How to Do Keyword Research With Free Methods That Work

Table of contents9

Two years ago I paid for my first keyword research subscription. The price was over a hundred dollars a month, and I was so excited that I spent a whole weekend inside the dashboard. On Monday morning, I had a spreadsheet with 4,000 keywords — and no idea which one to write about first. Six months later I canceled the subscription. The free methods I’m about to teach you had done the real work all along. I’ve done keyword research for blogs in food, tech, and travel since then, and my rule is simple: classify keywords by intent before you look at a single number. That’s the method here, and it costs nothing.

I know the fear. Every guide you read says you need expensive tools to “do SEO properly.” You have zero budget, or you’re not ready to spend, and it feels like you’re already behind. Here’s my honest nudge back: expensive tools never did the thinking for me. The thinking is the work, and it’s free.

In this guide you’ll learn how to do keyword research without paying for anything — in this exact order:

• How to collect real words from real people, not tool guesses
• How to make Google hand you hundreds of free keyword ideas
• How to sort every keyword by intent before you touch any numbers
• How to read the competition with your own eyes and spot keywords you can actually win
• How to build a simple keyword map so every page has a job
• How to keep your list alive with data from your own site

Ready? Let’s get started.

Step 1: Steal Words From Real People, Not Tools

The first time I did keyword research properly, I didn’t open any tool. I opened the comments section of a competitor’s blog post and read 200 comments. Buried in those comments were phrases like “my sourdough starter smells like vinegar, is it dead?” — a phrase no keyword tool had suggested to me, and a phrase real people actually typed into Google.

Here’s the deal: tools guess at what people search. Real people tell you exactly what they search. Your job in step one is to collect raw language from three places: reviews (Amazon, Google, any review site in your niche — the “pros and cons” sections are gold), forums and communities (Reddit, Quora, niche Facebook groups — the questions people repeat are keyword ideas), and your own inbox (if you have a site already, every support email and comment is a keyword in disguise).

Write everything down in one big list. Don’t filter yet. A phrase like “cheap running shoes for flat feet that don’t hurt” is ugly, but it tells you three things: the problem (flat feet), the budget (cheap), and the pain (they hurt). Tool-generated keywords never come with that much context. This step takes about an hour, and it’s the foundation everything else builds on.

Step 2: Let Google Finish Their Sentences for You

Now take your raw list to the search box itself. Type the first few words of a phrase and stop. Google’s autocomplete suggestions are real searches made by real people, ranked by how often they’re typed. It’s the closest thing to reading searchers’ minds, and it’s completely free.

The best part? There’s a technique for squeezing hundreds of ideas out of it. Type your seed phrase, then add each letter of the alphabet after it — “how to bake sourdough a”, “how to bake sourdough b” — and watch new suggestions appear. Then do the same with numbers and question words (who, what, where, when, why, how). One seed phrase can turn into 200 suggestions in half an hour.

Then scroll down. Every results page has two free idea machines at the bottom: “People also ask” (click one question and three more appear — it’s nearly endless) and “Related searches” (Google’s own summary of what people search next). Open an incognito window so your personal search history doesn’t skew the suggestions. Copy every phrase that looks like something your readers would ask. Don’t judge them yet — judgment comes in step three.

Bonus Tip: Google Trends is another free window into the same mind-reading. Type a keyword and you see whether interest is rising, falling, or seasonal. A keyword that spikes every December is not a keyword you write about in December — it’s one you write about in September so Google has time to find it.

Step 3: Sort Every Keyword by Intent Before Looking at Numbers

This is the step most beginners skip, and it’s the one that changes everything. A keyword’s search intent is the reason behind the search — what the person actually wants when they type it. There are four types, and they’re easy to spot once you look:

• Informational — the person wants to learn something. “How do I fix a leaking tap?” They want an answer, not a product.
• Commercial investigation — the person is comparing before deciding. “Cordless drill vs impact driver” or “best laptop for students.” They’re researching.
• Transactional — the person is ready to act. “Buy organic coffee beans online.” They want a checkout button.
• Navigational — the person wants a specific site. “Facebook login.” You can’t win these, so ignore them.

Now go through your big list and label every keyword with one of these four. Then — and this is the important part — pick the intents that match what your site actually does. A food blog wins with informational keywords. A small shop wins with transactional ones. A review-style site wins with commercial investigation. I learned this the hard way: I once wrote a beautiful informational guide for a transactional keyword, and it ranked — for readers who never bought anything. Traffic that can’t convert is just a number.

One page should target one intent. When a keyword’s intent doesn’t match any page you want to build, delete the keyword. You’re not ignoring data; you’re making a decision. That’s what the numbers can’t do for you.

Step 4: Read the Competition With Your Own Eyes

Forget difficulty scores. You don’t need a number between 0 and 100 to tell you if you can win a keyword — you need to look at who’s actually winning it. Search the keyword in an incognito window and read the top ten results like a detective.

