How to find keywords customers actually search for

High search volume feels like a win and behaves like a trap. A term forty thousand people search every month is usually forty thousand people who aren’t ready to buy anything — students, browsers, the merely curious. You can win that ranking, celebrate the traffic spike, and watch your sales stay exactly where they were. The keywords that actually grow a small business are quieter, and the whole skill is learning to tell the two apart.

This is a practical, end-to-end way to find the terms your customers really search — not the ones a tool tells you are popular, but the ones that bring people who could buy. It comes down to five moves: understand intent, gather real phrases, cluster them, score for winnability, and write for the searcher. None of it requires expensive software or an SEO background, and you can run the whole thing in an afternoon.

Start with intent, not volume

Every search has an intent behind it — a reason the person typed those words. Sort any keyword list into three buckets and the right targets start to separate themselves from the noise.

• Informational — they want to learn (“how does SEO work”, “what is a sales funnel”)
• Commercial — they’re comparing options (“best email tool for founders”, “Mailchimp vs ConvertKit”)
• Transactional — they’re ready to act (“hire a fractional CMO”, “buy standing desk”)

Informational terms build trust and traffic; they’re how strangers first find you. Commercial and transactional terms build revenue; they’re how those strangers become customers. A healthy plan includes all three, but the order matters. If you’re early and time-poor, weight toward the bottom two — that’s where the buyers are, and a page that ranks for “best CRM for plumbers” is worth ten that rank for “what is a CRM”.

The mistake most small businesses make is writing almost entirely informational content because it’s easier and the volumes look bigger, then wondering why all that traffic never converts. Traffic isn’t the goal. The right traffic is. A useful exercise: for every informational post you plan, ask which commercial or transactional page it’s meant to feed. If the answer is “none,” you may be writing for an audience that will never buy.

Where to actually find the terms

You don’t need a keyword tool to start — you need to listen where your customers already talk. Google’s autocomplete shows you real phrases as you type a seed term. The “people also ask” box reveals the follow-up questions searchers have. The “related searches” at the bottom of the results page maps the neighbourhood around a topic. Each of these is Google handing you real demand for free, no subscription required.

Then mine your own inputs, which are richer than any tool. The questions in your support inbox, the words customers use on sales calls, the threads in the communities your buyers hang out in — these are keywords in disguise, phrased exactly the way a real person would search. If three customers this month asked the same question in nearly the same words, that’s a page waiting to be written, and you already know it has demand because real people asked it.

A quick way to build a starter list: write down ten seed terms — the obvious words for what you do — and run each through autocomplete and related searches, copying anything relevant into a single document. Twenty minutes of this produces dozens of genuine phrases, and unlike a tool’s export, every one of them came from a real human typing into a real search bar.

Cluster, don’t collect

A long list of individual keywords is overwhelming and leads to scattered, thin content. Instead, group your terms into clusters around a single problem. One cluster becomes one cornerstone page plus a few supporting posts that link to it — far more powerful than ten disconnected articles, because depth on a topic is what earns rankings, and internal links pass authority between related pages.

Clustering also fixes a quieter problem: keyword cannibalisation, where two of your own pages compete for the same term and neither wins. When you plan by cluster, each page owns its own angle, and the cornerstone owns the head term. The result reads like a section of genuine expertise instead of a pile of one-offs, and Google rewards sites that demonstrate depth over those that scatter shallow posts across unrelated topics.

Map each cluster to the right page type as you go. Head terms and big questions become cornerstone guides. Comparison and “best” terms become commercial pages with a clear recommendation. Narrow how-to questions become short, focused supporting posts. Doing this mapping up front means you’re never staring at a keyword wondering what kind of page it deserves.

Score for winnability

A perfect keyword you can’t rank for is a daydream. For each cluster, ask two honest questions: is the intent a genuine fit for what I sell, and can a site my size realistically compete on the first page? You’re looking for the overlap between “people who could buy” and “searches I can actually win” — usually more specific, lower-volume, higher-intent terms than your instinct first reaches for.

