Research questions before search volume
Free keyword research works best when it starts with an audience problem. A list of popular phrases is not a content strategy. You need to know what people are trying to accomplish, whether your site can genuinely help, and which page would satisfy that need.
Imagine a small shop selling refillable household products. “Cleaning” is broad and difficult to turn into a useful page. “How to clean a refillable spray bottle” describes a concrete task. Research should help you identify that task, its practical details, and the kind of answer a reader expects.
No method in this guide guarantees rankings or traffic. Google states that following its guidance does not guarantee indexing or a particular search position. 1
Start with language you already have
Collect questions from support conversations, sales calls, on-site search, and customer feedback that you are authorized to use. Remove names and sensitive details. Write down the wording people actually use rather than immediately replacing it with industry terminology.
Group questions by task: choosing a product, using it, fixing a problem, or comparing options. For the refill shop, those groups might include choosing bottle materials, mixing concentrates safely according to instructions, removing odors, and finding local refills.
Create a worksheet with columns for query, audience, task, evidence source, likely page type, existing URL, and next step. Keep commercial relevance separate from popularity. A question asked by a small number of qualified customers may be more useful than a broad phrase with little connection to your offering.
Expand seeds manually
Type a seed phrase into a search engine and inspect visible suggestions and related questions. Use them as possible wording, not as verified measures of demand. Results can vary by location, language, device, and previous activity.
Try modifiers that correspond to real decisions: “for beginners,” “without,” “versus,” “size,” “clean,” “repair,” and a relevant location. Do not create every combination mechanically. Keep only phrases that make sense for your audience and expertise.
Review a few public forums or community discussions manually when permitted. Look for recurring uncertainties and missing explanations, not material to copy. Do not collect personal information or automate scraping against a site's rules. Customer language should inform an original answer, not become a license to reproduce someone else's work.
Inspect intent in the results
Open the results for a candidate phrase and note what dominates: tutorials, product pages, comparison articles, videos, or local listings. This is evidence about likely intent, not an instruction to imitate every existing page.
For “refillable bottle sizes,” readers might expect a buying guide. For “refill shop near me,” a locally relevant business page may be more appropriate than a generic article. If the results show mixed intent, define your page's purpose especially clearly.
| Query pattern | Likely task | Useful page format |
|---|---|---|
| How to clean a bottle | Complete a procedure | Illustrated step-by-step guide |
| Glass versus metal bottle | Make a comparison | Decision table with trade-offs |
| Refill store in a town | Find a nearby provider | Accurate location page |
| Bottle leaking after refill | Diagnose a problem | Troubleshooting guide |
These are planning examples, not predictions about a particular live results page. Always inspect the query and market you intend to serve.
Use Search Console when you have a site
Search Console can show queries and pages associated with your site's search performance. Google documents it as a tool for understanding crawling, indexing, and search visibility. It is especially useful for discovering language connected to pages you already publish. 2
Filter to one relevant page and review queries over a meaningful period. Look for questions the page partly answers, confusing wording, or a mismatch between the title and actual content. Small datasets are noisy, and not every query is shown, so avoid drawing sweeping conclusions from a few impressions.
Do not create a new page for every wording variation. “Cleaning refill bottles” and “how to wash refill bottles” may belong on the same comprehensive page. Separate pages when the underlying task or audience differs enough to require a genuinely different answer.
Understand optional planning tools
Google Keyword Planner provides keyword ideas and search-related estimates for advertising. Its figures and competition labels should not be treated as a direct organic ranking difficulty score. Read the current account and setup requirements before using it; do not launch an advertising campaign merely to investigate a tool. 3
Trend tools can help compare relative interest or seasonality, but relative interest is not the same as exact search volume. Note the region, period, and search type when comparing terms. Avoid presenting approximate or indexed figures as precise demand forecasts.
If a tool asks for billing details, paid upgrades, or permissions beyond your needs, stop and evaluate whether the benefit justifies it. A free workflow can remain useful without access to every dataset.
Prioritize with transparent criteria
Score each idea on audience value, relevance to your site, available evidence, and production effort. Add whether you already have a page that should be improved instead. Treat these as editorial judgments, not mathematical predictions of traffic.
For example, a bottle-cleaning guide may score well because your support team receives the question repeatedly and you can provide accurate product-specific instructions. A broad history of household cleaning might attract interest but offer little reason for your site to publish it.
Choose a small set of topics you can cover well. For each, write a brief containing the reader's task, the answer promised, important limitations, official references where appropriate, and a useful next action. Plan internal links that genuinely help the reader continue.
Publish, measure, and revise
After publishing, confirm the page is accessible and eligible for indexing. Then monitor search performance over a reasonable period while recording major changes. Search results can change for many reasons, so do not attribute every movement to your latest edit.
Evaluate usefulness beyond clicks: fewer repeated support questions, clearer product selection, or better completion of a legitimate task. Improve weak sections using actual reader feedback rather than repeating a phrase more often.
Explore the free tools category for supporting utilities. Keep your research worksheet small enough to maintain and revisit it when customer needs change. The goal is not the largest keyword list. It is a defensible set of pages that answer real questions better than your site did before.