SEO guide · 9 min read

How to do keyword research

Every piece of content should start here. This is how to find the terms and questions worth targeting, judge them properly, and map them so your pages never compete with each other.

On this page
In one sentence

Keyword research is finding the words, phrases and questions your audience actually searches for, then judging which are worth targeting. In 2026 it is not just about search volume: it is about matching the right intent, and increasingly about the questions people ask AI answer engines, not only the keywords they type into Google.

What is keyword research?

Keyword research is the practice of identifying what your audience searches for, and deciding which of those searches you should try to win. It is the foundation of both SEO and content strategy: you cannot optimise for demand you have not identified. Today it spans two things at once, the keywords people type into Google and the questions they ask answer engines like ChatGPT and Google AI Overviews.

Why does keyword research matter?

Without it, you are guessing. Skip the research and you end up writing content nobody searches for, competing for terms you cannot realistically win, or targeting the wrong intent so the traffic never converts. Done well, it points your effort at demand that actually exists and that you have a chance of capturing. It also underpins the keyword map that keeps your pages from competing with each other, which we come to below.

How do you do keyword research?

A repeatable process beats a big one-off list:

  1. Start with seed terms. The core topics of your business, described in your customers' words, not your internal jargon.
  2. Expand them. Use Google autocomplete, the "people also ask" and related searches on the results page, and, best of all, the queries you already earn impressions for in the Search Console Performance report.
  3. Add data. Rough search volume and difficulty help, but weigh them against relevance to your business.
  4. Judge intent. Work out what each searcher actually wants, covered next.
  5. Prioritise. Relevance, intent and winnability first. A term you can win that matches a buyer beats a bigger term you cannot.

The fastest wins are often hiding in your own Search Console data: terms you already rank for on page two, where a better-targeted page could push you up. Many of your best opportunities will be long-tail, the longer, more specific phrases; see how to find long-tail keywords for that in depth.

How do you read search intent?

Search intent is what the person is actually trying to do, and it comes in four broad types: informational (to learn), commercial (to compare), transactional (to buy) and navigational (to reach a specific site). Each wants a different kind of page, and targeting a keyword with the wrong type of page is one of the most common reasons content fails to rank. For how to work out a keyword's intent and match your page to it, see how to read search intent.

How do you research questions for AI search?

This is the part traditional keyword research misses. Answer engines are driven by questions, so your research should capture the questions your customers ask, not only the keywords they type, then turn each into a section or page that answers it directly, the core move in Answer Engine Optimisation. For where to find those questions, how to prioritise them and how to turn them into content, see how to research questions for AI search.

How do you map keywords to pages?

Research is only half the job; the other half is assignment. Give each page one primary target, and write it down. This keyword map is the single best defence against keyword cannibalisation, where two of your own pages compete for the same term and neither wins. Group closely related terms under the page that owns their intent, and never split one intent across several pages. A little mapping discipline up front saves a lot of untangling later.

How does Limecube help?

Limecube's keyword workspace is where this whole process actually happens, per client, not in a spreadsheet you forget. You do the research in the tool: build and expand your keyword lists, and surface the questions your customers ask. It flags the questions worth answering for AI visibility and scores how well your existing pages cover them, so the gaps are obvious.

It also does the mapping for you: every keyword is assigned to the page that should own it, and when a page has drifted onto too many terms, the tool suggests which existing page each one really belongs to, so you can re-map and head off keyword cannibalisation before it costs you rankings. Research, questions and mapping in one place, which is exactly what stops your content competing with itself. The 7-day trial covers it.

Frequently asked questions

  • What is a seed term in keyword research?

    A seed term is a starting keyword: a broad, core word or phrase for one of your main topics, which you then expand into a fuller list. If you sell running shoes, "running shoes" is a seed term, and the research grows it into hundreds of more specific, longer-tail variations and questions.

  • How many keywords should I target per page?

    One primary target, plus a small cluster of closely related terms that share the same intent. Trying to target many unrelated keywords with one page weakens it, and spreading one intent across several pages causes keyword cannibalisation. One clear job per page is the rule.

  • Do I need a paid keyword tool?

    Not to start. You can get a long way with free sources: Google autocomplete, the "people also ask" and related searches on the results page, and your own Search Console queries. A paid tool helps with volume and difficulty at scale, but the thinking, intent and relevance, matters more than the tool.

  • What matters more, search volume or intent?

    Intent and relevance, by a distance. A high-volume keyword with the wrong intent, or one you have no realistic chance of ranking for, is wasted effort. A lower-volume term that matches what your buyers actually want, and that you can win, is worth far more.

  • How is keyword research different for AI search?

    The unit changes from keywords to questions. Answer engines respond to questions, so you research the questions your customers actually ask, not just the phrases they type into Google. Each question becomes something a section or page can answer directly, which is the heart of Answer Engine Optimisation.

  • How do I stop targeting the same keyword on several pages?

    Keep a keyword map: assign one primary target to each page, written down, so every page has a clear job. That single habit prevents most keyword cannibalisation before it starts. If it has already happened, our guide on fixing cannibalisation walks through the repair.

Related guides

Ready to ship more
without hiring more?

Start the 7-day trial: the full Professional plan, no credit card. Add a client and put the whole workspace to work, from AEO content and AI visibility tracking to Google Ads, site audits and reporting.