What Is a Keyword in SEO? (And How GEO Changes the Rules)
A keyword is the query a customer types — but what search engines do with it has changed. Intent now matters more than exact phrasing, long-tail terms convert better than head terms for specialised niches, and AI engines do not count keyword density at all: they extract answers. GEO reframes the discipline from "rank for the phrase" to "be the citable answer for the question."
In this guide
- The 4 types of keywords by search intent
- Head terms vs long-tail — where new sites actually win
- The semantic shift: why density is dead and answers win
- FAQ: the 4 intent types, long-tail value for new sites, density vs answers
In traditional SEO, a keyword (or keyphrase) is the exact word or phrase a user types into a search engine like Google. Talking about keywords means talking about the bridge between what people search for and the content your website provides to answer that need.
A keyword exists in two places at once: the user's actual search query, and the words embedded in your website's text, titles, and meta tags. Google's crawler reads your page, identifies the main topics (keywords), and matches it to user queries.
Core Principle: There's no "rank #1 for a keyword" concept in GEO — your goal is to be the source an AI retrieves when answering, driven by semantic relevance and entity clarity, not keyword density.
The 4 Types of Keywords (Search Intent)
Google no longer just matches words — it matches intent. SEO categorizes keywords based on what the user is actually trying to achieve:
| Type | Keywords | Target content |
|---|---|---|
| Informational | "what is geo", "how to improve ai visibility" | Blog posts, guides, whitepapers |
| Navigational | "visora login", "foihk" | Your homepage or a dedicated landing page |
| Commercial | "profound vs visora", "best geo tools hong kong" | Comparison pages, listicles, case studies |
| Transactional | "buy visora credits", "ai visibility scanner free" | Pricing pages, free trial sign-ups |
Head Terms vs. Long-Tail Keywords
Keywords are also categorized by their length and specificity:
| Type | Example | Difficulty |
|---|---|---|
| Head terms (short-tail) | "SEO" or "Marketing" | Very hard to rank for |
| Long-tail keywords | "generative engine optimization tool for restaurants in hong kong" | Easier to rank for, brings highly qualified leads |
How Search Engines Use Keywords Today (The Semantic Shift)
Ten years ago, SEO meant "stuffing your page with the exact keyword 50 times." Today, search engines use Semantic Search — Google reads your page, maps the words into an embedding space (turning words into math), and understands meaning, synonyms, and context.
If your target keyword is "GEO Hong Kong," Google also understands that "Generative Engine Optimization services in HK" and "AI search visibility Asia" mean the exact same thing. You don't need to repeat the exact phrase — you just need to cover the topic comprehensively.
How This Connects to GEO (Generative Engine Optimization)
In the AI search world (ChatGPT, Perplexity, Gemini), keywords matter differently:
- AI engines do not track "search volume" or "keyword difficulty" the way Google does.
- AI engines retrieve information based on semantic relevance and entity clarity.
- Therefore, in GEO, your goal isn't to "rank #1 for a keyword" — it's to be the source the AI retrieves when it needs to answer a user's prompt. You achieve this through structured data (Schema.org), authoritative citations, and clear, factual content.
Frequently Asked Questions
Go Deeper
- Lexical vs Semantic Search Explained — the engine shift that changed what ranks
- What Is Natural Language Search? — writing for how customers actually prompt AI
- The 7 Metrics Behind Your AI Visibility Score — what gets measured after the answers land
- AIV Methodology — the public scoring model
Want to know your business's AIV score?
Free scan — see in under a minute whether ChatGPT, Gemini, and Perplexity recommend you.
Last updated: