Overview
Transform your keyword lists from chaotic spreadsheets into organized, actionable clusters. Our Keyword Clusterer uses advanced NLP and machine learning to group keywords by semantic similarity and search intent, helping you build better content strategies.
Perfect for
- Content planning and strategy
- PPC campaign organization
- Site architecture planning
- Competitive analysis
Performance Stats
Features
Multiple Algorithms
Choose from K-means, DBSCAN, hierarchical clustering, or let our system auto-select the best approach.
Semantic Understanding
Uses OpenAI embeddings to understand true keyword meaning, not just surface-level text similarity.
Intent Classification
Automatically categorizes clusters by search intent: informational, navigational, transactional, commercial.
Visual Clustering
Interactive 2D and 3D visualizations help you understand cluster relationships and quality.
Batch Processing
Handle large keyword lists (10K+ keywords) with efficient parallel processing and progress tracking.
Export Options
Export results to CSV, Excel, Google Sheets, or JSON with customizable formatting and metadata.
Usage Examples
Basic Clustering
from voidseo import KeywordClusterer
clusterer = KeywordClusterer()
# Load keywords
keywords = [
"best running shoes",
"top athletic footwear",
"marathon training tips",
"how to run faster",
"running shoe reviews"
]
# Cluster automatically
clusters = clusterer.cluster(keywords)
# Export results
clusterer.export_csv(clusters, 'keyword_clusters.csv')Advanced Configuration
# Custom clustering parameters
clusters = clusterer.cluster(
keywords,
algorithm='dbscan',
min_samples=3,
eps=0.3,
include_intent=True,
visualize=True
)
# Get cluster statistics
stats = clusterer.get_cluster_stats(clusters)
print(f"Silhouette score: {stats['silhouette']}")
print(f"Number of clusters: {stats['n_clusters']}")
# Visualize results
clusterer.plot_clusters(clusters, method='tsne')Clustering Methods
K-means
Best for: Large datasets with roughly equal cluster sizes
Pros: Fast, scalable, consistent results
Cons: Requires pre-defining number of clusters
n_clusters, init, max_iter
DBSCAN
Best for: Datasets with noise and varying cluster densities
Pros: Finds clusters automatically, handles noise well
Cons: Sensitive to parameters, struggles with varying densities
eps, min_samples, metric
Hierarchical
Best for: Understanding cluster relationships and nested structures
Pros: Creates cluster hierarchy, no need to specify cluster count
Cons: Computationally expensive for large datasets
linkage, distance_threshold, n_clusters
Auto-Select
Best for: When you're unsure which method to use
Pros: Tests multiple methods and selects the best performing one
Cons: Takes longer as it runs multiple algorithms
evaluation_metric, methods_to_test
Real-world Examples
E-commerce Site Architecture
Challenge: Online retailer had 50,000 product keywords with no clear organization.
Solution: Used hierarchical clustering to create a 3-level category structure.
Result: Improved site navigation, 34% increase in organic traffic, better internal linking.
Content Marketing Strategy
Challenge: SaaS company needed to prioritize content creation from 5,000 target keywords.
Solution: Clustered by intent and semantic similarity, then prioritized by search volume.
Result: Created 25 pillar pages covering 80% of keywords, doubled organic leads.