In this guide: Keyword research for beginners: understand search intent, use free keyword tools, judge keyword difficulty, and build a keyword list you can actually rank for.
Keyword research is the foundation of SEO. It tells you what your audience is searching for, which words they use, and which topics you have a realistic chance to rank for. Without it, you’re writing content and hoping someone finds it.
Keyword research is the process of finding and choosing the search terms (keywords) you want your pages to rank for.
Good keyword research answers:
Types of keywords:
For new websites, long-tail keywords are your best friends.
Search intent is the goal behind a search. Google ranks pages that satisfy the intent, so your content must match it.
| Intent | Example | Content type |
|---|---|---|
| Informational | “what is a cdn” | Guide, explanation |
| Commercial investigation | “best caching plugin” | Comparisons, reviews, lists |
| Transactional | “buy .com domain” | Product or pricing page |
| Navigational | “hostreta login” | Brand page |
How to check intent: search the keyword and study the top results. Are they tutorials? Lists? Product pages? Videos? Match the dominant format and angle — then do it better.
You can do solid research without paid tools:
SEO tools give a “difficulty” score, but you can judge manually:
A practical approach for new sites: target keywords where smaller sites already rank and where you can create clearly better content.
Create a simple spreadsheet with these columns:
Then:
Organize keywords into content pillars and clusters — see Content Marketing Strategy for Small Businesses.
How many keywords should one page target?
One main keyword plus closely related variations and subtopics.
Is search volume the most important factor?
No. Business relevance and intent matter alongside volume; neither a small nor a large estimate guarantees revenue.
Further reading: Google Search Central SEO Starter Guide
A useful research record starts with the country, language, period and source. Two numbers beside the same phrase may describe different populations or dates. Without that context, a spreadsheet can appear precise while comparing incompatible evidence. Keep the original settings with each observation rather than relying on memory.
For an English publication, review the United States and United Kingdom as distinct markets when the audience requires it. Canada, Ireland, Australia and New Zealand may share some questions while differing in terminology, currency or available services. Do not create six copies of one article merely because the country column has six values.
Keyword research should help identify whether a regional difference changes the answer. A guide to changing a WordPress heading may not need separate national versions. A comparison of locally available services may need genuine local evidence. Record the reason for regional content before commissioning another page.
Google’s Keyword Planner metrics guidance explains that average monthly searches reflect the selected period, location and network, including close variants. Its competition field concerns advertisers. Do not rename that column SEO difficulty or assume that high advertising competition establishes how difficult an organic result will be.
Google Trends documentation explains relative, normalized interest. A score of one hundred is not one hundred searches. Identical interest values from two regions need not represent identical search volumes. Keep Trends observations separate from a volume estimate and preserve the settings used for the comparison.
When using Ubersuggest or another SEO platform, record the displayed metric and date rather than translating its score into another tool’s scale. When using AnswerThePublic, record the questions and comparisons you actually see. If a field is hidden, unavailable or quota-limited, mark it unavailable instead of estimating a value and presenting it as supplied data.
Pinterest interest can inform a visual distribution plan and seasonal creative choices. It is not a measurement of Google search volume. Keyword research becomes more trustworthy when each source answers a defined question instead of being averaged into a single impressive-looking score.
Use one row per opportunity rather than one row per possible wording variation. Include the main phrase, supporting questions, intent, market, source, collection date, available volume or range, available difficulty, trend observation, existing coverage and proposed action. Keep a notes column for evidence limitations.
Separate observed data from editorial judgement. A tool may supply an estimate, while the decision that the topic fits your audience is your own assessment. Label each accordingly. A priority such as investigate or improve existing is more honest than calling an untested keyword easy.
Add an owner and a review date. Keyword research does not end when a spreadsheet is exported. A product comparison can become outdated, a seasonal question can lose relevance and a new page can overlap with an existing guide. The worksheet should make the next decision visible.
Search your public and scheduled archive before creating a new article. Read the actual coverage rather than relying only on titles. If an existing page already answers the intended task, improve it and record the supporting queries there. Several phrasings can belong to one coherent explanation.
Choose a primary destination for each intent and explain the relationship with supporting guides. A hosting comparison may link to a separate backup procedure when recovery becomes relevant. That does not require both pages to repeat the same full introduction or target exactly the same decision.
When an old URL has changed, update contextual references to the current public destination. Do not link readers to scheduled material before it becomes available. Our website growth workflow connects this inventory work to practical measurement and distribution.
Inspect the visible result types, the reader task and the questions left unresolved. A tutorial, product listing and discussion thread may serve different needs despite sharing words. Record the observed mix and its date. One search on one device is a useful observation, not a complete model of ranking competition.
