How to find keywords without paid tools: a free, repeatable research workflow
Most "free keyword research" advice is a list of links. This is a workflow: where to look, what to copy out, and how to turn what you find into one search-intent decision per post. If you want to know how to find keywords without paid tools, you can build a steady pipeline using Google Search Console, public sitemaps, Google Trends and a handful of free endpoints — none of it requires a subscription, only the discipline to write what the queries actually want.
Table of contents
- Key takeaways
- What "free" keyword research can and cannot do
- Where do the best free keyword ideas actually come from?
- Step 1: pull the queries you already earn clicks for
- Step 2: mine competitor sitemaps for topics you do not cover
- Step 3: use Google Trends to check demand and seasonality
- Step 4: group queries by search intent before you plan posts
- Free keyword research template you can copy today
- Frequently asked questions
- Sources
Key takeaways
- Search Console is the only source that tells you which queries your site already earns impressions and clicks for — start there before any third-party tool.
- A competitor's public sitemap is a free, dated list of topics they have decided are worth a page; mining it is one of the fastest free keyword research methods available.
- Grouping by search intent before you plan keeps one post from competing with another on your own site.
- Google Trends is for checking demand and seasonality, not for finding new queries — use it to time what you already have.
- A repeatable research workflow beats a clever one: same sources, same column order, same date every cycle.
What "free" keyword research can and cannot do
It will surface the queries you already earn, the topics your competitors have shipped, and the seasonal shape of demand for any term. It will not give you clickstream data, bulk difficulty scores, or a backlink view of the SERP, because those datasets are not freely available.
That trade-off is workable. What most teams need from keyword research is not a difficulty number; it is a list of queries a real person typed, grouped by what they were trying to do, paired with a page that already earns those clicks. Free sources answer that question; paid tools mainly answer the secondary one about competitive difficulty. A workflow built around Google Search Console keyword ideas, public sitemaps and Trends gives you the primary signal, which is the one that decides whether a post compounds.
Where do the best free keyword ideas actually come from?
Three places, in this order. First, your own Search Console: queries you already rank for, often on pages that were never written with that query in mind. Second, the public sitemaps and HTML sitemaps of competitors in your space, which behave like a free, dated table of contents of what they have decided is worth a page. Third, Google's "People also search for" and "People also ask" boxes, which surface the queries real searchers typed next.
Anything else — autosuggest scraping, "answer the public"–style tools, forum mining — is a variation on these three with extra steps. If your workflow is repeatable, stick to the sources that publish their method openly and let you verify the data with one click.
Step 1: pull the queries you already earn clicks for
Open Google Search Console, filter to the last 16 months (the maximum it retains), set the dimension to "Query", and export. Sort by impressions, then by clicks. You now have two lists: queries Google has shown your page for (impressions), and queries someone actually clicked through for (clicks).
Look for three patterns. First, queries with many impressions and very few clicks — usually a title and meta description that do not match the query, fixable in 20 minutes. Second, queries sitting on page two with a position between 11 and 20 — these are the cheapest wins, because the page already exists and the topic already has some relevance signal. Third, queries that appear on more than one of your pages at once — those are your internal cannibalisation problems, and the fix is consolidation, not new content.
This step alone usually produces a backlog bigger than the next two months can ship. Resist the temptation to start writing; the next step is what tells you which of these queries deserve a brand-new page and which deserve a rewrite.
Step 2: mine competitor sitemaps for topics you do not cover
Pick three or four competitors whose blog you would not be embarrassed to be associated with. Find their sitemap — usually at /sitemap.xml, sometimes at /sitemap_index.xml, occasionally split per content type. Open it raw, copy the URLs that point at blog posts or articles, and paste them into a spreadsheet.
Strip the URLs down to the slug, and you have a list of topics a competitor has shipped. Read the slugs the way a reader would: how-to-migrate-a-postgres-database-without-downtime is a topic, pricing-faq is a page that probably does not earn search traffic. Keep the topics, drop the chrome (about, pricing, careers, tag pages).
This is the heart of competitor sitemap keyword mining: you are not copying their titles, you are reading the map of decisions they have already made. If a serious competitor has shipped a page on a topic and you have not, that is a signal — either they have demand you are missing, or they are wrong together, and a quick check on Trends tells you which. If your workflow covers five competitors a quarter, the backlog of "we should probably write this" never empties, but it also never goes stale.
Step 3: use Google Trends to check demand and seasonality
Google Trends is not a keyword finder. It is a demand meter. The mistake most free workflows make is using Trends to discover queries; the search suggestions inside Trends are thin and biased toward news. Use it for what it is good at: comparing the relative demand of two or three candidate topics, and reading the seasonality curve of the one you picked.
