How to Do Keyword Research (Without Expensive Tools)

The keyword research industry has convinced a lot of founders that you cannot start SEO without a 99 dollar per month subscription. We disagree, and meeeters.com is the proof: most of the keywords this blog ranks for were found with free methods, mostly by reading Google itself.

Paid tools are convenient. They are not necessary. Everything they know about keywords, they learned by scraping Google. You can read the source directly. This guide is the full method: how to find keywords, how to judge whether you can win them, and how to decide what to write first.

First, understand search intent (or nothing else works)

Every query carries an intent, and Google's entire business is matching results to that intent. If your page type does not match the intent, you will not rank, no matter how good the content is or how many backlinks you point at it.

The four classic intent types:

• Informational: the searcher wants to learn. "what is domain rating", "how do backlinks work". Blog posts and guides win here.
• Commercial investigation: the searcher is comparing before a decision. "best backlink checker", "ahrefs vs semrush", "meeeters review". Comparisons, lists and reviews win.
• Transactional: the searcher wants to act now. "buy backlinks", "link building service pricing". Product and landing pages win.
• Navigational: the searcher wants a specific site. "search console login". You only rank for your own brand here, ignore the rest.

The test is brutal and simple: search the keyword and look at what ranks. If the top 10 for your target keyword is entirely blog guides and you planned a product page, Google has already told you the answer. Match the dominant page type or pick another keyword.

Intent also predicts value. Informational queries bring volume and trust, commercial queries bring comparisons shoppers, transactional queries bring buyers. A healthy site targets all three, weighted toward whatever stage your funnel is starving for.

Start with seed keywords

Seeds are the 5 to 15 core topics your business lives on. You do not find them in a tool, you find them in your head and your customers' mouths:

1. Write down what you sell, in the words a stranger would use. Not your internal jargon.
2. List the problems that push people to look for you. For us: "how to get backlinks", "site not ranking", "link exchange safe or not".
3. Steal vocabulary from customers: support emails, sales calls, onboarding questions. The exact phrasing people use when describing their problem is keyword gold.
4. Add your competitors' names. "Alternative to X" and "X vs Y" queries convert absurdly well.

Each seed will explode into dozens of real queries in the next step. Keep seeds broad ("keyword research") and let the expansion methods find the specific angles ("keyword research without tools").

Free expansion method 1: Google autocomplete

Type a seed into Google and pause. The suggestions are real queries, pulled from actual search behavior, updated constantly. Then go deeper:

• Type the seed plus each letter of the alphabet: "keyword research a", "keyword research b", and so on. Tedious, and it works.
• Put an underscore or asterisk in the middle: "keyword _ research" surfaces variations you would not guess.
• Prefix with question words: "how keyword research", "why backlinks", "can seo".

Do this in an incognito window so your history does not contaminate the suggestions. Twenty minutes of autocomplete work on one seed typically yields 30 to 60 real query variations. No tool subscription involved.

Free expansion method 2: People Also Ask and related searches

Search your seed and open the People Also Ask box. Every question is a query real people typed, and clicking one loads more. Four or five clicks deep, you have a map of every subquestion surrounding your topic.

These questions serve two purposes. The big ones become their own articles. The small ones become H2 sections inside a bigger article, which is exactly how you build pages that rank for hundreds of long-tail variations at once.

At the bottom of the results page, the related searches block gives you sibling queries, often revealing adjacent intents you had not considered. Copy everything into your sheet. Between autocomplete, People Also Ask and related searches, one seed usually produces a hundred candidate queries in under an hour.

Free expansion method 3: mine your own Search Console

This is the single most underrated keyword source in existence, because these are not candidate keywords. They are keywords Google has already decided your site is relevant for.

The striking-distance workflow:

1. Open Search Console, Performance report, last 3 months.
2. Switch to the Queries tab and sort by impressions.
3. Filter to queries with average position between 4 and 15.
4. These queries sit just below the positions that get real clicks. Google already shows you for them; it just does not trust you enough yet.
5. For each one, either strengthen the page that ranks (better section, better title, more internal links) or, if no page truly targets the query, create one.

Moving one query from position 9 to position 3 routinely means five to ten times the clicks. It is far cheaper than ranking a new keyword from zero. If you have never set this up, our complete Search Console guide walks through everything, and it should be your first stop before spending a single euro on tools.

The other Search Console goldmine: queries where you get impressions on page 3 and beyond for topics you never wrote about directly. Google is telling you it wants to rank you for this. Write the page.