Identify what your page can contribute honestly. That might be a clearer explanation, a reproducible procedure or a comparison based on current documentation. Do not copy another publisher’s structure and add length without adding practical value. Keyword research should inform the answer, not replace original editorial thought.
If the topic requires a test, define the method before claiming tested results. If no experiment exists, label the comparison as documentation-based. Never invent customer experience, prices, screenshots or performance figures to make the article seem more authoritative.
The brief should name the audience, direct answer, supporting questions, required evidence and acceptance checks. Put the most useful answer near the start. Organize later sections around steps or decisions rather than repeating the target phrase in every heading.
Use relevant internal references and a small set of primary sources for material facts. An article about a changing product limit needs current provider documentation. A broad opinion or anecdote should not be presented as a verified universal rule. State what is unknown and adjust the scope when evidence is missing.
A good brief also identifies what the article will not attempt to prove. A checklist does not establish a measured speed improvement, and a comparison does not certify a service. This boundary makes keyword research more actionable because the writer knows which claims require additional work.
Suppose a small business repeatedly asks whether hosting backups include restoring an individual order. This is an example question, not a claim about measured search demand. The editor first checks existing backup and hosting guides, then determines whether the practical task is already covered.
If coverage exists, the action may be to add a responsibility table and a link to the detailed recovery explanation. If the question exposes a distinct gap, a new brief can specify the restoration scenario and necessary sources. Do not invent a monthly volume or difficulty score for this example.
The keyword research record would preserve the original question, the hypothetical status of the example and the chosen action. Later, actual search and engagement observations may help revise the priority. They should not be retroactively inserted as if they were available during the original decision.
Promotional recommendation: KAWEDMA offers website development and hosting services that may help implement the destination your editorial plan requires. Ask for a written scope covering usable pages, editing access, contact forms and maintenance. This is not a claim that purchasing a service makes a keyword easier to rank for.
Affiliate disclosure: Hostreta may earn a commission if you purchase through this link. If the planned publication needs WordPress infrastructure, compare hosting.com’s managed WordPress offering against current requirements and alternatives. Verify resources, renewal terms and recovery responsibilities directly. Hosting does not replace keyword research or guarantee visibility.
After publication, record the public URL, actual publication date and distribution state. Use Search Console to inspect page and query relationships rather than joining unrelated exports by guesswork. Compare equivalent complete periods and preserve country filters when reviewing an international audience.
Website sessions, users and page views are not the same as search clicks. Keep these metrics separate and use a documented tracking convention for external campaigns. Our UTM parameters guide explains practical interpretation without assuming that every click creates a recorded session.
Keyword research priorities can change when new evidence arrives. A page may need a clearer answer, better internal links or a corrected factual claim. Avoid changing several unrelated elements simultaneously and then attributing every result to one keyword edit.
Keep the worksheet with your editorial calendar. The purpose of keyword research is to make better decisions about useful content, not to promise a position or a traffic total. Review the evidence, maintain the public archive and refine priorities as the site’s actual audience becomes clearer.
Give a colleague a sample opportunity row and ask them to reproduce its meaning. They should be able to identify the intended country, the source of each number and the proposed page. If a cell simply says high competition, ask whether that refers to advertisers, an SEO platform score or an editorial assessment. Rewrite ambiguous labels before using the row to allocate writing time.
Next, ask the colleague to explain the reader task in one sentence without repeating the query mechanically. For a backup question, they might describe a business owner deciding who can restore recent orders. That explanation is more useful than a list of synonyms because it identifies the evidence and limitations the article must address. Keep the original query alongside the explanation so the editorial interpretation remains visible.
Then review the existing destination together. Check whether the answer already appears in a subsection, whether its instructions are current and whether readers can find the next practical step. A necessary update may be smaller than a new article. Record the reason for the decision so a future editor does not recreate the same competing page when they encounter another wording variation.
Finally, choose an acceptance check that the editor can actually verify. A corrected product limit can be checked against current documentation. A new internal link can be opened and evaluated for usefulness. A claim that the edit will increase traffic cannot be verified at publication. Keep that as a hypothesis to review later using complete reporting periods and sufficient context.
Save the exercise outcome with the worksheet. Include unresolved questions, the responsible person and the next review date. This makes the research process repeatable without pretending that every judgement is an exact measurement. A transparent record of uncertainty is a stronger planning tool than a spreadsheet filled with invented precision or unsupported promises of easy rankings.
Before sharing a keyword research worksheet with a contractor, remove confidential customer details and unnecessary account information. The brief usually needs the audience question and evidence, not private support transcripts. Confirm access and ownership so the business can maintain the research record after the engagement ends.
We may earn a commission on purchases made through these links, at no extra cost to you.
Don’t miss the chance to participate in the biggest giveaway of this year.