Analyze AI visibility
Sentiment
The tone of model language when your brand is mentioned-in answers scoped to your prompts, date range, and filters.
View Markdown · AI index · Full documentation for AI
Overview
Sentiment measures the overall tone of AI-generated answers when they mention your brand, on a scale from 0 to 100. It answers whether models frame you favorably, neutrally, critically, or with mixed signals-not only whether you appear.
Scores reflect your monitored prompts and respect the date range and filters you apply (for example time period and provider). Use sentiment alongside visibility: a brand can be visible often yet described in lukewarm or negative terms when mentioned.
A score near 50 means positive and negative classifications in scope balance out; higher values tilt toward relatively more positive language. Read trends and comparisons rather than a single snapshot - see Welcome to igeo.ai.
How sentiment is measured
igeo.ai classifies how your brand is portrayed when it appears in an answer (or aggregated slice of answers), using discrete labels:
Positive - Language that favors trust, strength, or recommendation when your brand is discussed.
Negative - Critical, cautionary, or unfavorable framing.
Neutral - Factual or descriptive tone without a clear positive or negative lean.
Mixed - Both supportive and critical signals in the same context; for scoring, this is split between positive and negative (see below).
How sentiment is calculated
The headline score summarizes classified mentions in your scope. Conceptually, each mixed label contributes half toward positive and half toward negative when computing the balance; neutral increases the denominator and tends to pull the score toward the middle. In closed form:
Sentiment (0–100) ≈ 50 + ((adjusted_positive − adjusted_negative) ÷ total) × 50
Here adjusted_positive and adjusted_negative incorporate that split for mixed labels; total is the count of classified items in the aggregation. Exact rules can vary by view and product version - use direction and relative changes over time. For variability and interpretation, see Welcome to igeo.ai.
Reading sentiment in igeo.ai
Open Sentiment from the sidebar (see Sidebar navigation) for breakdowns tied to your workspace. Use the same time range and engine / provider filters as on the Dashboard so period comparisons stay fair.
Individual answers will vary; rely on smoothed trends in your chosen window rather than one response.
Brand reputation
Tone shapes whether buyers trust you when models cite you. Persistent negative or mixed sentiment on core prompts is a signal to revisit messaging, sources, and factual gaps-not only mention volume.
Competitive comparison
On the same prompts, compare how you and tracked competitors are described. Pair with Share of Voice and Position to see whether you win mentions but lose the narrative.
Strategic insights
Track sentiment over time to spot drift after launches or news, prioritize topics where tone slips, and align improvements with Sources and Recommendations.
Visibility
How often your brand appears in AI answers to your tracked prompts-not relative share.
Position
How average rank among brands works in the same prompts and slice as sentiment.