Overview
Monitor your AI visibility and citation performance across LLMs.
The GEO (Generative Engine Optimization) Dashboard tracks how often your site is cited as a source by AI models like Perplexity, ChatGPT, Gemini, and Claude. It shows your overall citation rate, breaks down performance by topic cluster, and automatically flags "content gaps," the clusters where you aren't being recommended, so your team knows exactly what to write next to improve AI visibility.
Header
- Last updated timestamp: when the dashboard's data was last refreshed.
- + Add Cluster: create a new topic cluster to track.
- Refresh: re-run the scan and pull the latest citation data.
Top stats
- Citation Rate: the percentage of tracked prompts where your site was cited, shown alongside the raw fraction (e.g. "cited in X of Y AI responses").
- Prompts Tracked: total prompt count being monitored, and how many clusters and AI models it spans.
Citation Trend
A line chart of your citation rate over a selectable window (default Last 7 Runs), with a trend indicator comparing the most recent run to the one before it.
AI Model Citation Rates
A per-model breakdown (Perplexity, ChatGPT, Gemini, Claude) showing how often each one cited your site out of all responses evaluated for that model, alongside the raw fraction. Bars are color-coded by performance band (green ≥ 50%, yellow ≥ 20%, red < 20%).
Citation rate varies significantly by model
Different models can cite your site at very different rates even when pulling from a similar number of evaluated responses. Treat each model as a separate distribution channel rather than expecting consistent citation behavior across all of them.
Biggest gap
A callout banner surfacing the single cluster with the most uncited prompts, with a Create content → shortcut and a note on when the cluster was last scanned and when its next scan is due.
"Next scan: Overdue" behavior is unverified
It's not confirmed here what triggers a scan to run once it's marked overdue: whether it's fully automatic, or needs a manual Refresh to kick off.
Clusters, Recommendations, and Insights tabs
Below the summary cards, three tabs organize the deeper data:
- Clusters: the full list of topic clusters being tracked. See Clusters for the per-cluster deep-dive.
- Recommendations: a badge shows the current count of open recommendations. See Recommendations.
- Insights: additional analysis surfaced from your citation data.
Building clusters from your site
Two actions help you go from a raw sitemap to trackable clusters:
- Scan Website Themes: crawls your sitemap to extract and categorize content themes.
- Generate Theme Prompts: generates ranking prompts per theme and creates clusters from them. Requires themes to be scanned first.
Theme scanning order matters
Generate Theme Prompts explicitly depends on Scan Website Themes having already run: attempting to generate prompts before scanning themes is likely to fail or return nothing.
Related Pages
All Prompts
Review the full list of AI prompts that triggered citations of your content.
Clusters
Group related prompts into clusters for theme-level analysis.
Competitors
Track competitor visibility in AI-generated responses.
Citation Domains
See which domains are being cited alongside your content in AI responses.