How to Do Keyword Research Without Expensive Tools

Keyword research for beginners gets taught backwards: start with a $99 tool, drown in 4,000 suggestions, give up. The working method runs the other direction — start with the words your customers already use, expand them with Google’s own free machinery, and judge each phrase on two questions: would ranking for this make money, and can my site actually win it? Here’s that workflow, start to finish, on a budget of zero.

01Harvest the language you already own

Before any tool: open your enquiry emails, reviews, and sales-call notes, and write down the phrases customers used when they didn’t know they were being quoted — “fix my slow website,” “logo that doesn’t look homemade,” “someone to run our ads.” That vocabulary is your seed list, and it beats brainstormed jargon every time, because customers search in their words, not your industry’s. (The gap is often comic: firms selling “bespoke digital transformation solutions” to people googling “fix my slow website” — and wondering why the two never meet.)

02Expand with Google’s free machinery

Type each seed into Google slowly and record what autocomplete offers — those are real queries, ranked by how often real people type them. Run the search and mine People Also Ask (each question is a future FAQ or post) and the related searches at the bottom. One seed phrase typically fans out into ten to twenty variants in minutes; the mechanics of each tool are in our free SEO tools guide.

03Sort by intent before volume

Every keyword carries a purpose: buying intent (“emergency plumber near me,” “web design quotes”) converts today; research intent (“how much does a website cost”) converts later via content; learning intent (“what is DNS”) rarely converts at all. Label each phrase B, R, or L. Small sites profit fastest by writing pages for B-terms and posts for R-terms — and cheerfully ignoring most L-terms until the library is deep.

04Check winnability by reading page one

Search each shortlisted phrase and study who currently ranks. Four signals say “winnable”: forum threads or Reddit posts on page one, thin or outdated articles, small brands rather than household names, and no result matching the query’s exact intent. If page one is wall-to-wall national giants, the phrase isn’t wrong — it’s a year-three target, not a month-three one. Volume tools guess demand; page one shows the competition with perfect accuracy, free.

05Map one keyword to one page — and stop

The rule that prevents most self-inflicted SEO damage: each primary keyword gets exactly one page, and each page gets exactly one primary keyword. Two pages chasing the same phrase split your strength and confuse Google about which to rank — cannibalization, the polite word for competing with yourself. Keep the map in a simple spreadsheet: keyword, intent label, target page, status. That document is your SEO strategy — the one artifact worth revisiting monthly — and everything else is execution on the usual timeline.

Once the site is live in Search Console, the method upgrades itself: the Performance report shows queries you already earn impressions for — pre-validated targets no external tool can see. And when the map outgrows the afternoon (hundreds of terms, competitor tracking, content at scale), that’s the honest hand-off point to a managed SEO campaign — send your draft map through the enquiry form and we’ll tell you which dozen phrases we’d chase first, and what chasing them costs, within the working day.

Keyword research FAQs