What Are Keywords in SEO? Types, Uses, and Examples

SEO gets sold like a magic trick: pick a phrase, sprinkle it on a page, wait for traffic. Then you try it, nothing moves, and you assume the whole thing is fake. We’ve been there. When people ask what are keywords in SEO, they usually mean “what words do I put on the page?” The more useful answer is: keywords are the labels we use for the queries real humans type (or say), and for the intent Google thinks sits behind those queries.

That difference sounds pedantic until you waste a month writing the “right” phrase while the top-ranking pages are winning because they answered a different job-to-be-done.

Keywords vs topics vs queries (and what Google is actually matching)

A keyword is not a spell. It’s a handle.

In practice, there are three layers:

A query is the literal text someone searches: “best backyard chicken breeds” or “what is considered reckless driving in Idaho.” A keyword is how we group and track that demand: we might call the keyword “backyard chicken breeds” and treat variations as part of the same target. A topic is the broader subject a page can credibly cover: “raising backyard chickens.”

What trips people up is thinking Google still runs on exact-match phrases. It doesn’t, not in the way beginners imagine. Google matches meaning, context, and intent across a set of related queries, then decides which pages satisfy the intent best. If you obsess over one exact wording, you end up blind to synonyms, adjacent questions, and the “People Also Ask” rabbit holes that are basically free editorial guidance.

One sentence we keep repeating internally: if two queries have the same intent and need the same kind of page, they are probably the same keyword target.

What are keywords in SEO today: picking a target when volume, difficulty, and intent disagree

This is the part nobody wants to talk about because it’s not a definition, it’s a decision. Keyword tools happily hand you thousands of terms with volume and competition labels. Then you stare at them and freeze.

The annoying part is that your three evaluation dimensions rarely line up:

Search volume tempts you toward head terms. Competition (or keyword difficulty) pushes you toward long-tail. Relevancy and intent pulls you toward whatever matches what your site actually offers.

We’ve watched smart teams pick the highest-volume term in the list, build a page, fail to rank, and then declare “SEO doesn’t work.” The real failure was choosing a keyword whose SERP was unwinnable for their authority level or whose intent didn’t match their page type.

Here’s the prioritization scorecard we use because “high volume” is not a strategy. For each candidate keyword, we give a simple 1 to 5 score in four buckets. No math wizardry, just forced clarity.

First: intent match. What is the searcher trying to do right now: learn, compare, buy, find a specific site, solve an immediate problem? We do not guess. We open the SERP and look at what Google is rewarding.

Second: business value. If we ranked, what would success look like for us: email signups, product trial, consultation request, store visit, ad revenue? This keeps you from chasing “interesting” traffic that never turns into outcomes.

Third: ranking feasibility. This is not the tool’s difficulty number alone. We look at the top results and ask: are these giant brands with deep link profiles, or are there smaller sites with focused pages? Is the content format something we can match or beat? If the SERP is full of government sites, major publishers, or the same entrenched players, you need a reason you will displace them.

Fourth: topical coverage potential. Can one page satisfy the cluster around this term without turning into a junk drawer? If we can cover multiple closely related queries with one strong page, the ROI goes up fast.

We learned this the hard way on a project where the “best” keyword by volume was also the worst fit. We chased the head term, wrote what we thought was the definitive guide, and it sat on page four. When we stopped sulking and actually read the SERP, it was obvious: Google wanted a comparison list with clear recommendations and quick criteria, not a philosophy essay. Our page was “good,” just wrong.

A worked example makes this real. Imagine you sell a niche home product. You see a head term with huge demand like “bedside table.” It looks like the promised land. It’s also vague, and the SERP is usually dominated by big retailers and category pages. Then you see a long-tail like “small bedside table with drawers” with a modest volume (one example floating around is 260 searches in an unspecified period). On paper, it looks tiny.

Now score them.

The head term has messy intent: some people want inspiration, some want to buy, some want dimensions, some want DIY plans. Your one page will struggle to satisfy all of that. The long-tail has a cleaner job: find a small table, specifically with drawers. That intent is tightly aligned to a product category or a curated buying guide.

Here’s what usually happens: the long-tail page ranks sooner, converts better because it matches a specific need, and pulls in a swarm of close variants you never explicitly targeted. Total conversions can beat the head term even if raw traffic is lower. Traffic is a vanity metric. Outcomes are not.

