Keyword research has a reputation for requiring expensive tools. For a small site, the most valuable parts of it require none, and the expensive tools mostly help with scale rather than insight.

Start with intent, not volume

The most common mistake is picking terms by search volume. A phrase with 50,000 monthly searches is worthless to you if the people searching it want something you do not have, and if the first page is occupied by sites a hundred times your size.

Search intent divides roughly four ways, and the type determines everything about what a page needs to do:

• Informational: "why is my sourdough dense". The searcher wants an explanation.
• Navigational: "royal mail tracking". They want a specific site and nothing else will do.
• Commercial: "best stand mixer under 300". They are researching a purchase but not buying yet.
• Transactional: "buy kitchenaid artisan 5qt". They have decided.

The way to determine intent is not to guess. Search the phrase and look at what is ranking. If the first page is all product listings, Google has concluded the intent is transactional, and an explanatory article will not rank there no matter how good it is.

Long-tail terms are where a small site can win

"Coffee" is unwinnable. "How to stop a moka pot tasting bitter" is winnable, gets fewer searches, and reaches someone with a specific problem you can solve.

This is not a compromise. Longer queries convert better precisely because they are specific, and a hundred visitors with an exact problem are worth more than ten thousand browsing a general term.

A useful rule of thumb for a new site: if the first page of results for a phrase is entirely major brands and established publishers, pick a more specific phrase. You are not going to displace them this year, and ranking eleventh earns nothing.

Free ways to find terms

• Google autocomplete. Start typing your topic and read the suggestions. These are real queries people type.
• People Also Ask. The expandable questions in results are a direct list of what searchers want to know, and each one expands into more.
• Related searches at the bottom of the results page.
• Google Search Console, once your site has any traffic at all. The Performance report shows the queries you already appear for, including ones you never targeted. This is the single most valuable free source, because it is your own data rather than an estimate.
• Forums and communities in your subject. Reddit threads, specialist forums and question sites show you the language real people use, which is frequently not the language the industry uses.

One page, one job

Do not write five pages targeting near-identical phrases. Search engines will pick one and largely ignore the rest, and you will have split your own effort. This is usually called keyword cannibalisation, and the fix is consolidation: merge the near-duplicates into one thorough page and redirect the others to it.

Be honest about difficulty

Keyword tools give a difficulty score, and those scores are estimates built from link metrics. They are useful as a rough filter and misleading as a precise figure.

The manual check is better and takes a minute: search the phrase and look at who ranks. Are they sites like yours, or are they national publishers and retailers? Do the top results genuinely answer the question, or is there an obvious gap you could fill better? That gap, when you find one, is worth more than any difficulty score.