You’re looking for weakness, and weakness has a shape. Can you see forums, Reddit threads, or Quora answers ranking on page one? That’s Google telling you nobody has written a proper answer yet. Are the top results old — published years ago and clearly not updated? Stale content is beatable content. Are the pages thin, vague, or missing the exact answer the question asks for? I once found a keyword with thousands of monthly searches where the top result was a 300-word article from 2019 that didn’t answer the question at all. My detailed guide outranked it in six weeks.

You’re also looking for strength you should respect. If the top ten are all huge established sites with deep, updated guides, that keyword is a long-term project, not a quick win. Mark it for later. Beginners should start where small sites, forums, and outdated pages are winning — that’s your entry point. Pick five keywords where the current winners look beatable, and circle them. Those five are your first targets.

Step 5: Build a Simple Keyword Map (One Page, One Job)

Now turn your circled keywords into a plan. A keyword map is just a simple table — you can make it in any spreadsheet — with one row per page you plan to publish. The columns are: page topic, primary keyword, search intent, and 3–5 supporting phrases (the related questions from step two that the page will also answer).

The rule is strict: one primary keyword per page, and pages must not target each other’s keywords. When two of your pages chase the same keyword, they compete with each other instead of with the competition — search people call this keyword cannibalization, and it quietly kills both pages’ rankings. Your map prevents it before it starts.

Here’s how a row looks for a food blog:

| Page topic | Primary keyword | Intent | Supporting phrases | |—|—|—|—| | Reviving a dead sourdough starter | revive dead sourdough starter | Informational | sourdough starter smells like vinegar; how long does sourdough starter last; can you save moldy sourdough starter |

See how the supporting phrases are real questions from real people? When you write the page, those become your subheadings. The map doesn’t just organize your keywords — it writes your outline for you. I keep mine in a simple spreadsheet and review it once a month. It takes twenty minutes and it keeps every new article pointed at a keyword that was chosen on purpose.

Step 6: Keep Your List Alive With Your Own Data

Here’s the free trick that paid tools can’t give you: once your site has any traffic at all, Google Search Console tells you exactly which searches are bringing people to your pages — including searches you never targeted. Open the Performance report, look at the Queries tab, and you’ll find phrases getting impressions that aren’t in your map yet.

The best part? These are keywords you’re already partly ranking for. A page sitting at position 12 for a phrase you didn’t plan is an invitation: add a section that answers that phrase properly, and you can climb to page one with one afternoon of editing. I do this check on the first Monday of every month. It takes fifteen minutes and it has produced more traffic than most of my new articles.

Note: Search Console only shows data for your own site, and it takes a few days to start collecting after you verify your domain. Set it up the day you start — future you will be grateful. If you’d like a walkthrough of reading these reports properly, my guide on Google Search Console: reading the data that actually matters goes through it screen by screen.

What to Do With Keywords Once You Have Them

Research without writing is just a hobby. Your map should now feed a publishing rhythm: one page per row, written for the intent you labeled, with the supporting phrases as subheadings. When the page is drafted, run it past the fundamentals — clean title tags that don’t kill your rankings, and make sure it connects into your site’s structure with internal links placed with a real strategy, not at random.

And don’t let thin category or archive pages sit there doing nothing — if your keyword map includes category-level topics, my guide on fixing thin content on WordPress category pages shows how to turn them into ranking pages instead of dead weight.

FAQ

How many keywords should I target as a beginner? Start with five. Seriously — five pages, each targeting one winnable keyword, written better than anything currently ranking. One great article beats fifty rushed ones every time. Expand the list only when those five are published and indexed.

Is Google Keyword Planner free? The tool itself is free with a Google Ads account, but the search volume numbers are shown as wide ranges unless you spend money on ads. The ranges still tell you relative size — “10K–100K” vs “10–100” is useful information. Use it to compare keywords against each other, not as exact numbers.

How long does free keyword research take? The full process in this guide takes one focused afternoon the first time — an hour collecting real language, an hour on autocomplete, an hour sorting by intent and checking competition, and thirty minutes on the map. Monthly maintenance is about half an hour.

What if all the keywords in my niche look too competitive? Go longer and more specific. “Running shoes” is unwinnable; “running shoes for flat feet beginners overpronation” is a conversation you can join. Every competitive niche has long-tail corners where forums and thin pages still rank. Your step-four detective work exists to find them.

Do I need to redo keyword research regularly? Review your map monthly, but you don’t restart from zero. Add new phrases from Search Console, mark keywords you’ve won, and retire keywords whose intent no longer matches your site. The map is a living document, not a one-time project.

Final Thoughts

I canceled that expensive subscription two years ago, and I don’t miss it. Everything in this guide — the comments-section mining, the alphabet method, the intent sorting, the detective work on the results page, the simple map — has built real traffic for real sites without costing a cent.

You now have the exact process. Collect real language this week. Sort it by intent. Circle five winnable keywords. Build the map. Then write the first page — because research only matters when it turns into published pages. What keyword are you going after first? Tell me in the comments; I read every one.