Skip to main content
To learn more about Sentiment and Sentiment Score metrics in see the About Sentiment section.
The Sentiment Score node pulls brand sentiment data from your Profound account into an Agent. Use this step when you want to analyze positive, negative, or neutral sentiment toward your brand in AI answers. This node is optimized for understanding perception trends and tone shifts over time. Use Sentiment Score for tasks such as:
  • Tracking positive vs. negative AI answer sentiment
  • Identifying sentiment changes after product updates
  • Comparing sentiment across models or regions

​
Node configuration

Selecting the Sentiment Score node opens its configuration panel on the right side of the Agent Builder.

​
Metric

Choose the sentiment metric to return. Options include:
  • Positive Count
  • Negative Count
  • Total Occurrences
  • Positive Ratio (%)
  • Negative Ratio (%)

​
Date Range

Specify the time window to query (for example, Last 7 Days). The node will return sentiment data only for answers within this range.

​
Filters

Use filters to narrow the dataset. Fields may include:
  • Analysis Types
  • Asset
  • Citation Categories
  • Hostnames
  • Regions
  • Personas
  • Platforms
  • Prompts
  • Tags
  • Topics
You can paste a static value or insert a value dynamically from earlier Agent steps by typing /. When filtering on a field, you must also include that field as a dimension in the field settings.

​
Output Label

Examples:
  • sentiment_score
  • weekly_sentiment
  • sentiment_by_model

​
Advanced settings

​
Date Interval

Group the date interval by day, week, month or year.

​
Limit

Set the maximum number of rows returned by the query. This is helpful when using multiple dimensions or sending data into LLM steps.

​
Sort By

Sort the output by metric or date.

​
Sort Direction

Sort the output in ascending (ASC) or descending (Desc) order.

​
Output

Returns structured sentiment metrics grouped by any selected dimensions. This output is commonly used in:
  • LLM narrative summaries
  • Reputation monitoring Agents
  • Executive dashboards