Keyword research is the process of finding the words and phrases people type into search engines, then deciding which of them are worth building content around. It is part market research, part demand forecasting, and part editorial planning.
Most people treat it as a volume hunt. Find the biggest number, write the article, hope for the best. That approach produces spreadsheets full of terms a small site will never rank for and traffic that never converts. A better process asks three questions in order: what are people searching, what do they want when they search it, and can we realistically be the answer.
Step One: Build a Seed List Before Opening a Tool
Start with what you already know. Write down the services you sell, the problems customers describe when they call, the phrases they use in emails, and the questions your sales team answers over and over. Twenty or thirty terms from memory is plenty.
Then mine sources that reflect real language rather than tool output. Look at your Search Console queries, your site search logs, the objections in sales calls, and the questions competitors answer in their FAQs. Reddit threads and review sites in your category are unusually good for this, because people there describe problems in their own words rather than in industry vocabulary.
Step Two: Expand With Data
Now bring in tools. Google’s Keyword Planner is free with a Google Ads account and gives you search volume ranges, related terms, and bid estimates that hint at commercial value. Paid tools such as Ahrefs, Semrush, and their alternatives add difficulty scores, competitor keyword sets, and question-based variations.
Two expansion techniques produce most of the useful results. Feed the tool a competitor’s domain rather than a keyword, which shows what is already working in your category. And filter for question phrasing, since questions map cleanly to article topics and tend to surface in AI-generated answers as well as traditional results.
Step Three: Sort by Intent, Not Volume
Search intent is what someone actually wants when they type a phrase. Ranking for a term whose intent you cannot satisfy produces traffic that leaves immediately.
- Informational intent means the searcher wants to understand something. A query such as “what is technical SEO” is best served by a guide or explainer.
- Commercial intent means the searcher is comparing options before making a decision. A query such as “best CRM for small business” is typically best served by a comparison or review.
- Transactional intent means the searcher is ready to buy or hire. A query such as “SEO agency Los Angeles” is best served by a service or product page.
- Navigational intent means the searcher wants to reach a specific brand or website. A query such as “Imprint Insights” is best served by the brand’s own page.The fastest way to confirm intent is to search the term yourself and look at what already ranks. If page one is entirely blog posts, a product page will not break in no matter how well optimized it is. Google has already decided what kind of answer that query deserves.
Step Four: Judge Difficulty Honestly
Every tool reports a difficulty score, and every score is an estimate. Use them as a filter, not a verdict, and check the results page yourself for anything you plan to target seriously.
- Who is ranking. If the top ten are national publishers and Wikipedia, that is a different fight than one against three local competitors.
- How strong those pages are. A weak page ranking in position four is an opening. Ten comprehensive, recently updated pages is a wall.
- How old the results are. Stale content ranking well suggests nobody has bothered to compete properly.
- What the page format is. If Google is showing videos or listicles, that is the format you need to bring.
For a newer site, the honest filter is simple. Skip anything where the entire first page is dominated by sites with far more authority than yours, and concentrate on specific, lower-competition phrases where you can genuinely be the best result. That logic is the whole argument behind long tail keywords, and it is why small sites that target narrowly tend to grow faster than ones that chase head terms.
Step Five: Group Into Topics and Assign Pages
A keyword list is not a plan. The last step converts it into one.
- Group keywords that share intent and would be answered by the same page. "How much does seo cost," "seo pricing," and "seo cost per month" are one page, not three.
- Assign each group a single target URL. Two pages competing for the same term split their own signals and neither ranks well.
- Choose a primary keyword per page and let the rest inform subheadings and body copy.
- Sequence the work by opportunity rather than volume, starting with terms that are close to a purchase decision and within reach.
- Note the internal links each new page should receive from existing pages, because that is what makes it findable.
The Mistakes That Waste the Most Time
- Chasing volume you cannot rank for. A term with 40,000 monthly searches is worth nothing if you finish on page four permanently.
- Ignoring intent. Optimizing a service page for an informational query produces impressions and no business.
- Building one page per keyword variation. Search engines understand synonyms. Ten near-identical pages compete with each other and dilute your authority.
- Trusting difficulty scores without checking the results page. The score is a model. The results page is the actual competition.
- Treating zero-volume terms as worthless. Tools underreport long, specific phrases, and those are frequently the ones closest to a purchase.
- Forgetting the pages you already have. Improving something that ranks eleventh usually beats publishing something new.
Refreshing the Research
Keyword research is not a one-time project. Search behavior shifts, competitors publish, and AI-generated summaries change how often people click through at all. Revisit your Search Console data quarterly and look for queries where you rank between positions five and fifteen. Those pages already have relevance and usually need improvement rather than replacement, which is far cheaper than starting new content from nothing.
Once a page is targeted properly, the remaining work is execution, which is what the on-page SEO checklist covers, and sustained authority building of the kind an ongoing SEO and GEO program is built around.
Research That Ends in a Decision
The purpose of keyword research is not to produce a large spreadsheet. It is to decide what to publish next and why. If your process ends with a ranked list of pages to build, each with a clear reason for existing, it worked. If it ends with two thousand rows nobody opens again, it did not.
If you want help turning keyword research into a focused SEO strategy, contact us.
