Keyword research is the foundation every SEO project starts from. But most “keyword lists” we see in an audit are just a tool export of thousands of rows, where nobody sorted out what the user actually wants or which page should answer. A list like that doesn’t work. Here’s how to build research that genuinely drives your site structure and content plan, instead of sitting as a dead file.
What keyword research is and why it matters
Keyword research is an organized set of the search queries your audience uses to look for what you offer. Its job isn’t to “collect every keyword” — it’s to show real demand: what people ask, in what words, and with what intent. The research is what decides which pages should exist on your site at all, and what to write about.
Without it, content gets made on a hunch — and then people wonder why pages don’t bring traffic. With it, every page is built for a specific demand, not at random. That saves the most expensive thing of all: time spent on content nobody needs.
Intent matters more than search volume
The most common mistake is chasing search volume (how many times a month a phrase is searched) and ignoring intent — what the user is actually after. High volume is worthless if you don’t understand what the person wants to find. “Buy running shoes” and “how to choose running shoes” are different intents: the first leads to a product page, the second to an article. Put the wrong page type against a query and you get no conversions, even if the traffic shows up.
Define the query’s intent first, volume second. One precise query with the right intent is worth more than a dozen high-volume ones you have no matching page for.
Types of search queries
To keep intents straight, it helps to sort queries into types. Each type calls for its own page format:
| Query type | What the user wants | Page that answers it |
|---|---|---|
| Informational | to understand, learn “how/why” | article, guide, FAQ |
| Commercial | to compare and choose before buying | roundup, comparison, category |
| Transactional | to buy or order right now | product page, service, form |
| Navigational | to find a specific brand/page | homepage, brand page |
How to gather queries: where they come from
There are several sources, and you should combine them all. Search suggestions and the “people also search for” block show the audience’s real language. Search Console gives you the queries you already appear for — the most valuable source, because it’s real demand for your specific site. Competitors in the top results show which topics already work in the niche. Research tools (Ahrefs, Serpstat, Keyword Planner) widen the list and add volume data. The goal isn’t to gather as much as possible — it’s to gather what’s relevant.
Clustering: grouping queries into topics
A raw query list is useless until it’s grouped. Clustering means combining queries with the same intent into one group that maps to a single page. For example, “SEO price,” “cost of SEO,” and “how much does SEO cost” are one cluster for one page, not three separate ones. Simple test: if a user expects to land on the same page for two queries, they belong in the same cluster.
Clustering also protects you from cannibalization — when two of your pages compete for the same query and both slip. One cluster, one page.
Mapping queries to pages
Once the clusters are ready, each one is tied to a specific page: commercial clusters go to commercial pages and categories, informational ones to blog articles, transactional ones to product and service pages. For an online store, account for filters separately — some clusters land on filter pages, and there you have to avoid creating duplicates (more on that in our piece on filters and indexing). The output of this step is a “query → page” map that becomes the backbone of your site structure.
Why a bloated list hurts
There’s a myth that the more keywords in your research, the better. In reality, a bloated list with thousands of irrelevant queries just scatters your effort. You either try to cram everything into a few pages (and it turns to mush) or spin up hundreds of thin pages for each query — and Google doesn’t like those. Good research isn’t the biggest, it’s the most precise: only the queries you can and want to build a useful page for.
A site came to us with “research” of 6,000 queries and a hundred empty pages, one per keyword. We trimmed it to about 400 grouped queries and a few dozen substantial pages. Traffic didn’t drop — it grew, because Google started seeing structure instead of clutter.
Common mistakes when gathering queries
Research usually fails for the same reasons:
- Gathering by volume while ignoring user intent.
- Skipping clustering — leaving a raw list of thousands of rows.
- Pointing high-volume queries at pages that don’t answer them.
- Building a separate page for each keyword instead of a group.
- Including queries irrelevant to the business just for their volume.
- Never updating the research — demand shifts, the file stays last year’s.
- Forgetting branded and near-niche queries, which often convert well.
From research to a content plan
A finished “query → page” map is already most of a content plan. What’s left is setting priorities (start with clusters that have demand and a realistically reachable top spot) and deciding each page’s format. Then each cluster gets copy written for people and search at once — as we described in our piece on quality content and in the one on SEO copywriting. Before you start, it’s worth checking the structure against a technical audit, so the new pages land on a healthy site.
The takeaway
Keyword research isn’t a list of keywords — it’s a map of demand: what people search for, in what words, and with what intent.
Building research that works means not piling up thousands of rows, but understanding intent, grouping queries into topics, and honestly mapping them to pages. A precise set of a few hundred grouped queries will do more than a bloated file of thousands that nobody ever turned into a useful page.
Frequently asked questions
What is keyword research in simple terms?
It's an organized set of the search queries your audience uses to find you, grouped by topic and intent. Put simply, it's a map of demand that shows which pages your site needs and what to write about.
What matters more — search volume or intent?
Intent. High volume is useless if you don't understand the user's goal and you point a query at a page that doesn't answer it. Define intent first, then look at volume.
What is query clustering?
It's grouping queries with the same intent into one group that maps to a single page. If a user expects the same page for two queries, they belong together. It protects you from cannibalization.
Is it true that more keywords are always better?
No. A bloated list of thousands of irrelevant queries just scatters effort and spawns thin pages Google doesn't like. Good research is the most precise, not the biggest.
How often should I update keyword research?
Demand shifts, so review it regularly — at least every few months, and more often in fast-moving niches. Use new Search Console queries and seasonal demand changes as your cue.
Where do I start with keyword research?
With understanding your audience and their intent. Then search suggestions, Search Console data (real demand for your site), competitors' top queries, and research tools. After that, cluster everything and map it to pages.
Content manager at heleum.studio. Produces and publishes articles, video, and social formats — and makes sure content works for both people and search.
