Sentiment

This guide covers the Sentiment feature within the AI Visibility tool.

The Sentiment feature in AI Visibility is a specialized feature that turns complex, unstructured LLM outputs into a clear, actionable snapshot of your brand’s reputation. With this tool, you can gauge how LLMs and AI agents are talking about your brand in response to your tracked prompts.

Your brand sentiment is the overall tone AI models express when discussing your brand across tracked prompts. Sentiment analyzes full statements in context, just like a human reader, to determine if the attitude is positive, negative, or neutral.

To access the Sentiment view within AI Visibility:

1. Click AI Research from the left-hand navigation menu
2. Click AI Visbiility
3. Select your Dashboard
4. Click Sentiment from the top navigation

The first set of modules within Sentiment is Brand Sentiment. This includes your overall Sentiment score, top themes, and a competitor comparison.

Overall Score

The Overall Score is determined on a scale of 1 to 100, where the higher the score, the more positively your brand is being discussed by LLMs for your tracked prompts.

Your score is calculated by looking at all the times your brand is mentioned in prompt responses across your tracked prompts, identifying which mentions express a positive or negative sentiment, and turning that into an overall score. Higher scores indicate more positive sentiment.

Top Themes

Within the Top Themes box, you’ll see the themes where your brand’s sentiment scores are especially high or low. This can help you identify areas of opportunity and capitalize on your brand’s strengths.

To identify your top themes, we analyze all your brand mentions across your tracked prompt responses and automatically group them. Each theme then receives an individual score based on how positively your brand is being discussed within that theme.

Competitor Comparison

The Competitor Comparison module helps you see how your brand sentiment compares with your tracked competitors. You can then use this information to identify where you need to close the gap.

In the table, you’ll see the brands being tracked, their overall Sentiment score, a level of confidence, and a Mentions Breakdown.

• Sentiment Score - We calculate sentiment for your brand and any tracked competitor brands. Each sentiment score also gets a confidence level based on how much the brand is actually discussed. Higher confidence means more brand mentions. 
• Mentions Breakdown - Breaks down how many times each brand was mentioned across all prompt responses for your tracked prompts in total, and a breakdown of how many mentions expressed a positive, neutral, or negative sentiment.

Below the Brand Sentiment section is the Sentiment Drivers section. This section identifies Key Themes for your brand and shows you the actual prompt responses associated with that theme.

Use the themes we found in your brand mentions, along with relevant prompt-response snippets, to analyze strengths and weaknesses and prioritize content opportunities.

To determine Key Themes, we look at all your brand mentions across prompt responses and group them into themes. Each theme then receives an individual score based on how positively your brand is discussed. Within the Key Themes table, you’ll see all the themes we identified for your brand as well as the prompt response segments we used to calculate sentiment.

To see the prompt response segments, click the arrow at the far left.

Brand mentions within the examples will be highlighted, and the Sentiment of the example will be indicated with a colored dot on the left-hand side. Green indicates positive sentiment, yellow is neutral, and red is negative.

You can also view the Sentiment drivers of your tricked competitors. Use the drop-down to select a Brand.