only source that is definitely about your business rather than the category. Sort by impressions and look for terms where you appear but nobody clicks. Those are pages that could be better.
Your Google Business Profile insights. Shows some of what people searched before finding your listing.
What to look for
Ignore search volume for local terms. National volume numbers are not useful for a business serving three counties, and the tools that report them round small numbers to zero. A term with “no volume” can still bring you six jobs a year, which for a lot of businesses is a good year.
Look for these instead:
Intent. Is the person searching ready to hire, or reading? “How to fix a running toilet” is a reader. “Emergency plumber Fort Walton Beach” is a customer. Both are worth having, but only one pays this month.
Specificity. Longer phrases are less contested and closer to hiring. “Plumber” is a war. “Tankless water heater installation Bay County” is winnable.
Whether you can honestly compete. If the first page is three national franchises and a directory, that term is expensive. Find the version of it you can win.
Match phrases to pages, not the other way around
The mistake is collecting 200 terms and trying to fit them into a site. Do it the other way.
List your pages. For each one, name the single phrase it should own and two or three variations. That is it.
- Homepage: what you do and where.
- Each service page: that service and the place.
- Each location page: the main service and that town.
- Blog posts: the questions you get asked.
If two pages want the same phrase, one of them has to change, or they should be one page. Two of your own pages competing for the same search is a self-inflicted problem, and it is common on sites that grew without a plan.
The questions are your blog
Everything on your list that is a question rather than a service is a post. “How long does a new roof take,” “do I need a permit for a fence in this county,” “what does a service call cost.” You already answer these on the phone.
They will not bring in work next week. They bring in the person who is three weeks from hiring somebody, and they make you the firm they already read something useful from.
A list you will actually use
Twenty rows in a plain document. Four columns: the phrase, the page that owns it, whether you currently rank for it, and the date you last checked.
Update it twice a year. That is the whole system, and it will do more than a subscription you open twice.
An afternoon
- Write down twenty things customers say on the phone.
- Type your top five services into Google and note the suggestions and the related searches.
- Open Search Console and list the ten terms where you get impressions but few clicks.
- Assign each phrase to one page.
- Note the phrases with no page. Those are the pages to write next.
