Keyword research is the process of finding the searches your audience uses, understanding the task behind each query, and mapping distinct needs to the right existing or planned page. A useful workflow combines audience knowledge, search-result review and available data; it does not rely on a universal search-volume or difficulty threshold.
This beginner workflow can be completed with free Google tools and a spreadsheet. Paid tools may add estimates, but their numbers remain models rather than guarantees.
Prerequisites
- A defined audience, offer or site purpose
- An inventory of published, draft and planned URLs
- A spreadsheet with columns for query, intent, page, evidence and priority
- Access to Google Search Console if the site already receives impressions
Step 1: Define the Audience and Seed Topics
Start with what the site actually helps people do. List products, services, recurring support questions, beginner problems and follow-up tasks. A seed topic is a starting point, not yet a target keyword.
| Source | Question to ask | Example seed |
|---|---|---|
| Product or service | What problem does it solve? | technical SEO audit |
| Support conversations | What repeatedly confuses users? | why a page is not indexed |
| Existing content | Which useful next step is missing? | keyword-to-URL mapping |
| Business goal | Which task can lead to a meaningful action? | choose an SEO tool |
Step 2: Collect Query Ideas From Several Sources
Expand each seed with sources that reveal different kinds of evidence:
- Search Console: queries and pages already earning impressions.
- Google Trends: relative interest over time, locations and related searches. Google’s Trends guide explains how it can inform content strategy.
- Autocomplete and related-search features: useful wording ideas, not guaranteed volume data.
- Google Ads Keyword Planner: ideas and advertising forecasts. Google describes it as a tool for building search-ad campaigns, so its competition metric is advertiser competition—not organic SEO difficulty.
- Customer and community language: the words real users employ when describing the problem.
Record the source and date beside each idea. Do not claim that a suggestion proves monthly demand, and do not invent country-specific volume when you have not retrieved it.
Step 3: Remove Irrelevant and Duplicate Ideas
Discard queries outside the site’s purpose, queries the business cannot answer responsibly and synonyms that clearly lead to the same task. Keep close variants together until you review intent; wording differences alone do not justify separate pages.
Step 4: Determine Search Intent
Search intent is the job a searcher wants to complete. Classify each query by the dominant need—learn, compare, navigate, buy or complete a task—then review the current results to confirm what Google is serving. The search intent guide explains the categories in depth.
Look beyond labels. “Keyword research tools” may imply comparison, while “how to use Keyword Planner” implies instructions. If the same query serves mixed needs, document the ambiguity rather than forcing certainty.
Step 5: Evaluate the Evidence, Not a Magic Threshold
| Factor | What it tells you | Limitation |
|---|---|---|
| Relevance | How closely the query matches the audience and site purpose | Requires human judgment |
| Search demand | Whether evidence suggests people use the query | Tool estimates vary by source, country and date |
| Organic difficulty | How challenging current results may be to compete with | Third-party KD scores use proprietary formulas |
| CPC | Advertiser bid estimates and possible commercial interest | It is not an SEO ranking factor or proof of conversion value |
| Business value | Whether satisfying the query supports a meaningful outcome | High value does not make content rank |
| Evidence gap | What current pages fail to explain or demonstrate | A gap must still matter to users |
Do not apply universal rules such as “KD below 40” or “500–5,000 searches.” Appropriate targets depend on the market, location, site, intent and available evidence. Read the separate guide to keyword difficulty before relying on any one score.
Step 6: Group Queries by Shared Intent
Cluster phrases only when one page can satisfy them completely. Separate them when the reader needs a different outcome. This prevents keyword-fanout pages that compete with each other.
Illustrative example: “keyword mapping template,” “how to map keywords to URLs” and “keyword-to-page map” can often share one implementation guide. “What is keyword difficulty?” needs a different page because the reader’s task is evaluation, not mapping.
