Key takeaways
- Autocomplete gives you phrasing, Keyword Planner gives a rough demand check, and Search Console shows queries you already appear for.
- Google says the keywords meta tag is unused and keyword stuffing breaks its spam policies [1].
- Keyword Planner needs billing details during account setup before basic features unlock, and very low-volume terms never show [3].
- Score every candidate 1–3 on six criteria; anything under 11 goes in the backlog, not this month's calendar.
- No method promises a position — Google says best practice improves the odds, not the outcome [1].
Free keyword research methods: the short answer
Free keyword research methods means building a topic list from sources you already have: Google's autocomplete suggestions, the queries in your Search Console Performance report, and monthly search estimates from Google Keyword Planner. You then score each idea for relevance and search intent before you write. The core of the process costs nothing to run.
Google frames SEO as helping search engines understand your content, and helping people find your site and decide whether to visit it [1]. Keyword research is the front half of that job: choosing which reader questions you want a page to answer.
Five free sources are worth combining:
- Google autocomplete and "people also search for" — the phrasings people actually type.
- Google Keyword Planner — monthly search estimates plus filters to narrow a list [3].
- Search Console Performance report — the queries your site already appears for [2].
- Google Trends — relative interest over time, useful for spotting a seasonal shape.
- The results page itself — the format Google is currently rewarding for that query.
None of them hands you a magic number. What they give you is enough evidence to pick your next five articles instead of guessing. If you want to see where the paid tools fit before spending anything, our comparison of keyword research tools covers that gap honestly.
Disclosure: Some links in this article are affiliate links. If you click and buy, we may earn a commission at no extra cost to you. Here's how this site makes money: commissions from recommended tools support our free guides.
What Google says about keywords — and what it says doesn't matter
Most free keyword advice fails for the same reason: it treats research as a numbers hunt. Google's own guidance is more modest. The SEO Starter Guide says following best practices makes it more likely a site can be crawled, indexed and understood, but there is no promise that any particular site will be added to Google's index [1].
Sit with that before you spend a night in a spreadsheet. The same guide lists things Google says do not move rankings:
- The keywords meta tag — Google Search doesn't use it [1].
- Repeating your phrase — keyword stuffing is against Google's spam policies [1].
- Hitting a word count — content length alone doesn't matter for ranking purposes [1].
- Stuffing the phrase into your domain — keywords in the name of the domain or URL path have hardly any effect beyond appearing in breadcrumbs [1].
What it does ask for is content that is easy to read and well organised, unique rather than copied or rehashed, kept up to date, and helpful, reliable and people-first [1]. It also tells you to expect your readers' search terms, noting that an expert and a newcomer might search for the same thing with completely different words — its example is "charcuterie" versus "cheese board" [1]. Google's language matching can connect a page to many queries even when the exact words aren't on it [1].
Timing matters too. Google says some changes take effect in a few hours and others could take several months, and that you generally want to wait a few weeks before judging whether a change helped [1].
Step 1: Build a seed list with Google autocomplete
Autocomplete is the cheapest research in existence: no account, no cost, and it reflects phrasings people genuinely type. Treat it as a list generator, not a demand measurement — it tells you a phrase exists, not how big it is.
Write three seeds before you type anything
Name the topic you want to cover, the problem your reader has in their own words, and the product category you might eventually recommend. Every query you collect should hang off one of those three, or you'll end up with a list of interesting phrases that have nothing to do with your site.
Run the alphabet and question drills
Type your seed into Google and note the suggestions. Then repeat with each letter of the alphabet after the seed, and with question openers: how, what, why, is, can, should, best, versus, for. Record everything that sounds like a real person's sentence. It is mechanical work, and that is the point — twenty minutes of typing beats an hour of staring at a blank content calendar.
Follow the "people also search for" trail
Search your main phrase and work through the related searches at the foot of the page and the "people also ask" box. Each one is a question Google is confident is connected to your topic. Copy the ones that match your reader, then repeat the drill on the two most promising.
Why wording beats volume at this stage
Google's "charcuterie" versus "cheese board" example is the whole lesson [1]: the same reader may search with jargon or with plain English. If you only write in your own vocabulary, you write for people who already know the topic. Autocomplete hands you the plain-language version — and the plain version is usually the affiliate-friendly one.
Step 2: Sanity-check demand in Google Keyword Planner
Keyword Planner is Google's own tool for researching keywords for Search campaigns: it suggests related keywords and shows estimates of the monthly searches they receive and the cost to target them, with filters to narrow the list [3]. It is built for advertisers, not bloggers, so give it one job — a rough check that a phrase has some monthly demand behind it.