Free expansion method 4: competitor SERPs and communities

You cannot see a competitor's keyword list without a paid tool, but you can reverse-engineer the important parts:

• Search your seeds and study who keeps appearing. The sites ranking for three of your seeds probably rank for fifty queries you want. Browse their blog. Their article titles are their keyword strategy, published in plain sight.
• Use the site: operator. Searching site:competitor.com keyword shows every page they built around a topic, which reveals how they cluster it.
• Read Reddit, forums and communities in your niche. The questions asked over and over are queries, phrased in raw customer language. When a Reddit thread ranks on page one of Google for a query, that is a double signal: real demand, weak competition.

Estimating difficulty by reading the SERP

Here is the part paid tools genuinely get wrong most often. Tools compress difficulty into a single score, usually derived from the backlink counts of ranking pages. The score misses page type, brand weight, intent match and content quality. The SERP itself misses nothing. You just have to read it.

Search the keyword and grade what you see:

Who ranks?

• Big brands and aggregators everywhere (Forbes, HubSpot, Wikipedia, major SaaS): a new site will not crack this page for years. Skip or find a longer-tail angle.
• Mid-size niche sites: winnable with strong content and some links.
• Small sites, personal blogs, DR under 30: green light. If they can rank, you can.
• Forum threads, Reddit, Quora in the top 5: flashing green light. Google is ranking user-generated scraps because nobody wrote a proper page. Write the proper page.

You can check the authority of ranking domains with free DR checkers (Ahrefs offers one). If the concept of domain authority is fuzzy, we explain what it measures and what it does not in what is domain rating.

How dedicated are the ranking pages?

A page that targets your exact query is hard to displace. A page that ranks incidentally, where your query is one paragraph inside a general article, is soft. If the top 10 is mostly incidental rankings, a dedicated page from a smaller site beats them regularly.

How good is the content, honestly?

Open the top 3 results and read them. Outdated screenshots, thin answers, generic AI filler, missing subquestions: every weakness is your opening. If you finish reading and think "I could genuinely answer this better", the keyword is winnable regardless of what any difficulty score says.

A simple scoring habit

For each candidate keyword, spend ninety seconds on the SERP and note three things: strongest domain type in the top 5, whether dedicated pages dominate, and whether you can honestly do better. Three answers, and you know more than any difficulty score would tell you.

The long-tail strategy (how small sites actually win)

Head keywords ("seo", "backlinks") get the volume and are unwinnable for years. Long-tail keywords, the specific 4-to-8-word queries, get modest individual volume but:

• They are dramatically easier to rank for, often with zero dedicated competition.
• They convert better, because specificity signals intent. Someone searching "link exchange safe 2026" is far closer to a decision than someone searching "backlinks".
• They compound. One solid article ranks for dozens of long-tail variations at once, because Google matches content to meaning, not exact strings.
• They build topical authority. Ranking for forty specific questions about link building is precisely what earns you the right to rank for "link building" later.

This is the entire early-game playbook, and it is why keyword research matters more for a small site than a big one. Big sites can afford to be lazy. You cannot, and fortunately, you do not need to be. If your site is young, our 90-day roadmap for new websites shows how this long-tail approach fits into a full launch plan, and how long SEO takes sets honest expectations for when the compounding kicks in.

Mapping keywords to pages

A keyword list is not a plan. The plan is the mapping: which page targets which cluster.

The core rule: one intent, one page. Not one keyword, one page. "keyword research without tools", "free keyword research methods" and "how to find keywords for free" are one intent, and one page should target all of them. "keyword research" and "keyword research tools" are different intents (learning a method vs comparing products) and need different pages.

The tiebreaker when you are unsure: search both keywords. If the top 10 results overlap heavily, one page covers both. If they barely overlap, split them.

Practical mapping steps:

1. Group your keyword list into intent clusters. Every cluster gets a primary keyword (usually the highest-volume phrasing) and a list of variations.
2. Assign each cluster to a URL, existing or planned. If two clusters map to the same URL, merge them or split the page.
3. Give each planned page a working title that contains the primary keyword naturally.
4. Note the dominant page type from the SERP (guide, list, comparison, tool) so you build the right format.
5. Plan the internal links between cluster pages up front. A cluster where every article links to its siblings and up to a hub page ranks better as a group, which is the whole logic behind internal linking for SEO.

This mapping also prevents the quiet killer of growing blogs: two of your own pages competing for the same query, splitting signals, and both ranking worse than either would alone.

Prioritization: what to write first

You now have more keyword clusters than time. Score each cluster on two axes, business value and winnability, on a simple 1 to 3 scale, and multiply.

