Clusters & Patterns
Discover hidden patterns in your feedback with AI-powered clustering.
Overview
When feedback comes from dozens of sources and hundreds of messages, it is easy to miss that twenty different people are asking for the same thing in different words. Clusters solve this by automatically grouping semantically related ideas together.
Instead of reading every idea individually, you can review clusters to understand the top themes in your feedback, see how many people are affected, and prioritize based on real demand.
How Clustering Works
IdeaLift uses semantic similarity — not just keyword matching — to group ideas. Two ideas that use completely different words but describe the same need will still be clustered together.
Embedding generation
Each idea is converted into a numerical vector that captures its meaning, not just its keywords.
Similarity analysis
Vectors are compared to find ideas that are close in meaning. A configurable threshold determines how tight clusters are.
Cluster formation
Related ideas are grouped into named clusters. IdeaLift generates a summary label for each cluster based on its contents.
Continuous updates
As new ideas arrive, they are automatically matched against existing clusters or form new ones.
Viewing Clusters
Access clusters from the main navigation under Insights > Clusters. The clusters view shows:
- •Cluster cards — Each cluster has a name, idea count, source breakdown, and average signal score
- •Sorted by volume — The largest clusters appear first, highlighting the most-requested themes
- •Drill-down — Click any cluster to see every idea in it, with links back to the original conversation
- •Filter by source or date range — Narrow clusters to a specific time period or feedback channel
Merging Ideas
When you find ideas within a cluster that are truly duplicates, you can merge them into a single idea:
- Open a cluster and select the ideas you want to merge
- Click Merge Selected in the toolbar
- Choose which idea becomes the primary (its title and description are kept)
- All votes, comments, and source attributions from merged ideas carry over
- The merged idea's signal score is recalculated based on combined data
Merging is non-destructive. The original ideas are archived, not deleted, and can be unmerged if needed.
Cluster Insights
Beyond grouping, clusters reveal trends you can act on:
Growth trends
See which clusters are growing fastest. A cluster that doubled in size this month signals urgent demand.
Cross-source patterns
When the same theme appears in Slack, Discord, and support tickets, it validates the demand across audiences.
Cluster health
Clusters with mostly stale ideas indicate an unaddressed pain point. Clusters with shipped ideas show progress.
Revenue impact
On Scale plans, clusters show the total revenue of requesting customers, helping you prioritize by business impact.
Growth plan and above
Clusters & Patterns is available on the Growth plan and above. See pricing for details.