What to expect before you sign in
Access is not instant. Google's help page states that you must complete your account setup by entering your billing information to access basic features such as "Get ideas for new keywords" [3]. Because setup steps and interface wording change, read the current help page and check the date on any tutorial you follow rather than trusting a year-old screenshot.
Two limits to keep in mind
First, keywords with very low search volumes are not discoverable or forecastable in the tool [3] — so a blank is not proof that nobody searches, only that the estimate is too small to display. Second, forecasts are refreshed daily and based on the last 7–10 days, adjusted for seasonality [3], so any single snapshot is exactly that.
Ignore the cost column for now. It measures advertising competition, not how hard a topic is to write about. The number you actually want is the monthly estimate, and even that is an approximation.
Step 3: Mine the queries you already appear for in Search Console
Once your site has been live for a while, the most honest keyword data you own sits in Google Search Console. The Performance report shows clicks, impressions, click-through rate and average position, and lets you group the data by query, page, country, device, search appearance or date [2]. The default view covers the past three months [2].
If Search Console is new to you, we walk through the reports worth a weekly look in Google Search Console for affiliate sites. For keyword work you need one tab: the Queries table.
Three patterns worth hunting
- High impressions, low click-through rate. People see the page and don't click. That is usually a title or description problem rather than a topic problem, and the report can show which pages have the lowest click-through rate from search results [2].
- Queries sitting below the first page. Average position tells you how close you already are [2]. Improving a query where you already appear is far cheaper than starting from nowhere.
- Questions you didn't plan for. The queries list shows what people typed to reach you [2]. Those are ready-made H2s and follow-up articles.
Two cautions about what the table shows
The list is built from searches where your site appeared, so it is not a picture of total demand [2]. And if you type a query from the list into Google and don't see your page, the data isn't wrong — Google notes that results are specific to the time, place, device and recent history of the person searching [2]. The newest figures can also be preliminary and may still change [2]. Export the table before you start editing anything, so you have a baseline to compare against later.

Step 4: Read the shape of demand with Trends and the live results page
Two checks remain, and both are about context rather than volume.
Google Trends for season and direction
Trends shows relative interest, not search counts — which is exactly what makes it a useful second opinion. If a topic spikes every November, a post published in December is late. Set the date range and country, note whether interest looks flat, rising or spiky, and write the season into your calendar. Trends charts get revised and re-drawn over time, so date your notes rather than relying on an old screenshot.
The results page tells you what format to write
Search your target phrase and look at what ranks: step-by-step posts, comparison lists, product pages, forum threads, videos? That format is what Google is currently rewarding for that query, and matching it is more reliable than fighting it. We go deeper on this in how search intent should shape an affiliate page.
If the top results are forum threads and personal accounts, that is a signal too — readers want lived experience, which is what Google's people-first guidance points at [1]. Our piece on what Google's helpful-content guidance means for small affiliate sites covers how to show that honestly when you haven't used the product yourself.
A free-keyword-research scorecard you can copy
A raw query list won't tell you what to write first. Score each candidate from 1 to 3 on six criteria, then add them up. This is our own heuristic rather than a Google metric, but it forces the judgement that a volume column hides.
| Criterion | Ask yourself | Score |
|---|---|---|
| Reader fit | Would my one reader recognise this as their problem? | 1–3 |
| Demand evidence | Did it appear in autocomplete, or show an estimate in Keyword Planner [3]? | 1–3 |
| Intent clarity | Can I tell whether the searcher wants an answer, a comparison or a product? | 1–3 |
| Competition realism | Do the current results look beatable by a small site? | 1–3 |
| Your angle | Can I add something original, or am I rehashing what already ranks [1]? | 1–3 |
| Staying power | Will this still be worth updating in a year [1]? | 1–3 |
Illustrative example. Three candidates from one autocomplete session score like this: topic A 3+3+2+1+3+2 = 14; topic B 2+2+3+3+1+2 = 13; topic C 1+1+2+2+1+1 = 8. Topic A goes into this month's calendar, B into the backlog for later, and C is dropped. The arithmetic matters less than the argument you have with yourself while scoring.
Two rules keep it honest. Anything below 11 doesn't get written — a low score is information, not a dare. And no topic earns a 3 for "your angle" unless you can name the specific thing you'll add: a worked example, a checklist, a dated comparison, or a plain-English explanation of a confusing term.
The weekly routine, as a fill-in template
Twenty minutes once a week is enough to keep a queue of scored topics ahead of your writing. Copy this into a document and fill it in by hand.
