Free Keyword Grouping Tool – Cluster Keywords by Topic

What Keyword Grouping Does and How the Clustering Works

Keyword grouping takes a long, messy list of keywords and sorts it into clusters of related terms, so you can plan one page per topic instead of guessing which keywords belong together. This tool clusters your list by how much the keywords overlap in wording and lets you tune how tight the groups are. Anyone doing keyword research faces the same problem: a research export can run to hundreds of phrases, and mapping them to pages by hand is slow and inconsistent. Content strategists planning a site's structure, SEO specialists building topic clusters, bloggers deciding what to write next, and agencies organizing a client's keyword universe all use grouping to turn a flat list into a content plan. The method is lexical, meaning it groups keywords by the words they share. For each keyword the tool lowercases the text, splits it into words, drops common stop words like the and for, and applies light stemming so running and run count as the same root. It then measures similarity between any two keywords using the Jaccard index, which is the size of the shared word set divided by the size of the combined word set, giving a score from zero to one. Using single-linkage agglomerative clustering, a keyword joins a group if it is similar enough to any member of that group, and groups keep merging until none are close enough to combine. You control the cut-off with a threshold slider: a low threshold makes fewer, broader groups, while a high threshold makes many tight ones. Each cluster is named after its most common two or three word phrase. Here is a concrete pass. Feed in buy running shoes, running shoes online, best running shoes, and coffee maker with the threshold at 0.3. The first three all share the stemmed words run and shoe, so their Jaccard scores clear the threshold and they merge into one cluster named running shoes. The coffee maker phrase shares no words with them, scoring zero, so it lands in its own group or the ungrouped bucket. Raise the threshold to 0.6 and the running-shoe cluster may split, because best running shoes and buy running shoes share fewer words proportionally once the modifier differs. Watching the clusters reshape as you drag the slider is how you find the grouping that matches how you want to build pages. The practical uses are steady. A strategist pastes a five-hundred-keyword export and instantly sees the dozen topics hiding inside it, each becoming a planned article. An SEO specialist groups product keywords to decide which deserve a dedicated landing page and which are variations that belong on one page. A blogger clusters a brainstorm list to spot that three ideas are really the same post. An agency carries a search-volume column through the grouping and sums it per cluster, so the client can see which topics represent the most demand and prioritize accordingly. A couple of honest notes shape how you read the output. This is lexical clustering based on shared words, not semantic or SERP-based clustering, so it groups keywords that look alike rather than keywords that rank for the same pages. Paid enterprise tools cluster by which URLs appear in the search results for each term, which catches synonyms that share no words, and this browser tool deliberately does not do that because it would need live search data from a paid interface. Use the threshold to tune the balance between too few giant groups and too many singletons, and treat the ungrouped bucket as keywords that need a human decision. On very short or highly varied lists the groups can look thin, so lean on your own judgment alongside the clusters. Everything runs locally in your browser, so your keyword list stays private and is never uploaded.

🔥 Trending Tools