Skip to main content

​
Definition of sentiment

Sentiment measures the qualitative tone of an AI model’s response when it mentions your brand. Unlike visibility, which tracks volume, sentiment tracks perception. KIME uses AI-driven classification to categorize every mention as positive, neutral, or negative, providing a clear picture of your brand’s reputation within the LLM’s output.

​
How sentiment is classified

The sentiment score is calculated by evaluating the language, adjectives, and context used in each brand mention.
  1. Positive: The AI recommends your brand, highlights specific benefits, or uses favorable descriptors.
  2. Neutral: The AI mentions your brand in a factual or list-based format without providing a qualitative opinion.
  3. Negative: The AI mentions your brand in the context of complaints, limitations, or poor reviews.

​
The Sentiment Score (0–100)

Your Overall Sentiment Score is a normalized index ranging from 0 to 100. A higher score indicates a predominantly positive presence. This score allows you to track whether your brand’s reputation is improving or declining over time across different AI models and regions.

​
Strategic value of sentiment data

Monitoring sentiment is essential for reputation management. If you notice a high volume of mentions (high visibility) but a low sentiment score, it indicates that AI models are frequently discussing your brand in a negative light. This often happens if the AI is training on outdated reviews or negative news cycles. Use this data to identify which specific sources are driving negative sentiment so you can address them through your broader communication strategy.