Assign each cluster to one canonical destination in a keyword map. If a suitable page already exists, improve it instead of creating a duplicate. Use the topic-cluster workflow to connect genuinely distinct supporting tasks.
Step 7: Review the Current Search Results
Search the query in the target location and note the dominant page type, freshness needs, features and level of specificity. Review results to understand the task—not to copy headings, word counts or formats.
- Are results tutorials, tools, product pages, category pages or comparisons?
- Which entities, conditions and follow-up questions recur?
- Where do current results leave a useful evidence or explanation gap?
- Would your proposed page add a distinct outcome or merely rephrase what exists?
Step 8: Prioritize With a Transparent Score
You can use a simple editorial score as a decision aid. The following is illustrative, not a Google metric:
| Criterion | Score |
|---|---|
| Audience relevance | 0–3 |
| Clear, distinct intent | 0–3 |
| Evidence of demand | 0–3 |
| Business or learning value | 0–3 |
| Ability to add credible original value | 0–3 |
| Duplication risk | Subtract 0–3 |
Document why each score was assigned. A low-volume task can still be worthwhile when it serves an important customer need; a high-volume phrase can be a poor choice when the site cannot satisfy its intent.
Step 9: Create a One-Page Content Brief
Record the primary task, audience, canonical URL, supporting query language, required entities, primary sources, examples, internal links and success measure. “Focus keyword” is an editorial label, not a phrase that must appear in every field.
Use the main phrase naturally in the title or introduction when it accurately describes the page. Do not force it into image alt text, headings or sentences where another description is clearer.
Step 10: Write for the Task and Verify the Claims
Answer the main question early, then provide prerequisites, steps, evidence, examples, troubleshooting and limitations as the task requires. Google’s people-first content guidance explicitly rejects writing to an assumed preferred word count.
Use primary sources for factual claims and label examples as illustrative when they are not real case studies. Review every commercial or performance claim before publication.
Step 11: Publish, Link and Measure
Add relevant internal links from discoverable pages, confirm indexability and monitor impressions, queries, clicks and conversions over a meaningful period. The Google Search Console guide explains the available data.
Use new query data to refine explanations or cover a missing subtask. Do not rewrite a page merely because it has not ranked immediately, and do not attribute short-term changes to one edit without sufficient evidence.
Common Keyword Research Problems
| Problem | Repair |
|---|---|
| Keyword Planner shows “low competition” | Treat it as advertiser competition, then assess organic results separately. |
| Several phrases look similar | Compare the reader task and current results before deciding whether they share a page. |
| A tool reports different volume than another | Record source, country and date; use ranges and direction rather than false precision. |
| The site already has a matching page | Improve or remap the existing URL instead of creating a competing article. |
| A query has no reliable volume estimate | Use Search Console, customer evidence and relevance; state that volume is unverified. |
| A page gets impressions but no clicks | Review intent, position, title and snippet with enough data before editing. |
Limitations
Keyword data is incomplete and tool-dependent. Search demand changes by country, language, season and device. Difficulty scores are estimates, advertiser competition is not organic difficulty, and no keyword choice guarantees ranking or traffic.
Conclusion
Do keyword research by starting with the audience, collecting evidence from several sources, identifying intent, grouping only true variants and mapping each task to one useful URL. Prioritize relevance and the ability to add value over any single metric.
Frequently Asked Questions
How often should keyword research be updated?
Review it when products, audience needs, search behavior or Search Console data change. There is no required calendar interval.
Can I do keyword research for free?
Yes. Search Console, Google Trends, search-result review, customer questions and a spreadsheet can support a useful workflow.
Is Keyword Planner competition the same as SEO difficulty?
No. Keyword Planner is built for advertising; its competition field describes advertiser competition, not organic ranking difficulty.
How many keywords should one article target?
Use one dominant intent and include the natural query variations needed to satisfy it. There is no ideal fixed number.
How long does keyword research take to produce SEO results?
There is no guaranteed timeline. Results depend on indexing, competition, content quality, site signals and demand.