Most people quit keyword research before they start, because the first advice they read is "buy a ₹7,000-a-month SEO tool." You do not need one. A free keyword research workflow — built on Google Search Console, Keyword Planner and Google's own search box — will surface most of the keywords a small site can realistically rank for. This guide gives you that workflow, step by step, plus the one 2026 change that decides which keywords are still worth the effort at all.
Source: Similarweb and SEO Inc zero-click analyses, 2026.
Why keyword research is the SEO skill you cannot outsource
Keyword research is simply finding out what your audience types into a search bar, in their words, and how much competition sits behind each phrase. Skip it and you write for yourself. Do it well and every article starts with a reader who is already looking for the answer.
The reason it feels expensive is that the loudest voices in SEO sell subscriptions. Their difficulty scores and volume numbers are genuinely useful, but they are conveniences, not the actual work. The real signal — what your own audience searches, and what Google already shows for it — is free and public.
For a coaching institute, a local shop, a solo consultant or a student building a first portfolio, the free path covers roughly 90% of what a paid suite does. If you want that foundation built properly rather than pieced together from scattered videos, a structured digital marketing course compresses months of trial and error into a few guided weeks.
Google Search Console: the free keyword data you already own
If your site has any traffic at all, your single best keyword source is Google Search Console (GSC). It is free, and it reports the exact queries real people used to reach you — not an estimate, but observed searches.
Open the Performance report and click the Queries tab. You get five columns per query: Query, Clicks, Impressions, CTR and Average Position. Sort by impressions and you will find dozens of searches where Google already shows your page but few people click yet.
The gold sits in positions four to nine. A page ranking there is one improvement away from page-one clicks. Positions one to three usually just need a sharper title or meta description to lift the click-through rate. Anything on page two is a content-depth problem worth solving.
Two limits to know. GSC caps each report at 1,000 rows, so a large site will not see every query — filter by page or folder to dig deeper. And a brand-new site with no impressions has no GSC data yet, which is where the next two free sources take over.
Two GSC habits sharpen the signal further. First, use the compare date range — set the last 28 days against the previous 28 — to spot queries that are climbing, which are cheaper to push than cold ones. Second, add a query filter for question words like “how”, “why” and “what” to isolate the informational searches your blog can answer, separate from the branded terms that always sit at the top. Both take seconds and cost nothing. Once you are acting on this data, pair it with the five Google Analytics 4 reports that actually matter to see which of those keywords convert, not just which get clicks.
Expanding the list: autocomplete, People Also Ask and Keyword Planner
GSC tells you what you already rank for. To find new territory, mine Google itself.
Autocomplete is intent research in disguise. Type your core phrase and read the suggestions Google offers — these are real, popular queries. Add a letter (“keyword research a…”, “keyword research b…”) and the list refreshes each time. People Also Ask boxes and the “Related searches” strip at the bottom of the results page hand you the exact questions and adjacent phrases people chase next.
Take a real example. Start with “stock market course” and Google may suggest “stock market course for beginners”, “stock market course fees”, and “stock market course near me”. Three suggestions, three different intents — a learner, a buyer, and a local searcher — and each deserves its own page. That is a full content cluster surfaced in under a minute, from a search box, for nothing.
For rough demand, use Google Keyword Planner. It is free the moment you create a Google Ads account, and you do not need to run a single ad to open it. There is one catch worth understanding before you trust the numbers.
Free Keyword Planner shows volume in broad ranges, not exact numbers
Source: rankdots and w3era Keyword Planner guides, 2026.
Without active ad spend, Keyword Planner reports volume in wide bands — 10, 100, 1K to 10K, 10K to 100K. A single band can span an order of magnitude, so treat it as a rough tier, not a precise figure. That is enough to separate a dead phrase from a busy one, which is all you need at the research stage. Exact monthly numbers only unlock once a live campaign runs, and you do not need them to decide what to write.
The 2026 filter: which keywords are still worth targeting
Here is the change that reshapes every keyword list this year. AI Overviews — the AI-generated answer boxes at the top of many results — now sit on a large share of searches, and they change what a ranking is worth.
On searches that trigger an AI Overview, roughly 83% end without any click, against about 60% on searches without one. Organic click-through on AI-Overview queries has fallen around 61% since mid-2024, from 1.76% down to 0.61%.
AI Overviews push far more searches to zero clicks
Source: Similarweb zero-click marketing data, 2026.
