Keyword Research

Keyword research is the process of understanding what your audience searches for, and using that understanding to shape content, architecture, and priorities across an SEO strategy.

What is keyword research?

Keyword research is the practice of identifying the search queries people use when looking for information, products, or services relevant to your site. It answers two questions: what are people searching for, and what do they actually want when they search it?

Done well, keyword research produces more than a list of terms to target. It reveals the shape of your audience’s intent. That insight drives content decisions, site structure, internal linking, and prioritisation. Every other pillar of SEO depends on it.

The keyword spectrum

Keywords sit along a spectrum from broad to specific.

Head terms (also called seed keywords) are short, one- or two-word queries with high search volume: “SEO”, “keyword research”, “local SEO”. They are rarely worth targeting directly; competition is fierce, intent is ambiguous, and a single page cannot satisfy the full range of searchers a head term attracts. Their primary use in keyword research is as starting points: entering a head term into a keyword tool generates the broader range of related queries to assess.

Mid-tail keywords add one or two words of qualification: “keyword research tools”, “local SEO for restaurants”. Lower volume, clearer intent, more achievable competition.

Long-tail keywords are specific multi-word queries with low individual volume but high collective importance. They account for the majority of all searches and are the main opportunity for most sites. See long-tail keyword strategy for how to capture them at scale.

Core keyword research elements

• Building a keyword universe. Assembling the complete set of search queries relevant to your business from multiple discovery sources, before any filtering or prioritisation.
• Search intent. The underlying goal behind a query: informational, navigational, commercial, transactional. Matching content type to intent is the most important factor in whether a page can rank.
• Search volume explained. What monthly volume figures mean, where they come from, and where they mislead.
• Keyword difficulty. How tools estimate ranking difficulty, what those scores actually measure, and when to trust them.
• Long-tail keyword strategy. Specific, lower-volume queries that collectively represent the majority of all searches and tend to convert better.
• Topic clusters and pillar pages. Grouping semantically related keywords around a central theme rather than chasing terms in isolation.
• SERP analysis. Reading the results page to understand what Google believes satisfies the intent of a query.
• Competitor gap analysis. Identifying queries your competitors rank for that you don’t.
• Keyword mapping. Assigning target queries to specific pages to prevent cannibalisation and surface content gaps.
• Keyword cannibalisation. When multiple pages compete for the same query, how to detect it, and when to consolidate, differentiate, or remove.
• SERP features. The feature types that appear alongside organic results, how each affects CTR, and how to target featured snippets and AI Overview citations.

Why does keyword research matter?

Without keyword research, SEO is guesswork. You can produce content that is technically excellent and well-linked, but if it doesn’t align with how your audience actually searches, it won’t be found.

Keyword research also determines where to invest. Some queries are too competitive for your current domain authority. Some have intent that doesn’t match your offering. Some have volumes too low to justify the effort. Good research filters for opportunity as well as relevance.

For sites targeting more than one country or language, this process must be repeated per market. International keyword research treats each language and country as a separate exercise: search volume, dominant phrasing, and user intent differ significantly across markets, even for the same underlying topic. International SEO mistakes often begin here: applying source-market terms without checking local usage. A multilingual content strategy built on locally researched keywords consistently outperforms one built on translations of source-market terms.

Keyword research and AI search

AI engines retrieve passages that answer specific questions. Long-tail, question-shaped queries map more cleanly onto retrievable passages than head-term keywords, which is why long-tail strategy matters more, not less, in the AI search era. The shape of demand has also changed: queries that were once two- or three-word fragments are increasingly typed in full sentences, especially in conversational interfaces.

Keyword research and E-E-A-T

Keyword research shapes which topics you cover and how deeply. Sites that build in-depth, expert-level content around a focused set of topics signal topical authority more effectively than sites covering everything thinly. That concentration of depth is what earns E-E-A-T signals over time: it is harder to be the best source on ten focused topics than a mediocre source on a hundred.

What about LSI keywords?

LSI (Latent Semantic Indexing) keywords are sometimes promoted as a keyword tactic: identify “semantically related” terms and work them into your copy to trigger Google’s understanding of the topic. The term comes from a 1980s information retrieval algorithm. Google representatives have stated publicly that Google does not use LSI.

The underlying instinct is sound: writing about a topic fully and naturally will include related terms, and thorough content performs better than thin content. But LSI as a tactic, selecting a checklist of co-occurring terms to force into a page, does not reflect how Google’s systems work. Google uses neural language models and entity recognition. Write for the topic, not for a related-keyword checklist.