Sentiment Analysis — How AI Engines Describe Your Brand

Where it is

The sentiment view is not in the sidebar. Open it from the Workspace: the "Sentiment ↗" link next to the core-metric tabs, the "Positive sentiment" stat under the visibility gauge, or the "Discover trending themes" card under Recommended actions. Data comes from the same scans as Prompt Tracking; run a check there first.

Sections

• Sentiment overview. Sentiment index, average score, number of checks and the negative share.
• Sentiment trend. Appears after several checks.
• You vs competitors. How engines talk about each brand in the same answers.
• Sentiment by attribute. Pricing, reliability, features, support, ease of use, performance and reputation.
• Sources shaping negative perception. The cited sources most associated with negative statements. These are the first outreach targets.
• Themes. Topics engines associate with your brand, filterable by positive, negative or neutral, with search.
• Recent mentions. Raw answer, sentiment reading, link to the full record, and a button to generate an optimization suggestion.

Export CSV is available on paid plans with data export.

How to use it

1. Look at the negative sources list. Each entry is a page you can act on: reply, request a correction, or publish a better source.
2. Check the attribute breakdown. A weak "pricing" or "support" attribute usually points to a missing public page (pricing, support policy) that engines cannot find.
3. Use the theme filter to find topics you would rather not be associated with and address them in Content Rules or new content.

Frequently asked questions

Why is it empty? Sentiment is derived from prompt scans. Add prompts and run a check in Prompt Tracking.

Is this the old "AI Insights feed"? Yes. The page now focuses on sentiment and source tracing; trend and opportunity views moved to the Workspace and Citation Sources.