Keyword Clustering Tool
Paste a list of keywords (one per line) and the tool groups them into topic clusters using Jaccard word-overlap similarity. Adjust the threshold to control how aggressive the clustering is, then export the clusters as JSON for your content plan.
How to use
- 1
Paste your keywords
Enter one keyword or keyphrase per line. Stop-words like 'the', 'for', 'how' are removed before comparison.
- 2
Adjust the similarity threshold
The slider controls how aggressively keywords are grouped. 0.30 = loose clustering (more clusters); 0.80 = strict (fewer, tighter clusters).
- 3
Review the clusters
Each cluster lists its member keywords and the shared words that define it. The largest clusters usually indicate core topics.
- 4
Export as JSON
Copy or download the cluster structure as JSON. Drop it into your content-planning spreadsheet or use it to build topic silos.
Common use cases
- Group 200+ keyword-research exports into 10–20 content pillars.
- Identify cannibalisation risk: keywords that are too similar and may compete.
- Plan topic clusters for a new blog — one pillar page per cluster.
- Spot outlier keywords that don't fit any cluster (often new content opportunities).
Limitations
- Similarity is based on word overlap only (Jaccard), not semantic meaning. "Running shoes" and "sneakers for jogging" will not cluster despite meaning the same thing.
- Stop-words are removed before comparison. The shared-words list shows only meaningful tokens.
- Clustering uses single-link union-find. At low thresholds, chains of weakly-similar keywords can merge into large clusters.
- For very large lists (10,000+ keywords) the O(n²) comparison can be slow — paste in batches.
Frequently asked questions
Related tools
View allSERP Snippet Preview
Preview how your page looks in Google search results.
FAQ Schema Generator
Generate FAQPage JSON-LD schema from Q&A pairs.
Schema Validator
Validate JSON-LD structured data against schema.org.
Meta Tag Generator
Generate HTML meta tags for SEO and social sharing.