The takeaway is not "give up." It is that you should sort keywords by how likely a click still is, and spend your free effort where clicks survive. A simple three-tier read does the job.
| Zero-click tier | Typical query | What to do |
|---|---|---|
| 0–30% (click-rich) | “digital marketing course fees”, “NIFM login” | ✓ Target hard. Standard SEO still wins the click. |
| 30–70% (mixed) | “how to do keyword research” | Earn the SERP feature and stay compelling enough to win the click. |
| 70–100% (answer-only) | “what is a keyword”, “GST rate” | • Measure brand mention and snippet capture, not traffic. |
Transactional and navigational phrases — fees, near-me, login, buy, admission — stay click-rich because the searcher needs a destination. Pure-definition queries increasingly get answered on the page. We covered this shift in depth in our note on Google's 2026 spam update and AI Overviews, and it should sit at the top of your prioritisation, not as an afterthought.
Want to read search data like this yourself?
NIFM's online digital marketing training walks through GSC, Keyword Planner and on-page SEO with live examples, in Hindi and English, with a certificate on passing the course exam.
Explore the Digital Marketing course →The free workflow, start to finish
Put the pieces in order and you have a repeatable routine you can run for any page or topic in an afternoon.
Step four is where free beats lazy. Without a paid difficulty score, you judge competition by reading the results page. Search your target phrase and ask: are the top results big national brands or forums and small blogs? Do the pages actually answer the phrase, or only mention it? A query where the top ten are thin, off-topic or user-generated is a query a focused page can beat — no score needed.
For step five, a plain spreadsheet with five columns is enough: keyword, source, volume tier, zero-click tier, and search intent (informational, commercial or navigational). Sort by intent, then by click-likelihood, and your content calendar writes itself. Local businesses should give an extra pass to city-and-service phrases — our guide to local SEO for small business shows how those keywords feed straight into a Google Maps listing.
Mistakes that quietly waste the free approach
The free method works, but a few habits blunt it:
- Chasing volume alone. A 10K-band keyword you cannot rank for is worth less than a 100-band phrase you can own. Relevance and winnability beat raw size.
- Ignoring intent. Ranking a “how to” page for a phrase people search to buy something will convert almost no one. Match the page type to the intent.
- Treating Keyword Planner bands as exact. They are tiers. Do not build a business case on the midpoint of “1K to 10K.”
- Skipping the SERP read. Volume and difficulty guesses mean nothing until you have looked at who currently ranks and why.
- Forgetting AI Overviews. Pouring effort into pure-definition keywords that now get answered on the page is the most common 2026 waste.
None of these cost money to fix. They cost attention, which is the one input the free workflow actually depends on.
What to do next
Open Google Search Console today, sort your Queries report by impressions, and list ten phrases sitting in positions four to nine. Run each through the free workflow — expand, tier, read the SERP, sort by intent — and you will have a costed-at-zero content plan by evening. The paid tools can come later, once traffic justifies them; they speed the work up, they do not unlock it.
Keyword research is a learnable skill, not a subscription. Learn to read the free signals well and you will out-target competitors who pay for tools but never look at what the data is telling them.
Learn SEO and keyword research the structured way
Trusted by 50,000+ learners since 2012 · Hindi + English · Learn at your own pace
Start the Digital Marketing courseFrequently Asked Questions
Can you really do keyword research for free?
Yes. Google Search Console, Keyword Planner, autocomplete and the People Also Ask boxes together cover most of what a small or mid-sized site needs. Paid suites add speed and precise difficulty scores, but the core signals — real queries and who ranks for them — are public and free.
Is Google Keyword Planner free without running ads?
It is free to open once you have a Google Ads account, and you do not have to run an ad. The trade-off is that without active spend it shows search volume in broad bands rather than exact numbers, which is fine for research-stage tiering.
How do I judge keyword difficulty without a paid tool?
Read the results page. Search your phrase and look at who ranks: national brands and authority sites signal a hard keyword; forums, small blogs and loosely relevant pages signal an opening. Whether the ranking pages truly answer the query matters more than any single score.
How do AI Overviews change keyword research in 2026?
They make click-likelihood part of the decision. Definition-style queries increasingly get answered inside the AI Overview, so favour transactional, local and navigational keywords where searchers still need to visit a site, and measure answer-box presence for the rest.
Which free keyword tool should a beginner start with?
Start with Google Search Console if your site already has traffic, because it shows searches you can act on immediately. If the site is brand new, start with autocomplete and People Also Ask to map intent, then use Keyword Planner to tier demand.
How many keywords should one blog post target?
One primary keyword and two or three closely related variations. A single post cannot rank for a dozen unrelated phrases, and trying to force them in reads badly to both people and search engines. Group the rest into their own posts — that is exactly what the intent-sorted sheet in this workflow is for.
Do free keyword methods work for local businesses in India?
They work especially well. City-plus-service phrases such as “digital marketing course in Delhi” are often lower competition and highly transactional, and they stay click-rich even with AI Overviews because the searcher wants a nearby option. Google Search Console and autocomplete surface these local variations for free.
.jpg)