If you want a quick rule when you’re stuck between two options: pick the keyword where you can clearly describe the searcher’s next step after reading your page. If you can’t, you’re probably looking at an intent mismatch.

Anyway, we once built an entire content calendar off a tool export without opening a single SERP first. It was… ambitious. It was also mostly wrong.

Keyword types that matter in practice (not the taxonomy quiz)

Most guides split keywords into head (short-tail) and long-tail. That’s useful because it predicts competition and specificity. Head terms are usually 1 to 2 words, broad, high volume, and crowded. Long-tail terms are usually 3 or more words, specific, lower volume, often easier to rank for, and increasingly common with voice search.

Intent-based types are the second dimension that actually influences what you should publish. We treat intent labels as a content format decision, not an academic classification.

Informational queries want understanding: definitions, steps, explanations, checklists. “What does keyword mean” is a classic example style.

Commercial investigation queries want help deciding: comparisons, “best X,” pros and cons, alternatives, reviews. A Keyword Planner example you’ll see in agricultural extension materials is “best backyard chicken breeds,” with volume shown as a range like 100 to 1K monthly searches and low competition when targeted to a specific geography. The point isn’t the exact numbers, it’s that the same topic can look very different depending on location and framing.

Transactional queries want action: buy, book, download, sign up. These usually map to product pages, category pages, or landing pages.

Navigational queries want a specific site or brand. You rarely “target” these unless you are that brand, but they tell you what people already associate with you.

Where this falls apart: people dismiss long-tail because the volume looks small, then they bet everything on one head term that they cannot win yet. Long-tail is often the training ground where you build topical credibility, internal links, and real user engagement signals. It’s not “low value.” It’s often where the money is hiding.

Keyword research workflow that mirrors how humans actually search

Tools are helpful, but starting with a tool encourages the worst habit in SEO: chasing whatever has the biggest number.

We start with a seed list built from audience language. That means reading customer emails, sales call notes, support tickets, Reddit threads, and the exact phrasing people use when they are confused or frustrated. If your seed list sounds like internal jargon, you already lost.

Then we expand using the SERP itself because it is a live record of demand. We pull ideas from Google Autocomplete, People Also Ask, and Related Searches. These surfaces are brutally practical: they show the next questions a real searcher asks, not what your team wishes they asked.

After that, we validate and widen with tools like Google Keyword Planner, Ubersuggest, AnswerThePublic, or browser add-ons like Keyword Surfer. We use tools for two things: rough volume ranges and finding variants we missed. We do not treat any single tool as truth.

Competitor review is the final expansion pass, but it has a trap. You can input competitor URLs into tools and get keyword ideas, and you should. Just do not mirror their exact keyword set. If you only copy, you only tie. The better move is to find the questions they answer poorly or the angles they skip because they can’t afford to be specific.

Then we prioritize with the scorecard from earlier. The output is not a list of 500 keywords. The output is a shortlist of page targets that a human can actually build.

From one keyword to many: clustering and mapping without the one-page-per-keyword trap

If you try to create a unique page for every keyword, you drown. If you cram every keyword onto a few pages, you confuse search engines and users. WordStream makes this point bluntly: sites can rank for thousands or millions of keywords, so you need grouping and mapping.

The messy middle is building clusters that reflect intent, not just wording.

Our repeatable method is simple, and it’s not glamorous. We group by shared intent and required content type.

Intent is the user’s goal. Content type is what the SERP expects: a guide, a category page, a tool, a definition, a comparison list, a local service page.

We collect candidate queries, then we open the SERP for each and label what Google is rewarding. If the top results for two queries are the same kind of pages, they can probably live together. If one SERP is dominated by “best” listicles and the other is dominated by ecommerce category pages, trying to combine them usually produces a page that ranks for neither.

A quick heuristic for “separate page vs same page” that saves arguments: split the content if the searcher would need a different outcome. Consolidate if the outcome is the same and the SERP page types are the same.

Keyword cannibalization is the silent killer here. You publish two posts that both target the same intent, both kind of match the same terms, and Google keeps swapping which one it shows. Rankings wobble, clicks get split, and you blame the algorithm. It’s usually your site structure.

We keep a lightweight mapping template in a shared doc. No tables, just a repeating block per page target: Page name, primary term, secondary variants, intent, call to action, and the internal links it should point to and receive. It’s boring. It works.

