Keyword Research: How to Find the Terms Your Buyers Actually Use

Keyword research is the process of finding the words and questions your buyers type into search engines, and deciding which of them your site should try to rank for.

It matters because companies rarely describe their product the way customers do. Your team might say "workforce payment automation". Your buyer types "payroll software for small business". If your pages use only the first phrase, they will not appear for the second.

Good keyword research closes that gap. It tells you what to write, what to call it, and in what order to do the work.

Key takeaways

• Keyword research is how you learn the exact language buyers use, which is often different from the language used inside your company.
• The best seed keywords come from customers: sales calls, support tickets, reviews and the search queries already reported in Google Search Console.
• Every keyword should be judged on three things together: how many people search for it, how hard it is to rank for, and what the searcher wants.
• Related keywords should be grouped so that one page targets one topic, instead of many thin pages chasing small variations.
• A keyword list is only useful once it becomes a prioritised plan that says which page to create or improve next.

What is a keyword, and what makes one worth targeting?

A keyword is any word or phrase someone types, or speaks, into a search engine. A keyword is worth targeting when enough of the right people search for it, you have a realistic chance of ranking, and a visit could lead to business.

Keywords are often described by length:

• Head terms are short and broad, such as "payroll". Many people search them, competition is heavy, and it is hard to tell what the searcher wants.
• Long-tail keywords are longer and more specific, such as "payroll software for restaurants with tipped staff". Fewer people search each one, but the need is clear and the competition is usually lighter.

For most businesses, long-tail keywords are the practical place to begin. Each one is small, but there are a great many of them, and the people searching tend to be closer to a decision.

How do you find keyword ideas?

Start with what your customers already say, then use tools to widen the list. Tools are good at expanding ideas but poor at knowing your market, so the first inputs should come from you.

1. Collect customer language. Read sales call notes, support tickets, reviews and chat logs. Write down the phrases people use to describe their problem, word for word.
2. Check what you already appear for. The performance report in Google Search Console lists queries that showed your pages. Look for queries with many impressions and few clicks: these are topics where you are visible but not yet convincing.
3. Write seed keywords. From the first two steps, pick ten to twenty short phrases that describe your products, the problems they solve and the people who use them.
4. Expand with a keyword tool. Put each seed into a tool such as Semrush or Ahrefs to see related phrases, questions and variations.
5. Look at the results page. Search your seeds and note the autocomplete suggestions, the "People also ask" questions and the related searches. These are free and reflect real behaviour.
6. Study competitors. See which topics rival sites cover that you do not. Treat this as a list of gaps to consider, not a list to copy.

By the end you should have a long, messy spreadsheet. That is expected. The next step is to judge it.

How do you judge which keywords to target?

Judge each keyword on demand, difficulty and intent together. A keyword that scores well on one and badly on the other two is rarely worth a page.

Two cautions apply. First, volume and difficulty figures are estimates produced by each tool, and they differ between tools. Use them to rank options, not as facts. Second, a low-volume keyword can still be valuable if the people searching it are ready to buy.

Intent deserves the most attention. If the results for a term are all comparison articles, a product page is unlikely to rank there, however well it is written. Our guide to search intent explains how to read a results page for these signals.

How do you group keywords into pages?

Group keywords that mean the same thing, or that the same page would satisfy, and assign each group to one page. This is often called keyword clustering.

Search engines understand that "how to run payroll" and "how do I do payroll" are the same question. Writing a separate page for each splits your effort and makes your own pages compete with one another.

A simple test: search two keywords and compare the results. If mostly the same pages appear for both, they belong on one page. If the results differ, they need separate pages.

For each group, choose:

• A primary keyword, the clearest and most searched phrase in the group, which goes in the title and main heading.
• Secondary keywords, the variations and sub-questions, which become headings and sections within the page.

Then map each group to a URL. If a page already exists for that topic, improve it. If none exists, plan a new one. This mapping also shows you where pages should link to each other, which supports building topical authority over time.

How do you prioritise the list?

Prioritise by combining business value with your realistic chance of ranking, then start with the quickest wins. A plan you can deliver beats a perfect list nobody acts on.

A workable order:

1. Pages that already rank on the second page of results. A modest improvement can move them into view.
2. High-value, low-difficulty topics with no page yet. These are the clearest gaps.
3. Supporting topics that answer earlier questions buyers have, and that link to your main pages.
4. Competitive head terms. Leave these until the supporting content exists.

To check it is working, track impressions and average position for each target group in Google Search Console. Impressions usually rise before clicks do. Review the list every few months, because the way people search changes.

Frequently asked questions

How many keywords should one page target?

One page should target one primary keyword and a handful of closely related secondary keywords that share the same intent. If a keyword would need a different kind of answer, it belongs on a different page.

Are long-tail keywords worth the effort?

Yes, long-tail keywords are usually worth targeting because they are specific, less competitive and often searched by people close to a decision. Each one brings few visits, but together they can make up a large share of a site's search traffic.

Do you need paid tools for keyword research?

No, you can do useful keyword research without paid tools. Google Search Console, autocomplete suggestions and "People also ask" questions reveal real queries. Paid tools mainly save time and add estimates of volume and difficulty.

How often should you repeat keyword research?

Keyword research should be revisited every few months and whenever you launch a new product or enter a new market. Search behaviour shifts, new questions appear, and your own site becomes able to compete for harder terms.

Should keywords appear a set number of times on a page?

No, there is no correct number of times to repeat a keyword. Use the primary keyword in the title, main heading and opening, then write naturally. Repeating a phrase unnaturally makes a page worse to read and does not help it rank.

Keyword research is less about finding secret phrases and more about listening carefully and choosing well. If you want help turning a messy list into a clear plan, start a conversation.