How to do keyword research without a budget
Free tools can genuinely do this job. Here is how to use them, and how to read what a search actually wants.
Keyword research answers one question: what words does the person you want to reach actually type? Get that wrong and everything built on top of it — page structure, headings, the content itself — is aimed at a question nobody is asking. The good news is that answering it properly does not require a paid tool. It requires patience and a willingness to read what search engines are already telling you for free.
Start with data about your own site, if you have any
If your site already has any history at all, Search Console's Performance report is the single best source of keyword data available, because it is not an estimate — it is the actual queries people typed before landing on your pages. Look specifically for queries where your site appears frequently but ranks low, or where impressions are healthy but clicks are not. Both point at content that is already somewhat relevant but not yet fully serving the query, which is often easier to improve than starting from nothing.
Autocomplete is free, real, and constantly updated
Typing a phrase into a search box and reading what it suggests before you finish is keyword research a search engine is doing for you, in real time, based on what people actually search. It reflects genuine query volume and genuine phrasing, not a guess. Do this with partial phrases too — typing your core topic followed by each letter of the alphabet in turn surfaces a wider spread of what people actually ask around that subject than the completions for the exact phrase alone.
People Also Ask and related searches
Search results pages carry two free, direct sources of related questions: the "People also ask" box that expands as you click through it, and the "related searches" list at the bottom of a results page. Both are generated from real query data and reveal the surrounding questions people have that a single target keyword does not capture on its own — which is often exactly the material worth covering in a piece of content aimed at satisfying a topic thoroughly rather than answering one narrow phrase.
The free tier of paid tools is not nothing
Several keyword tools built primarily for paid subscribers still offer a free, capped tier or a limited number of lookups per day without payment. These typically will not give you exact search volumes at scale, but they are enough to compare a handful of candidate phrases against each other and get a rough sense of relative popularity — which is usually the actual decision you are trying to make, rather than needing a precise number.
Reading intent, not just volume
A keyword's monthly search volume tells you how often it is typed. It tells you nothing about what the person typing it actually wants, and writing for the wrong intent is a far more common failure than picking a low-volume phrase. Before writing anything, search the exact phrase yourself and look at what is already ranking for it.
| What is currently ranking | What that tells you about intent |
|---|---|
| Product or category pages | People searching this want to buy, not read — a long explainer article will struggle here regardless of quality |
| How-to guides and tutorials | People want a process they can follow — comparison content or a sales page will not satisfy this |
| Comparison or "best X" articles | People are deciding between options — a single-product page competes poorly here |
| Definitions or short explainer pages | People want a quick, direct answer — a long guide risks burying the actual answer they came for |
Matching what you build to what is already succeeding for that query is a far more reliable signal than any volume estimate, because it reflects what a search engine has already tested against real user behaviour.
Long-tail phrases are where a new or small site can realistically compete
A broad, high-volume term is dominated by sites with years of accumulated authority behind them, and competing directly against that from a standing start is rarely realistic. A longer, more specific phrase — closer to how someone actually thinks when they have a precise problem — carries less volume individually but is far less contested, and a handful of them together can add up to meaningful, realistic traffic while a site is still establishing itself. This is not a lesser strategy to graduate away from; it is simply where the achievable opportunity sits for most sites, most of the time.
Several closely related phrases are often best served by a single, thorough page rather than a separate thin page for each minor variation. Creating a near-duplicate page for every keyword variant tends to create competition between your own pages rather than genuine additional coverage — see fixing duplicate content if that has already happened on your site.
Turning a list of phrases into a plan
Once you have a working list, group it by topic and by intent rather than treating it as a flat list to work through in order. See how to structure a website so it can rank for turning grouped topics into an actual site structure, and how to write content that actually ranks for what to do with a specific phrase once you have decided a page should target it.
Related reading
The structural decisions that make a site easy for both a visitor and a crawler to understand, made once and rarely revisited.
How to write content that actually ranksNo trick outranks a page that actually answers the question better than the alternatives already sitting in the results.
How to set up Google Search ConsoleVerifying a site takes minutes. Here is the fastest verification method for your setup and what to look at once you are in.
Does HTTPS affect SEO?Google has confirmed it counts, though as a lightweight signal. The larger effect is what happens without it.