How to Do Keyword Research for Free: A Beginner's Step-by-Step Guide

What Keyword Research Actually Is (and Why I Skipped It for Months)

I'll be honest with you: when I started my first site, I didn't do keyword research. I just wrote about whatever popped into my head and hoped Google would send me traffic. It didn't. Months of articles, barely any visitors. The mistake wasn't my writing - it was that nobody was searching for what I wrote about, at least not in the words I used. Keyword research is just the fix for that. It means finding the actual phrases real people type into Google, then writing the best answer to those phrases. That's it. No stuffing keywords into every sentence like it's 2010, no secret formulas. Just: what are people asking, and can I answer it better than what's already ranking? The method below costs nothing. I've ranked posts with it, and so can you.

Start With What You Already Know: Your Seed List

Before touching any tool, grab a blank document and dump out two lists from your own head. List one: topics you could write about without much research, the stuff you actually know. List two: questions people ask you about those topics - friends, family, commenters, customers, anyone. Let me walk you through this with an example. Say your site is about home espresso. Your seed list might be: grinder settings, milk frothing, sour shots, cheap beans, cleaning the machine. And questions people ask: why does my espresso taste sour, how fine should I grind, do I really need a tamper. That's your raw material. It's not keyword research yet, but everything downstream depends on how honest and specific this list is. Vague seeds like 'coffee' give you vague results. Specific seeds give you specific, winnable phrases.

Google's Search Box Is the Best Free Tool You'll Ever Use

Now type one of your seeds into Google, but don't hit enter. The suggestions that drop down are autocomplete, and they're gold - these are phrases people genuinely search, in huge numbers. Three tricks make this way more powerful. First, the alphabet trick: type your phrase, add a space, then each letter a through z, one at a time. 'espresso grind a', 'espresso grind b', and so on. Every letter gives you a fresh set of long-tail phrases. Second, the underscore trick: put an underscore where a word would go, like 'espresso _ grinder' - Google fills in the blank with popular options. Third, try your seed as a question: 'how to', 'why does', 'can I', 'best way to'. Write down every phrase that looks like something you could answer well. You're not filtering yet - just collecting.

• Type the seed, don't hit enter - read the dropdown suggestions
• Alphabet trick: add 'a', 'b', 'c'... after the phrase for new long-tails
• Underscore trick: 'your phrase _ word' shows popular middle words
• Question starters: how to, why does, can I, what's the best

People Also Ask Is a Question Machine

Search one of the phrases you just collected and look for the 'People Also Ask' box. Click one question and it expands - and three more appear. Keep clicking. Each answer gives you a mini-outline of what Google thinks satisfies that query, which is basically a cheat sheet for your own article. Then scroll to the very bottom of the search page for 'related searches' - eight more phrases, handed to you free. Collect these into your document too. Between autocomplete, People Also Ask, and related searches, a good session gives you 20 to 40 real phrases from one seed, all things people actually typed. Questions from this step make great article titles almost word-for-word, because the headline matches the search exactly.

Google Trends: Your Reality Check

Before you fall in love with a phrase, check it on Google Trends. Type in two competing versions of the same idea - say 'cold brew recipe' versus 'iced coffee recipe' - and see which one is actually growing. Trends also shows you seasonality, which saves you from publishing a Christmas gift guide in March. That iced coffee phrase probably spikes every summer and dies in winter; write it in spring and you'll catch the wave. Watch out for phrases that trend down year after year - sometimes interest is just fading, and no amount of good writing fixes that. Trends won't give you exact search numbers, but for a beginner it's plenty. Rising or steady? Good. Dying or wildly seasonal without a plan? Pick something else.

The Free Tools Worth Your Time (and Their Honest Limits)

Google's own tools plus two freebies cover everything a beginner needs. Google Keyword Planner is free, though it needs a Google Ads account - you can create one and skip through the campaign setup without running ads. It shows search demand in wide ranges rather than exact numbers, like '100 to 1K searches a month.' That's enough to compare ideas: one phrase in the 1K-10K range clearly gets more searches than one in the 10-100 range. AlsoAsked shows you People Also Ask trees in one visual map, and AnswerThePublic turns a seed into a wheel of question phrases. Both have free tiers with daily limits, and I'll be straight with you: those limits are tight. Use them for your best seeds only, and do the bulk of your work with autocomplete and People Also Ask, which never run out.

