Searching for keywords without paying for a license is possible, but it requires setting up a process, not opening a random website. What follows is the exact sequence I use when a new project doesn’t have the budget for Ahrefs or Semrush: first the universe of terms is defined, then the volume is validated with cross-referenced sources, and finally it is enriched with data that only appears in Search Console. If you follow the order, the result looks quite a lot like that of a paid tool; if you start at the end, you’ll end up with a list of a hundred keywords that you don’t know how to prioritize.
Step 0: prepare the ground before touching any tool
Before installing anything, you need two things: a spreadsheet (Google Sheets works) with these fixed columns —keyword, intent, estimated volume, perceived difficulty, proposed URL, source— and a list of 5 to 10 seed queries that describe your business using the words a customer would use, not the ones your sales department uses. This detail seems trivial, and it’s where half of all keyword research exercises fail: if you seed with internal jargon, you’ll drag that jargon through the entire process.
A mistake I see often: importing a free tool’s CSV directly into the sheet without normalizing it. The tools return duplicates for matching variants (singular/plural, with and without a preposition), and you’ll end up with the same keyword three times with different volumes. Spend ten minutes on a “canonical keyword” column before continuing.
Step 1: extract the initial universe with Keyword Planner
Google Keyword Planner is the most reliable primary source because the data comes from Google itself. The catch: to see volume ranges you need an active Google Ads account, and if you don’t have campaigns running, Google shows very broad ranges (for example, “1K–10K”). That’s not a flaw in the tool; it’s how it protects the data when there’s no ad spend. According to Google Ads’ official documentation, the ranges narrow as activity accumulates in the account.
Specific setup
- Go to Keyword Planner and choose “Discover new keywords.”
- Enter your seed queries and set the exact location and language. A common mistake: leaving the location as “all countries” and then being surprised that irrelevant terms appear.
- In the network filter, leave only “Google search network.” The search partners option adds noise from third-party sites.
- Download the CSV and in the sheet add a “source = Planner” column.
Honest limitation: Keyword Planner is designed for advertising, so it groups terms aggressively and doesn’t distinguish very long informational intents well. For long tails of more than five words, its coverage drops noticeably.
Step 2: expand with real search engine suggestions
This is where free tools add more value than paid ones for small projects, because they capture the user’s literal language. Set it up like this:
Google Suggest and alphabetical variants
Type your seed into Google and watch the autocomplete. Then repeat the query adding each letter of the alphabet (“keyword a,” “keyword b”…). It’s tedious by hand; free tools like AnswerThePublic (with a daily limit) or Keyword Tool Dominator automate part of the process. My advice: use manual autocomplete for the 10 main seeds and save the tools for the secondary ones. The reason is that many of these free utilities limit results or add watermarks to the export.
“People also ask” and related searches
They’re two gold mines for informational intent. When you extract them, don’t copy the question as is: rephrase it as it would appear in an H2. Google shows these questions because it detects demand, but you need a content structure, not a transcript.
Step 3: validate volume with cross-referenced sources
No free tool gives exact volumes, and anyone who promises you otherwise is oversimplifying. What you can do is triangulate: if a term appears with volume in Keyword Planner, with a similar volume in Ubersuggest (free plan limited to a few searches per day), and also has suggestions in autocomplete, the signal is solid. If it only appears in one source, treat it as unknown volume.
Here’s a concrete example that illustrates the difference between data and reality: in a local services project, Keyword Planner showed 20 monthly searches for a very specific variant of the service. We almost discarded the term. When we checked Search Console three months later, that variant was generating steady impressions and some conversions, because the real volume included proximity searches that the Planner doesn’t reflect accurately at the level of a small city. The lesson: for local businesses, the Planner’s volume systematically underestimates. Always cross-check with the Search Console performance report if the domain already has history.
Step 4: turn Search Console into your keyword tool
If the site already receives traffic, Search Console is the best free keyword tool there is, and almost no one configures it properly for this use. Procedure:
- Go to Performance → Search results.
- Filter by query and export the last 3 months (the maximum the interface usually keeps is 16 months, but with limited samples on large properties).
- In the sheet, cross-reference with your Planner list and mark matches.
- Look for the most valuable pattern: queries with many impressions and low CTR. There’s content that already ranks in position 8-15 and only needs on-page optimization.
Important nuance: Search Console hides queries for privacy when volume is low. On sites with little traffic, the “(other)” row can group most of your terms. It’s not a configuration error; it’s a product limitation worth knowing before getting frustrated.
Step 5: prioritize without fooling yourself with difficulty
Free keyword difficulty tools (the KD-type metric) are estimates based on links, and their reliability plummets for local markets and for Spanish. My recommendation: don’t trust the number, do a manual check of the top 10 SERP positions. If you see forums, YouTube videos, or directory pages, the real difficulty is lower than what the tool says. If you see established brands with specific pages and updated content, mentally add two points.
A prioritization criterion that works: sort by (commercial intent × SERP ease) and leave volume as a tiebreaker, not as the main criterion. A term with 200 searches and clear transactional intent usually performs better than one with 5,000 informational searches that requires a 3,000-word article to compete.
What you won’t get with free tools
Let’s be transparent about the limits:
- Long position history: free versions usually give 7-30 days. For seasonal trends you need a year of data.
- Page-level competitor data: you can see their main keywords in some free tools, but not the full breakdown or the evolution.
- Mass exports: almost all of them limit rows or require registration. For a cluster of 500 terms you’ll have to break the work into chunks.
- Daily updates: free volumes are refreshed less frequently.
If the project needs any of these four things on a recurring basis, the cost of the license pays for itself. To validate an idea, a fledgling blog, or a local website, the free workflow described covers 80% of the real need.
Monthly maintenance routine
The workflow doesn’t end with the list. Once a month: export new queries from Search Console, check whether any keyword from the Planner changed volume range, and update the status column. Thirty minutes. That habit avoids the classic problem of having a sheet with keywords from a year ago that no longer reflect how people search. And if at some point you notice that a free source changed its limits —it happens often with free plans—, record the extraction date in the sheet. Without that data, you won’t know whether the volume in front of you is from this quarter or the previous one.