WEEK OF: ____ SEEDS (three, written before typing): 1. ____ 2. ____ 3. ____ AUTOCOMPLETE PHRASES COLLECTED (aim for ten or more): - ____ - ____ KEYWORD PLANNER CHECK (estimate noted per phrase; blank means too small to show [3]): - ____ SEARCH CONSOLE QUERIES TO REVISIT (copied from the exported Queries table [2]): - ____ SCORECARD (18 max, see the table above): Topic: ____ | Score: __ | Decision: write / backlog / drop NEXT ACTION: Working title for the top topic: ____ Page it will link to: ____ Date to review this page again: ____
Filling this in weekly leaves a paper trail. When a post underperforms later, you can see whether the topic scored badly or whether the writing did.
What these methods can't tell you — and when to spend money
Being clear about the limits is what separates a process from a pitch.
- Estimates are estimates. Keyword Planner's figures are built for planning ad campaigns, and very low-volume phrases never appear at all [3].
- Search Console looks backwards. It shows what you already appear for, not what you could rank for [2].
- Nobody can promise a position. Google is explicit that following best practices improves the odds of being crawled, indexed and understood — not of ranking first [1].
- Free data rarely includes difficulty. You judge competition by reading the results page, which is slower but not wrong.
So when is paying worthwhile? When you have published pages, you know which ones are close to breaking through, and the bottleneck is comparison data rather than writing time. Before that point, a subscription mostly buys a nicer interface for the same decisions.
Illustrative example. Suppose a subscription tool were priced at $50 a month. Over 12 months that is 50 × 12 = $600. That isn't automatically a waste — but if your site has nine posts and no impressions yet, the $600 buys insight you can't act on. Spend after you have evidence, not before; our keyword research for beginners lesson shows how to find a topic before buying a tool.
Two bits of housekeeping before you publish. Check the disclosure rules that apply to you, and our FTC affiliate disclosure examples show how that can look in a blog post, an email and a video. And remember that Google primarily finds pages through links from other pages it has already crawled [1], so plan your internal linking for a 20-page affiliate site before you publish, and use these checks for refreshing old posts when Search Console shows a page slipping [2].
For how all of this fits into the wider build, read our guide to SEO for affiliate marketing. Affiliate program terms and tax rules change often — always check the current official page and the date on it, and ask a qualified professional about your own situation.
Related guides: Search Intent for Affiliate Content; 12 Clauses to Read in an Affiliate Program's Terms Before You Join; How to Judge a Product Before You Promote It; Affiliate Commission Types Explained.
Start with one seed and the twenty-minute drill
You don't need a subscription to choose your next five articles — you need a seed word, a notebook and a weekly slot. When you're ready to see which paid tools would add something the free ones can't, our dated comparison sets out what each one does and what to skip.
Frequently asked questions
Is Google keyword research free?
The sources in this guide cost nothing: autocomplete, Google Trends and the results page itself. Keyword Planner is different — Google's help page says you must complete account setup by entering your billing information before basic features like "Get ideas for new keywords" unlock [3]. Check the current help page for what that involves before you sign up, and date any tutorial you follow.
How do I research keywords for SEO with no budget?
Collect phrasings from autocomplete and related searches, check monthly estimates in Keyword Planner [3], then look at the Search Console Queries table for phrases where your site already appears but ranks low [2]. Finish by reading the live results page to see what format ranks. Score the survivors, write the top one, and repeat weekly.
Is Ubersuggest free to use?
Free tiers are set by the company running the tool and can change without notice, so check that tool's own pricing page on the day you sign up rather than a screenshot in an old blog post. Nothing in the process above depends on one third-party tool: autocomplete, Keyword Planner [3] and Search Console [2] are enough to choose your next few articles.
What kinds of keywords should a beginner track?
Three groups are enough to start: question phrases (how, what, why), comparison phrases (one option versus another), and product-led phrases. Google's own guidance notes that people with different levels of knowledge search with different words for the same thing [1], so collect the plain-English variant of each phrase as well as the jargon version.
How long until keyword research shows results?
Research is the choosing, not the ranking, so expect a gap. Google says some changes take effect in a few hours while others could take several months, and that you generally want to wait a few weeks before deciding whether a change helped [1]. Pick a review date when you publish, then use the Search Console data to decide what to rewrite or expand.
Sources
- SEO Starter Guide: The Basics | Google Search Central | Documentation | Google for Developers (source page, retrieved 2026-10-20)
- Performance report (Search results): Overview and basic setup - Search Console Help (source page, retrieved 2026-10-20)
- Use Keyword Planner - Google Ads Help (source page, retrieved 2026-10-20)
This guide is general education. Program terms and tools change; check the linked official pages for the current version. How we write: editorial policy. This page has no affiliate links.