Page 1 Tells You Whether You Can Win

This step is the whole game, and most beginners skip it. Search your phrase and actually look at who's ranking. If page one is full of forums, Reddit threads, Quora answers, YouTube videos, and small personal blogs with thin articles - that's a green light. Google is telling you there's no great answer yet, and a proper, complete article from you can take that spot. I've seen tiny sites outrank everyone on phrases like this within weeks. But if all ten results are giant brands, major publishers, or government and medical sites, walk away. You won't beat them in your first year, and that's fine. Your job isn't to fight giants. It's to find the hundred little phrases where nobody good has shown up yet.

• Green light: forums, Reddit, Quora, small blogs, thin outdated articles
• Green light: YouTube videos ranking where no good written guide exists
• Red light: all results from big brands or major media sites
• Red light: government or medical authority pages on health/legal topics

Match the Intent, or Don't Bother Ranking

Here's the part that tripped me up longest. A keyword isn't just words - it's a person wanting a specific thing. Search 'how to fix a leaky faucet' and every result is step-by-step instructions, because that's the intent. Search 'best espresso machine' and every result is a comparison or review, because the intent is to buy. As a beginner, you want phrases with how-to and problem-solving intent: how to, why does, fix, guide, tutorial. You can satisfy those with knowledge and clear writing alone. Buying intent - best, review, vs, cheap, discount - is where the big affiliate sites fight hardest, and their pages have years of authority. I don't touch those until a site is established. Write for the person who has a problem today, not the one shopping for a product.

Keep a Simple Spreadsheet (Messy Beats Memory Every Time)

You now have phrases, intent, and a sense of the competition - but if you keep it all in your head, you'll forget half of it by next week. One spreadsheet, four columns: keyword, intent (how-to / buying / comparing), what's ranking now, and your angle. The last column is the important one: how will your article be better than what's there? More complete, clearer steps, real photos, an actual answer instead of fluff. A few rows might look like this:

• why does my espresso taste sour | how-to | thin forum threads, no real guide | full 900-word guide with grind-dose fixes and photos
• how to clean espresso machine without chemicals | how-to | one old blog, weak steps | tested vinegar-vs-tablet method with photos
• espresso grind too fine symptoms | how-to | Reddit threads only | visual guide showing what bad shots look like

Rookie Mistakes That Cost Me Months

Three mistakes, all mine, so you can skip them. First, chasing one big keyword. I tried to rank for a massive single-word term my whole first year. Never happened. One long-tail phrase with fifty searches a month beats a giant keyword with zero chance of ranking - and fifty of those long-tails add up to real traffic. Second, ignoring intent. I once wrote a product review when everyone searching the phrase wanted instructions. It ranked nowhere because Google could see searchers bouncing. Read page one and give people what they're clearly asking for. Third, copying competitor topics blindly. A big site can rank for anything because of its authority; you can't. Their topic list is not your topic list. Pick phrases where the current results are weak, write the best answer on the internet for that phrase, and repeat. That's the whole strategy.

Frequently Asked Questions

Do I need Ahrefs or Semrush as a beginner?

No. They're excellent tools when a site is earning money and you're scaling up, but they solve problems you don't have yet. Everything in this guide - autocomplete, People Also Ask, Trends, and reading page one - does the real work for free. I ranked my first posts with zero paid tools. Spend your early budget on your domain renewal, not subscriptions.

How many keywords should one article target?

One main phrase, plus a few closely related ones you'll naturally cover while answering the question. If your main phrase is 'why does my espresso taste sour', you'll automatically touch on grind size, dose, and water temperature - those are your supporting phrases. Writing one article aimed at fifteen keywords usually produces a mess that ranks for none of them.

How long before keyword research pays off in rankings?

Honestly, weeks to a few months for the long-tail phrases this method finds. New sites don't rank overnight - Google needs time to find your pages, and longer to trust them. Don't judge one article after ten days. Judge the system after three to six months of steady publishing, and you'll see the curve bend.

Is Google Keyword Planner really free?

Yes, but you need a Google Ads account to open it. You can create the account and skip through the campaign setup steps without spending anything or running ads. Just know it shows demand in ranges, like '100 to 1K searches a month', not exact numbers. That's fine for comparing which of your ideas is bigger - it's not precise enough for detailed planning, and beginners don't need that precision anyway.