When we assign primary vs secondary terms, we do not obsess about exact matches. We pick a primary term that reflects the main intent, then we list secondary variants that represent the same job in different language. Those secondary terms become sections, FAQs, examples, or subheadings, not separate pages.

Where keywords go on a page (and how to do it without sounding broken)

On-page placement matters because it helps search engines and humans confirm what the page is about fast. High-attention areas include title tags, meta descriptions, headers, body content, URLs, and even image file names when relevant.

We use a simple discipline: the primary keyword or a close variation belongs in the title tag and usually in the main on-page heading, assuming it reads like something a human would write. Then we write the page to satisfy the intent, and we let secondary terms show up naturally as we cover the required subtopics.

Meta descriptions are a common misunderstanding. They don’t typically “rank” the page by themselves. They influence clicks. If your snippet reads like keyword soup, your click-through rate drops even if you rank.

Keyword stuffing is still a real failure mode. We’ve seen pages get worse when someone forces exact-match into every header. The copy becomes awkward, users bounce, and the page sends terrible engagement signals. The fix is annoyingly simple: write for comprehension first, then do a pass to make sure key terms appear where they clarify meaning.

If you want one practical check: read your opening paragraph out loud. If you feel embarrassed, you stuffed.

Why keyword strategies stall (and what to check before you rewrite everything)

When rankings don’t move, the instinct is to change the keywords. Half the time the keyword is fine. The page is not.

Moz points out a common plateau cause: you lack perceived authority or relevance compared to the pages already ranking. That can mean weaker links, thinner topical coverage, or a brand that Google hasn’t seen as trustworthy in that niche.

Another trap is manual rank checking. Personalized results, location, search history, and device all skew what you see. If you search your own keyword ten times a day, you are not measuring anything. Use a rank tracker over time or at least check in a clean browser with location settings controlled.

Intent mismatch is the most brutal stall. You can have great writing and still lose because the SERP wants a different format. If the top results are “definition” pages and you wrote a product category, you are swimming upstream.

The pro move: design for outcomes, then use keywords as proof you understood the job

Most “keyword optimization” advice is backwards. It tells you to pick a phrase, then write around it. We get better results when we design the page around the outcome first, then verify we covered the language people use.

Here’s the SERP intent deconstruction checklist we actually run, especially for high-stakes pages. We open the top results and look for three things: dominant content type, dominant angle, and required sections.

Content type is what kind of page wins: guide, list, tool, category, local service page. Angle is how it’s framed: “for beginners,” “cheap,” “2026,” “near me,” “step-by-step.” Required sections are the repeatable blocks that show up across winners: pricing, step lists, comparison criteria, templates, pros and cons, FAQs.

Then we translate that into an outline that earns the click and the stay. If the winning pages all include a comparison table or clear criteria, and we don’t, we’re not “different,” we’re incomplete. Different is only good when it is better.

Keyword proximity gets thrown around like a hack, so here’s our cautious take. Putting related terms near each other can help clarity because it signals the relationship between concepts. Treat it as a writing tactic, not a ranking trick.

Before: “Keywords help SEO. You add them to your page to rank. Search engines look at content.”

After: “SEO keywords help search engines understand what your page is about, especially when the keyword appears near the related topic terms your audience also uses. Your goal is clarity, not repetition.”

Notice what changed: we didn’t cram the phrase five times. We made the sentence more specific. That’s usually what “optimization” looks like in real life.

If you only take one thing from this: stop trying to win by finding the perfect phrase. Win by matching intent, choosing feasible battles, and building pages that cover a topic thoroughly enough that dozens of related queries make sense for Google to send to you.

FAQ

What is an example of a keyword in SEO?

A keyword could be "backyard chicken breeds." It represents a group of similar searches like "best backyard chicken breeds" or "good chicken breeds for beginners" that share the same intent.

What are the four types of keywords for SEO?

Informational, navigational, commercial investigation, and transactional. These types describe why someone is searching, which usually determines what kind of page will rank.

How do I know if two keywords should be on the same page or separate pages?

Check the SERP for both queries. If Google rewards the same content type and the searcher wants the same outcome, combine them. If the SERPs expect different page types or different outcomes, split them.

What is the 80/20 rule for SEO keywords?

Spend most of your effort on the small set of keywords and pages most likely to drive outcomes, not on a massive list of low-priority terms. In practice, that means focusing on the best intent match and feasible wins first.