BUY WEBSITES
BUY DOMAINS
LOGIN
SIGNUP
Keyword Grouper & Clusterer
A keyword grouper takes a flat list of search terms and sorts it into clusters that share a common significant word, turning a thousand-row export into a handful of themes you can map to pages, ad groups, or content briefs. Instead of eyeballing a spreadsheet, the keywords organise themselves around the words that actually repeat across the list, which is the same signal that usually separates one topic or one landing page from another.
This tool is free and runs entirely in your browser. Paste the list, choose whether a cluster needs one shared word or two, remove stopwords so filler like the and best does not anchor a group, and read the clusters back with a live count and a one-click copy. Nothing you paste is uploaded, so it is safe for unpublished keyword research and client lists.
🔒 Private: everything runs in your browser. Nothing you paste is uploaded.
How to use the keyword grouper
Paste your keyword list, one term per line, and the clusters build instantly. Tune the two dropdowns and the stopword toggle to move between broad themes and tight clusters, then copy any group or export the whole map as CSV.
Choose between one-word themes and two-word clusters
The label-words dropdown controls how tight each cluster is. With one word, the tool groups every keyword around the single token it shares most across the list, so running shoes, trail shoes, and shoe reviews can all land under shoes. Switch to two words and the labels become adjacent pairs like running shoes, which splits that broad bucket into sharper, page-ready clusters. Start with one word to see the big themes, then move to two when a theme is large enough to deserve its own set of pages.
Remove stopwords so filler does not anchor a group
Left unchecked, the most common words in any keyword list are connectors and modifiers such as for, with, best, and cheap. Because they repeat everywhere, they outrank the real topic and pull unrelated keywords into one meaningless bucket. The stopword toggle strips a built-in list of English connectors and low-signal commercial modifiers before counting, so the frequency that drives clustering reflects subject words, not glue. Turn it off only when a modifier like a brand or a year is itself the theme you want to isolate.
Set the minimum group size to control the ungrouped bucket
The minimum dropdown decides how many keywords must share a word before it becomes a real cluster. At two, a token has to appear in at least two keywords to anchor a group, and anything too unique drops into the ungrouped bucket where you can review one-off terms separately. At one, every keyword finds a home, even if some clusters hold a single term. Use two to keep the output tidy for a large list, and one when you want a complete partition with nothing left behind.
Keyword grouper frequently asked questions
Q1How does the keyword grouper decide which group a keyword belongs to?
It tokenises every keyword, counts how often each significant word appears across the whole list, and assigns each keyword to the group of its most frequent shared word. When a keyword could fit several groups, the larger, more established cluster wins, with ties broken by order of first appearance. The logic is fully deterministic, so the same list and settings always produce the same clusters.
Q2What is the difference between one-word and two-word grouping?
One-word grouping clusters keywords around a single shared token, which surfaces broad themes but can lump several subtopics together. Two-word grouping uses adjacent word pairs as labels, so a wide theme like shoes splits into tighter clusters such as running shoes and trail shoes. One word is best for a first overview; two words give you page-level or ad-group-level clusters.
Q3Why should I remove stopwords before grouping?
Connector words like for, with, and the, plus generic modifiers like best and cheap, appear in almost every keyword. If they are counted, they become the most frequent words and anchor large, meaningless groups. Removing them lets subject words drive the clustering, which produces themes that actually map to topics and pages. You can switch the toggle off when a modifier is the theme you want to keep.
Q4Is this real clustering or just an alphabetical sort?
It is real frequency-based clustering, not a sort. The tool builds a frequency table of every significant word across your list, then groups keywords by the word they share most strongly, demoting any cluster that ends up smaller than your chosen minimum. Keywords are reassigned to their next-best shared word so no undersized group survives, which a simple alphabetical or substring sort cannot do.
Q5Does the keyword grouper send my list to a server?
No. All tokenising, counting, and clustering happens in your browser with client-side JavaScript, and the tool makes no network requests with your data. Nothing you paste is uploaded, logged, or stored anywhere, and the page keeps working offline once it has loaded, so it is safe to use with confidential or unpublished keyword research.
