Mention Clustering uses AI to automatically group your incoming mentions by topic and narrative. Instead of showing you thousands of individual posts, it surfaces the key stories driving conversation about your brand.
Each group is called a cluster. A cluster might be a breaking news story, a viral social thread, a customer complaint pattern, or a PR event. Clusters are ranked by volume, so the biggest conversations always appear first.
Note: AI Mention Clustering processes mentions based on the selected trackers, filters, date range, and clustering settings. Changing the date range, tracker, or grouping settings and clicking Apply Changes will re-run clustering and reset your current selection.
Your trackers need to have at least 500-600 results at minimum for the platform to create clusters.
Setting up your clusters
Before applying clustering, configure the four settings in the bar above the cluster table. When you're ready, click Apply Changes.
Grouping
Grouping controls how closely related mentions need to be to end up in the same cluster.
Strict keeps distinct stories separate.
Loose merges loosely related mentions into broader themes.
Query Size
Query size is the number of mentions AI processes.
Start at 10k for speed.
Increase toward the maximum for broader coverage and more comprehensive clusters.
AI Summary Language
The language used when AI creates summaries of your clusters.
Defaults to Auto.
Change this if your team works in another language.
Summary Length
Summary length determines how detailed the AI summary will be.
Choose Shorter for quick executive briefs.
Choose Longer for more in-depth media monitoring reports.
Reading the Cluster Table
Column | Description |
Cluster Name | A representative post or AI-generated label describing the theme of the cluster. |
Mentions | Total number of individual mentions in the cluster. Click the column header to sort ascending or descending. |
Reach | Estimated total audience size exposed to mentions in the cluster. |
Engagement | Combined interactions across all mentions: likes, shares, comments, and reactions. |
Sentiment | A visual bar showing the positive/negative balance. Red indicates predominantly negative content; green indicates predominantly positive content. |
Channels | Platform icons showing where the conversation is happening: X/Twitter, Reddit, Facebook, YouTube, Instagram, TikTok, and others. |
Trackers | Which of your trackers captured the mentions in the cluster, with a count per tracker. |
Generating an AI Summary
Once you've reviewed the clusters, you can ask AI to create a narrative brief across any selection.
1. Select your clusters
Tick the checkboxes next to the clusters you want included. Use Select All to include everything, or hand-pick the clusters most relevant to your report.
2. Click "Create Summary"
The button appears in the top toolbar once at least one cluster is selected. AI processes all mentions within your selection.
3. Review and share your brief
The AI produces a summary in your chosen language and selected length. You can copy it directly into a report or stakeholder update.
Pro Tip: For a quick executive brief, select the top 3–5 clusters by mentions and set summary length to Shorter. For a full media monitoring report, select all clusters and switch to Longer.
Exporting cluster data
Use the download icon in the top-right of the toolbar to export your cluster data. This is useful for:
Sharing with team members who don't have access to Mentionlytics.
Archiving a snapshot of the conversation at a specific point in time.
Frequently Asked Questions
How is a cluster different from a regular mention feed?
A mention feed shows every individual post in chronological order. Clustering groups those posts by topic, so you can see the bigger conversation themes instead of a stream of isolated mentions.
Why might two similar stories appear in separate clusters?
Your Grouping slider may be set to Strict. Try moving it toward Loose and clicking Apply Changes. This broadens how AI defines similarity and can merge related clusters together.
Does changing the query limit affect my mention count?
Yes. A lower query limit means fewer mentions are processed, which can cause smaller clusters to disappear entirely. If you suspect you're missing conversations, increase the limit toward the maximum.
Can I run mention clustering for multiple trackers at once?
The Clustering works with one or multiple selected trackers. If you change your selected trackers, use the tracker dropdown at the top of the page and perform the changes.
Can I include AI Clusters in my reports?
Yes, you can add AI clusters to your Reports from if you’re a Pro or higher-tier plan user, as you require the Pro+ feature of Customizable Reports. Here’s how to do it:
Then, you may have an executive summary slide in your reports based on your mention clustering.
And you can also have your mention clusters as Top Trending Topics in a separate slide.