Two practical checks. First, for any topic that came out of Steps 1 or 2, type it into Trends and look at the 12-month curve. A flat line means the topic is evergreen and the publish date does not matter; a sharp seasonal peak means you either publish four to six weeks before the peak or you save the post for next year. Second, compare two candidate topics against each other in the same view — Trends lets you stack up to five terms. The relative heights tell you which one has more demand in your actual market, in the geography you write for, in the language you ship in.
This step is also where you stop confusing volume with opportunity. Trends has no absolute volume, only a normalised index, and that is the right unit for the decision you are making: not "how big is this market", but "is this the topic I should ship first".
Step 4: group queries by search intent before you plan posts
Before anything becomes a row in a content plan, decide what the searcher wants to do when they type the query. Four buckets cover almost every blog query:
- Informational — they want to learn something, the right format is a guide or explainer with a clear answer in the first 150 words.
- Commercial — they are comparing options, the right format is a comparison, a list of criteria, or a "best X for Y" page.
- Transactional — they want to do or buy something, the right format is a product, a signup, or a tutorial that ends in the next action.
- Navigational — they want a specific page, the right format is usually a redirect or a single canonical answer, not a blog post.
One query, one intent, one post. The most common mistake in keyword research done without paid tools is letting two intents share a page, which usually means one of them never ranks. A solid content plan vs content brief process catches the conflict at the planning stage, when it is still cheap to fix. If two of your candidate queries share a page, ask which intent is stronger; the weaker one either gets its own page or it gets cut.
A useful rule of thumb: if the SERP for the query shows mostly listicles, write a listicle; if it shows mostly how-to guides with step numbers, write a how-to. The SERP is Google telling you what shape it expects.
Free keyword research template you can copy today
A workable template has six columns and one date. The columns are: target query (the way it is typed, not the way you would write it in a sentence), search intent (one of the four above), source (Search Console, competitor sitemap, Trends, "People also ask"), candidate page (the URL on your site that should rank, or "new"), format (guide, comparison, tutorial, listicle), and publish date.
Fill it in this order, one query at a time. Decide the query and its intent first — without those, the rest of the row is guessing. Pick the source, because the source tells you how trustworthy the query is: a Search Console query is something Google has already shown your site for, a sitemap query is something a competitor decided was worth a page, a Trends query is something people searched for but you have no ranking signal on yet. Match the candidate page to the query: existing URL means rewrite, "new" means draft. Match the format to the SERP you would actually see if you typed the query in. Finally, set a publish date you can defend — not "soon", but the calendar slot you would protect if a customer asked.
The date is the column that turns a list into a plan. A backlog with no dates is a wishlist; a backlog with one date per row per week is something the team can actually ship against. Run the same four steps every month on the same day, in the same order, and you have a repeatable research loop instead of a heroic one.
Doing this loop by hand every month is honest work and it works. The same loop, run by a pipeline that scans a connected domain, builds a dated plan from the queries the site actually earns, and publishes each post on your own blog once it clears a set of checks, is what a content team with no writer or content manager buys instead — that is the trade the page What an SEO content plan should contain describes in full. The free workflow above is the manual version of the same idea; the automated version is what you reach for when the manual one stops fitting inside a working week.
Frequently asked questions
What is the fastest way to find keywords without paid tools?
Google Search Console is the fastest starting point because it shows you queries your own site already earns impressions and clicks for, which means the demand and the relevance signal are already there. After that, public competitor sitemaps give you a dated list of topics other sites have shipped, which is the second-fastest source of real queries. Together those two cover most of what a small or mid-sized blog needs to plan a quarter of content.
Can Google Trends replace a keyword research tool?
No. Trends is a demand meter, not a keyword finder; it tells you how interest in a query changes over time and how two or three queries compare, but it does not return a list of new queries and it does not show click or impression data. Use it after you have candidates from Search Console or sitemaps, to time and prioritise them, not to discover them in the first place.
How do I find competitor keywords without Ahrefs or a paid tool?
Read their sitemap — usually at /sitemap.xml or /sitemap_index.xml — and copy the URLs of their blog posts into a spreadsheet. The slugs are a free, dated list of topics they have decided deserve a page. You are not stealing their keywords, you are reading the map of decisions they have already made, and then deciding which gaps on your own site are worth closing.
How often should I run a free keyword research workflow?
Once a month is the cadence most teams can sustain. Run it on the same day every month, in the same order: Search Console first, competitor sitemaps second, Trends third, intent grouping fourth. A repeatable monthly loop beats a clever quarterly one, because the backlog stays warm and the publish dates stay realistic.
How do I know if a keyword is worth writing a post for?
Check three things. First, does the SERP for the query show a clear shape — guides, comparisons, tutorials, or product pages — that you can match? Second, does your site already earn any impressions for it in Search Console, or does a competitor who covers your space have a page on it? Third, does Trends show steady or seasonal demand over twelve months? If two of the three are yes, the query is worth a row in the plan.