A simple way to gauge difficulty without a tool: search the term and read the first page like a competitor. If it’s wall-to-wall huge brands and decade-old definitive guides, move on for now. If you see thin pages, forum threads, or results that don’t quite answer the question, that’s a gap you can fill. Note who ranks, how thorough they are, and where they’re weak — that reconnaissance is your brief for beating them.

The best early keywords are the ones where the current top result is clearly beatable — where you can be more useful, more specific, or more current than what’s there. A page that ranks fifth today for a winnable term is worth far more of your time than a fantasy ranking for a term owned by Wikipedia and three Fortune 500s.

Write for the searcher, not the engine

Once you’ve picked a term, the job is simple to state and hard to fake: answer the question better than the result that currently ranks — faster, clearer, with the specific detail a real buyer needs and the current top page is missing. Rankings follow usefulness more reliably than they follow any trick. Front-load the answer, use the searcher’s own words in your headings, and include the practical specifics (numbers, steps, examples) that turn a generic article into the one people bookmark.

Match the format to the intent, too. A “how to” query wants clear steps; a “best” query wants an honest comparison with a recommendation; a “what is” query wants a crisp definition followed by useful depth. Giving searchers the shape they expect is half of satisfying them, and satisfied searchers — the ones who don’t bounce back to Google — are the strongest ranking signal you can send.

Chase the buyer and the traffic tends to follow. Chase the traffic — the big, vague, high-volume terms — and you often get neither the ranking nor the customer.

When a keyword tool actually helps

Everything so far works with free signals and your own judgment, which is exactly where a beginner should start — tools are dangerous before you understand intent, because they tempt you to chase the biggest numbers. But once you can read intent and difficulty by eye, a keyword tool earns its place by adding scale and rough volume estimates to the instincts you’ve already built. It turns a handful of seed terms into hundreds, and it puts a ballpark figure next to each so you can prioritise.

Use it as a multiplier, not a crutch. Feed in the clusters you found by listening to customers, let the tool expand them, and then apply the same intent-and-winnability filter to the longer list. The danger is letting a tool’s volume column override your judgment about who’s actually searching and whether they could buy. A term with a thousand searches and the wrong intent is still the wrong term, no matter how confidently a dashboard recommends it.

Mine the keywords you already rank for

One of the richest sources of new targets is the set of terms you already rank for without trying. Open Search Console, sort your queries by impressions, and look for searches where you appear on page two — close enough that Google already considers you relevant, far enough that you’re getting almost no clicks. Each of those is a keyword you’ve half-won, and a small push often takes it onto page one.

This is usually faster and more reliable than chasing brand-new terms from scratch. The hard part of ranking — earning some authority and relevance for a topic — is already done; you’re just nudging an existing page over the line, or writing one focused piece to claim a term you’re accidentally adjacent to. Always look at what you almost rank for before you go hunting for something entirely new.

A quick word on long-tail terms

Beginners obsess over the big head terms — the one or two-word searches with enormous volume — and ignore the long tail, which is where small sites actually win. Long-tail keywords are the longer, more specific phrases (“bookkeeping software for self-employed electricians” rather than “bookkeeping software”). Each one has modest volume, but there are thousands of them, the intent is sharper, and the competition is a fraction of what the head term attracts.

Add up enough long-tail wins and you have a serious amount of high-intent traffic — built from terms the big players didn’t bother to target. As a bonus, ranking well for a cluster of long-tail phrases is often how you eventually start ranking for the competitive head term above them, because Google has watched you satisfy searcher after searcher on the topic. Start narrow and specific; the breadth follows.

A repeatable 30-minute routine

Make this a habit, not a one-off project. Once a week, pick one customer problem, gather a cluster of real phrases from autocomplete and your inbox, check what you already half-rank for, sort them by intent, eliminate the ones you can’t win, choose the single best target, and note the page type it deserves. That’s your next cornerstone, fully briefed before you write a word.

Do it weekly and within a couple of months you’ll have a content plan built entirely from terms your customers actually search — and a blog that brings buyers instead of bounces. The teams that win at SEO aren’t the ones with the best tools; they’re the ones who consistently point their writing at winnable, high-intent searches. This routine is how you become one of them.