Two refinements from experience:

• Striking-distance queries from Search Console jump the queue. They are half-won already. Improving an existing page beats writing a new one, hour for hour.
• A few P3 pieces early are not wasted. Easy wins teach you the process, give Google something to index, and start building the topical footprint that makes P2 keywords reachable later.

Revisit the scoring quarterly. Winnability changes as your site earns authority: a SERP that was out of reach at DR 10 becomes a fair fight at DR 30.

A worked example: from seed to published page

Theory is easy to nod along to, so here is the method applied end to end, the way we actually ran it for this blog.

Seed: "link exchange". It is a core topic for us, and customers use those exact words.

Expansion: typing "link exchange" into an incognito Google window surfaced autocomplete suggestions like "link exchange seo", "link exchange safe", "is link exchange good for seo" and "link exchange vs guest post". The People Also Ask box added "Do link exchanges still work?", "Are reciprocal links bad for SEO?" and "How do I find link exchange partners?". Ten minutes in, we had about forty real queries.

Intent grouping: those queries collapse into a handful of intents. "Is it safe, does it still work, is it good or bad" is one worried-evaluator intent. "How to find partners" is a separate how-to intent. "Vs guest posting" is a comparison intent. Three intents, three pages, not forty pages.

SERP check: for the safety cluster, the results included a couple of dated articles, a thin listicle and a Reddit thread in the top 10. Winnable. For the broad query "link building", the SERP was Ahrefs, Semrush, Moz and Backlinko wall to wall. Parked without regret.

Priority score: the safety cluster scored high on business value (people asking it are our exact future users) and high on winnability. P1. It got written first, and it ranked, and it still sends signups today.

The loop: once the page had a few weeks of data, Search Console showed impressions for phrasings we had not targeted, which fed the next round of articles. That is the system working: every published page becomes a keyword research instrument for the next one.

Run this exact sequence on one of your seeds this week. The first pass takes an afternoon. By the third seed, it takes an hour.

Two free tools that fill the remaining gaps

Two more free sources worth knowing before anyone reaches for a credit card:

• Google Trends answers the direction question that autocomplete cannot: is this topic growing, seasonal, or dying? Comparing two candidate keywords in Trends also gives you their relative popularity, which partially substitutes for volume data. Writing about a topic six months before its seasonal peak is one of the oldest free wins in SEO.
• Google Keyword Planner (inside a free Google Ads account, no spend required) shows volume ranges like 100 to 1K for any keyword list you paste in. The ranges are coarse, but coarse is enough for prioritization: you mostly need to know whether a query is a 10, a 100 or a 10,000.

Between Google's own surfaces, your Search Console, and these two, the free stack covers discovery, intent, difficulty, direction and rough volume. The only thing left unpaid-for is convenience.

What paid tools actually add (and when to pay)

To be fair to the tool vendors, here is what a subscription genuinely buys you:

• Search volume estimates. Free methods tell you a query exists, not how big it is. Volumes are estimates with wide error bars, but they help rank opportunities.
• Competitor keyword exports. Seeing every keyword a competitor ranks for, sorted by traffic, compresses weeks of SERP reading into an afternoon.
• Scale and speed. Everything in this guide, faster, in bulk, with alerts.

Our honest recommendation: run the free method for your first three to six months. You will learn to read SERPs, which is a skill tools cannot replace and tool users often never develop. Pay for a tool when your bottleneck becomes hours rather than knowledge, not before. Many solid keyword-to-page decisions on this very site came from nothing more than autocomplete, People Also Ask, and Search Console.

From keywords to rankings

Keyword research finds the demand. Ranking for it takes three more ingredients: content that actually answers the query (our on-page SEO checklist covers the mechanics), internal links that concentrate your authority, and for the competitive clusters, backlinks from relevant sites.

That last ingredient is where most small sites stall, and it is the problem Meeeters was built to solve: you swipe relevant partner sites, give an editorial link, and earn a verified link back from the network. The exchange model removes the reciprocal pattern, meaning your return link never comes from the site you linked to, and every placement is niche-relevant and editorial.

Start where every good SEO decision starts, with data: get your free SEO analysis, see which keywords and pages are closest to winning, and earn your first verified backlink through the Meeeters network.

Christopher Fernandes

Founder of Meeeters

I built Meeeters to make link building safe and simple: real, relevant backlinks with no reciprocal footprint and no black-hat shortcuts. Questions about your site? Write to me directly.

Email us

We reply fast, usually within a few hours.