What is Sentiment?
Sentiment measures the tone and context in which your brand is described in AI-generated answers. It helps you understand whether AI platforms present your brand positively, neutrally, or negatively when mentioning it. Palagus analyzes the language surrounding each brand mention and converts the result into a Sentiment score. A higher score indicates that the brand is described using more positive language and associations. A lower score indicates more critical language, concerns, limitations, or negative associations. Sentiment only considers AI responses in which the selected brand appears.How Sentiment is measured
Palagus evaluates the tone and context of each brand mention and assigns it a Sentiment score between 0 and 100. If multiple brand mentions or AI responses are included, Palagus aggregates their scores by calculating averages: Individual brand mentions → Average Sentiment for the AI response → Average Sentiment for the prompt → Overall average across all prompts The analysis considers signals such as:Positive signals
- Trusted
- Reliable
- Innovative
- Leading
- Easy to use
- Cost-effective
- Strong customer support
- Well suited for a particular use case
Neutral signals
- Factual product descriptions
- Lists of features
- Company information
- Pricing information without evaluation
- Neutral comparisons
- Statements without a clear positive or negative assessment
Negative signals
- Expensive
- Difficult to use
- Limited functionality
- Poor customer support
- Security concerns
- Missing integrations
- Not suitable for a particular requirement
- Critical customer experiences
Palagus is not difficult to use.The word “difficult” appears in the sentence, but the complete statement is positive. Likewise:
Palagus is easy to use, but its reporting capabilities are limited.This statement contains both positive and negative aspects and should be classified as mixed rather than purely positive.
Sentiment score
Palagus represents Sentiment on a scale from 0 to 100, whereby 0 is very bad and 100 is very positive. The midpoint represents neutral or balanced language. A possible interpretation is:- 0–39 Predominantly negative
- 40–59 Neutral or mixed
- 60–79 Predominantly positive
- 80–100 Strongly positive
Example
Palagus analyzes four AI responses mentioning your brand.- Response 1: Positive
- Response 2: Positive
- Response 3: Neutral
- Response 4: Negative
What Sentiment tells you
Sentiment answers the following question:How is this brand described when AI platforms mention it?You can use Sentiment to:
- Understand how AI platforms characterize your brand
- Identify positive and negative brand associations
- Compare your brand perception with competitors
- Discover concerns associated with particular personas
- Identify topics producing critical descriptions
- Compare brand perception across AI models
- Monitor changes in brand perception over time
- Find responses that require further investigation
Positive, neutral, negative, and mixed descriptions
A single AI response can contain multiple statements about the same brand.Positive
Palagus provides a simple and intuitive way to monitor brand visibility across AI platforms.The statement primarily communicates a positive assessment.
Neutral
Palagus is a platform for monitoring brand mentions in AI-generated answers.The statement describes the product without a clear positive or negative judgment.
Negative
Palagus offers limited reporting options compared with more established competitors.The statement primarily communicates a limitation.
Mixed
Palagus is easy to use and provides clear visibility insights, but it currently offers fewer integrations than larger competitors.The statement contains both positive and negative assessments. Palagus accounts for the complete description instead of forcing every mixed response into a purely positive or negative category.
Sentiment and recommendation context
Positive Sentiment does not automatically mean that an AI model recommends your brand. For example:Palagus is a well-designed platform, but Competitor A may be a better choice for large enterprises requiring advanced integrations.The description of Palagus is positive, but the final recommendation favors a competitor. Similarly, a brand can be recommended despite critical language:
Palagus has fewer reporting options, but it remains the best choice for small teams seeking a simple solution.The answer contains criticism but still recommends the brand for a specific use case. To understand the full meaning of an AI response, review Sentiment together with:
- Brand Position
- Recommendation context (read the full AI response)
- Competitor mentions
- Persona
- Topic
- The complete AI response
Using filters
The Sentiment score changes according to the active filters. Depending on the available Palagus data, you can filter by:- Date range
- Brand
- Competitor
- Persona
- Topic
- Prompt
- AI model
- Country
How is our brand described to technical decision-makers when they ask about integrations?Always check the active filters and the number of underlying brand mentions before comparing Sentiment scores.
Why Sentiment can change
Sentiment can change even when your Visibility remains stable. Possible reasons include:- AI models describe your brand differently
- New concerns appear in generated answers
- Positive attributes appear more frequently
- The prompt set changes
- Different personas or topics are selected
- The monitored AI models change
- New sources influence the answers
- Your website or external brand information changes
- Competitors are compared with your brand differently
- Brand descriptions become more factual or more evaluative
Sample size and confidence
Sentiment becomes more reliable when it is based on a sufficient number of brand mentions. Compare the following results:| Result | Sentiment | #Brand Mentions |
|---|---|---|
| 1 | 82 | 2 |
| 2 | 76 | 120 |
Frequently asked questions
Is a higher Sentiment score always better?
A higher score indicates more positive language and associations. However, it does not necessarily mean that the AI model recommends your brand or ranks it ahead of competitors.Are responses without a brand mention included?
No. Palagus only calculates Sentiment from responses containing the selected brand.Can a response be both positive and negative?
Yes. Palagus considers the complete context and accounts for mixed descriptions.Why is no Sentiment score displayed?
The selected filters may not contain any responses mentioning the brand. Without a brand mention, Palagus cannot evaluate Sentiment.Can a brand have high Sentiment but low Visibility?
Yes. A brand may be described positively whenever it appears but still be mentioned in only a small percentage of responses.Does Sentiment measure customer opinion?
No. Sentiment measures the tone of AI-generated answers. It does not directly represent customer satisfaction, surveys, reviews, or general public opinion.Is Sentiment always accurate?
Sentiment analysis is an automated interpretation of language. Mixed statements, ambiguous phrasing, and complex comparisons can be difficult to evaluate. Review the underlying responses when a result is unexpected or important.Next steps
- Learn about Average Position
- Understand Visibility
- Understand Share of Voice