What keyword research is
Keyword research is the practice of finding out how people describe what they want when they search, and then using that wording to decide what your pages should cover. A “keyword” is not magic. It is simply a phrase, and often a whole question, that someone types or says into a search engine.
The goal is not to find secret words. It is to understand your audience’s vocabulary. A bakery owner may say “celebration cake” while customers search “birthday cake near me.” A site that uses only the owner’s wording can miss people who are looking for exactly what it offers.
Start with what you already know
You do not need software to begin. Good sources are closer to home:
- The questions customers or readers ask you by email, phone or in person.
- The words they use in reviews, comments and messages.
- Your own products, services and expertise, written in plain language.
- Pages and categories already on your site, which show what you consider your topics.
- If you have a search reports page set up, the queries that already bring visitors.
Write each idea as a short phrase. After ten minutes you may have thirty or forty, most of them genuinely relevant, because they came from real people.
Expand the list carefully
Once you have a base list, you can widen it. Type a starting phrase into a search box and look at the suggestions that appear as you type, and at the related questions and searches shown with results. Both give a rough picture of how other people phrase the same need. Free and paid research tools can also suggest related phrases and give estimated search volumes, but the estimates are approximate and vary between tools, so use them to compare ideas rather than as exact counts.
Be careful not to expand for the sake of it. A long list of loosely related phrases leads to thin pages written only to match words. Keep what you can actually cover well.
Sort by what people want
Two phrases that look similar can need completely different pages. Someone searching “how to clean a cast iron pan” wants instructions. Someone searching “cast iron pan” may want to buy one or may be browsing for ideas. Group phrases by the underlying need, which is the subject of search intent, and decide what type of page each group deserves: a guide, a product page, a category page or a short answer.
A useful trick is to look at what already appears for a phrase. If the top results are all how-to guides, a sales page is unlikely to fit, and the reverse is also true.
Choose realistic targets
Popular phrases attract the most competition. A small site is usually better served by narrower, more specific phrases, sometimes called long-tail phrases, where the need is clear and the field is less crowded. “Watering a snake plant in winter” is more specific than “plant care,” and a single useful page can answer it completely.
Some questions to apply to each candidate:
- Does this fit what my site is about and can I add something useful?
- Do I have real knowledge or experience on the topic?
- Is there a clear need behind the phrase?
- Can I cover it in one page, or does it need several?
- Is it the same as a page I already have?
That last one matters more than it seems. Writing two pages for nearly the same phrase makes them compete and may create duplicate content problems. If a page already exists, improve it instead. Looking for what your site is missing is the work of a content gap review, and keyword ideas feed neatly into it.
Map one main topic to each page
A simple spreadsheet is enough. Give each page a row with a main phrase, a few close variations, the type of need and a note on its status. Use the main phrase in the title tag and the heading naturally, then write the page for people. Variations tend to appear on their own when you answer the topic properly. Forcing them in leads to the problems described under keyword stuffing.
Check results without chasing them
After pages have been live for a while, look at which queries actually bring visitors. You may find that people reach the page through phrases you never planned for, which is useful information for the next update. Keep expectations modest: rankings take time, depend on many factors and cannot be promised. If a page does not perform, ask whether it matches the need before assuming the keyword was wrong.
Common beginner mistakes
- Picking phrases only because they have high volume.
- Targeting a phrase the site has no real answer for.
- Using one phrase for several near-identical pages.
- Ignoring the way customers talk in favor of industry jargon.
- Treating research as a one-time task. Language and interests shift, so a quick yearly review is reasonable.
Keyword research should feel like listening. If your list reflects real questions from real people, and you plan one useful page per need, you are doing it well, whatever tools you use.