Content planning
Turn a flat list of ideas into a page plan: one cluster, one page, one search intent. The group name is usually the head term for the title and H1.
Keyword clustering groups search terms that one page can answer, so each cluster becomes one page and one target URL. MrKeyword clusters the ideas from one seed keyword in one market: up to 500 ideas from Google Keyword Planner, grouped by the words they add to the seed and a related-terms table, with volume, difficulty and intent on every group.
Free account, no credit card. You'll see the results right after you sign in.
Built on first-party data from Google and Bing
Start with one seed, then read the groups before you write anything.

Enter the broadest phrase that describes the topic, and pick the market: the United States, the United Kingdom, Canada, Australia, India or worldwide. The seed and market are carried through sign-in.

MrKeyword asks Google Keyword Planner for up to 500 ideas, then groups the ideas it receives. The tab appears when there are at least 20 ideas and 2 groups of 3 or more.

Read the group name, size, total searches, average difficulty and intent mix. Choose the groups that match pages you can write, then map every keyword in a group to the same URL.
The input is deliberately narrow, because the groups are only as good as the ideas behind them. This is exactly what the tool accepts and what it returns.

| Item | What the tool supports |
|---|---|
| Input | One seed keyword, ideally one to three words, plus one market. |
| Markets | the United States, the United Kingdom, Canada, Australia, India or worldwide. |
| Ideas per search | Up to 500 ideas from Google Keyword Planner, fetched live when you search. |
| Grouping method | The words each idea adds to the seed, clustered with MrKeyword's lexical-semantic clusterer and a related-terms table. No live Google results and no shared-URL data. |
| Group limits | At least 3 keywords per group, at least 20 ideas and 2 groups before the tab appears, and up to 40 groups shown. |
| Output per group | Group name, keyword count, total monthly searches of distinct keywords, average difficulty and intent mix. Selecting a group scopes the other research tabs to it. |
| Not supported | Pasting your own keyword list for clustering, clustering a competitor's list, and SERP-overlap clustering. Those need a different tool or a manual review. |
Clustering is a planning step, not a report. These are the jobs it does before a page is written or a campaign is built.
Turn a flat list of ideas into a page plan: one cluster, one page, one search intent. The group name is usually the head term for the title and H1.
Two groups that describe the same job should be one page, not two. Spotting that before writing is cheaper than merging the pages later.
Assign every keyword in a group to the same URL as a primary or secondary term, so the whole team works from one map instead of a spreadsheet of loose phrases.
Each tight group can become an ad group with one message, which keeps the ad and the landing page aligned with the query.
On a new site, the groups show which sub-topics exist before you commit to a content calendar, and which ones are small enough to win first.
A group table is easier to explain than a list of 500 rows: it shows the topic, the demand, the difficulty and the intent in one line per page.

MrKeyword clusters the words each idea adds to the seed, not the live Google results. Every idea shares the searched phrase, so the clusterer removes the seed's words and groups what is left: "price", "near me", "how to" and similar modifiers. A group is named after the word most of its keywords share beyond the seed, which is why the names read like topics rather than like the first keyword in the list.
The same search also returns a related-terms table, and the Groups tab uses it alongside the clusterer. Group totals count distinct keywords only: a plural or a reordered phrase is one variant set, counted once at its highest reported volume, so a group total is not inflated by near-duplicates.
This is lexical-semantic clustering, not SERP-overlap clustering. It is fast, costs no extra lookup and works from the ideas already on screen. It cannot tell you which pages rank for a phrase, because MrKeyword never scrapes Google. If shared ranking URLs are the signal you need, use a tool that collects live results and check its country and date settings.
Keyword clustering groups queries; a topic cluster is a group of pages linked around a pillar page. Clustering answers "which keywords belong on one page?" Topic clusters answer "which pages should link to each other, and in what order?"
Use clustering first to decide the pages, then build the topic cluster around the pillar page those pages support. Our guide to keyword clustering walks through the manual method and a worked example, and keyword mapping shows how to record the result as a URL-by-keyword table.

Pick the groups that match pages you can actually write, and start with the ones whose difficulty and intent fit your site. A group with a clear commercial intent belongs on a service or product page; a question group belongs in a guide. The keyword difficulty checker explains the score beside each group, and the keyword intent checker labels a list you already have.
Then map one URL to each group and keep the map. The master keyword list guide shows the sheet, and keyword cannibalization covers what to do when two pages already compete. To track the result, add the head term of each group to the free rank tracker, which reads positions from your own Search Console.
Type a seed above and the groups are one sign-in away.
Cluster keywords freeYes. Clustering runs inside Keyword Research on a free MrKeyword account: no credit card, and the Free plan never expires. The search that supplies the ideas uses 1 of your 10 monthly keyword lookups; opening a group or switching tabs uses none.
One seed keyword and one market at a time. MrKeyword asks Google Keyword Planner for up to 500 ideas around that seed, then groups the ideas it receives. The market can be the United States, the United Kingdom, Canada, Australia, India or worldwide. It does not accept a pasted list of your own keywords for clustering, and it does not cluster a competitor's keyword list.
It clusters the words each idea adds to the seed, using the same lexical-semantic clusterer as the rest of Keyword Research, plus a related-terms table. A group is named after the word most of its keywords share beyond the searched phrase, such as "price" or "near me". It does not use live Google results or shared ranking URLs.
A group needs at least 3 keywords, and the Groups tab only appears when a search returns at least 20 ideas and at least 2 groups. Up to 40 groups are shown; the rest are counted as ungrouped rather than split into tiny groups.
The group name, how many keywords it contains, the total monthly searches of its distinct keywords, the average difficulty and the mix of search intents. Close variants such as plurals and reordered words count once at their highest reported volume, so a group total is not inflated by duplicates. Selecting a group scopes the other research tabs to those ideas.
No. MrKeyword never scrapes Google, so it has no shared-URL or SERP-overlap data. Its groups come from the wording of the ideas and a related-terms table. If you need SERP-overlap clustering, use a tool that collects live results and check its country and date settings.
Use one cluster per page. Pick the group that matches a page you can write, make its head term the title and H1, and map the other keywords in the group to the same URL as secondary terms. That is the step that prevents two of your pages competing for one search; our guide to keyword cannibalization shows what happens when they do.
Clustering is a model of language, not a copy of Google. These are the edges of ours.
Free account, no credit card. 10 keyword lookups a month on the Free plan.
