How to Do Keyword Research for SEO Without a Mountain of Tools

The board beats the tool, every time

Most teams do not fail at keyword research because they lack data. They fail because the data never becomes a decision. A list of 4,000 rows feels like progress, but nothing in it says which page to write on Monday morning. That is the gap a board closes: a keyword is only finished when it has a page, an owner and a date.

Three things break when research stays inside a spreadsheet. First, two writers pick up the same topic and publish near-duplicate pages that compete with each other. Second, a promising question keyword gets filed next to a transactional one and the writer produces a sales page that answers nothing. Third, nobody revisits the winners, so a page stuck at position nine never gets the internal link that would move it.

A board fixes all three because it is visible, columned and finite. You can see what is waiting, what is blocked and what shipped. When someone asks what SEO is doing this month, you point at the write column. The board is also the cheapest place to say no to a keyword, which is a decision most teams never make out loud.

What goes in each of the four columns

The column names matter less than the rule that a note may only move forward. Four stages are enough for a small team, and each one has a clear exit condition.

The exit condition for "to research" is simple: the note has an intent tag and one sentence describing the searcher's problem. The exit for "to map" is a URL, even if that URL does not exist yet. The exit for "to write" is a brief with a title, an outline and the internal links the page will use. The exit for "live" is a date and a check-back date, because publishing is the start of the measurement, not the end of the project.

Five myths that slow keyword research down

• "We need every keyword before we start." You need enough to fill the next two months. A full market map built before the first page ships is usually obsolete by the time it is finished.
• "High volume means high value." A 90,000 search term about a definition might send students. A 400 search term with a buyer behind it can carry a quarter. Sort by business value first, then by traffic potential.
• "One page can rank for everything." Pages that try to serve five intents serve none well. If two keywords need different answers, they need different pages.
• "Zero volume means no demand." Long B2B questions rarely show volume because the tools bucket them. If your sales team hears a question on three calls, it belongs on the board.
• "Research is a one-time project." The board is a standing queue. New questions arrive from support tickets, sales calls and search console queries every month.

From seed to schedule in six steps

This is the sequence we use on client accounts. It deliberately keeps tooling light: a sheet, the search engine results pages, and your own site search. Everything else is optional.

1. SeedWrite five to ten phrases a real buyer would type. Pull them from sales calls, not from imagination. Include one problem phrase, one product phrase and one supplier phrase.
2. ExpandFor each seed, collect modifiers: how, best, cost, supplier, alternatives, vs, for [industry]. Add the questions customers ask in email. Aim for breadth here, not judgement.
3. Tag intentMark each phrase as informational, commercial, transactional or navigational. The tag decides the page type: guide, comparison, product or quote page.
4. Read the results pageSearch the phrase and look at the top five. If they are all category pages, a blog post will not win. If they are all guides, a product page will not either.
5. Map to a URLAssign exactly one URL per phrase. If two phrases need the same page, they are variants, not separate targets. If two phrases need different answers, split them.
6. ScheduleOrder the mapped pages by revenue proximity and effort. Two commercial pages then one informational page is a healthy rhythm for most B2B sites.

Mapping keywords to pages without competing with yourself

One keyword, one page is the whole rule. Everything else is bookkeeping. When two pages chase the same phrase, search engines have to choose, and they usually choose the weaker of the two on any given day. The result is a page that jumps between positions six and fourteen for months while neither version earns links.

The fix is a mapping table that lives next to the board. Every URL appears once, with the primary phrase in one column and its close variants in another. If you cannot decide which URL owns a phrase, the site has a structural problem, not a keyword problem.

Notice the last row. A quote page is a legitimate target for a transactional phrase, and it is often the highest-value row in the table. Research that ignores the contact page tends to produce traffic that never converts.

If your team already has a backlog of overlapping posts, our Keyword Research Services start with a cannibalization pass: we list every URL that targets the same phrase, pick the winner, and set the others to merge, redirect or re-target. That single pass usually frees more growth than a new content sprint.

Once the map is stable, the writing brief gets easier. Each brief carries the primary phrase, the question it must answer in the first hundred words, the three sections a reader expects, and the two internal links the page will use. That brief is what On-Page SEO Services work from, which is why the two work best as one project rather than two.

What we actually deliver

A short example: twelve pages, one quarter

An industrial components supplier came to us with 180 blog posts and almost no organic enquiries. Their board had never existed; every post had been commissioned by whoever shouted loudest. We rebuilt the map first, which took nine days, and found 31 posts targeting phrases that already belonged to product pages.

We kept eleven posts, merged fourteen into four longer guides, and redirected the rest. Then we filled the board with 22 phrases from sales calls, tagged them, and scheduled twelve pages for the quarter. By the end of it, the merged guides were earning links that had previously been split four ways, and the quote page was receiving organic visits for the first time in two years. Nothing about the writing got faster. The targeting got clearer.

Questions we hear before a project starts