---
title: Sentiment
description: Understand how AI answers describe your brand using the 0–100 score
  and the answer text.
canonical_url: https://orchestor.io/docs/en/visibility/sentiment
markdown_url: https://orchestor.io/docs/en/visibility/sentiment.md
contentType: explanation
---

# Sentiment

Sentiment indicates how positively or negatively an AI answer describes your brand. Together with [Visibility](https://orchestor.io/docs/visibility/visibility.md), which measures how often the brand appears, it helps you check both whether your name comes up and how it is described.

## Reading the score

Orchestor displays sentiment on a **0–100** scale.

| Score | Reference point |
| --- | --- |
| 0 | Very negative |
| 50 | Neutral |
| 100 | Very positive |

This is not the percentage of positive answers. For example, 75 does not mean “75% of answers are positive.” It is an average assessment of the descriptions' tone, converted for display.

## Calculation

AI analysis assigns a continuous score from −1 to 1 to descriptions of the brand. A score of −1 is very negative, 0 is neutral, and 1 is very positive. Valid sentiment scores are averaged, then converted to the interface's 0–100 scale.

```text
Displayed sentiment score = (Average sentiment score + 1) ÷ 2 × 100
```

For example, scores of −0.5, 0, and 0.5 average to 0, displayed as **50**. An average of 0.5 is displayed as **75**.

Answers without an available sentiment score are excluded from the average. If there are no valid scores, the interface displays `—`. An unscored answer is not treated as a neutral 50.

## Look beyond the average

The same score of 50 can mean all descriptions are neutral, or that strong positive and negative descriptions coexist. When the aggregate changes, check answer counts, mention counts, and the answer text.

An assessment of the brand overall may also differ from assessments of specific features such as price or ease of use. Reading which features are discussed and the wording used helps identify descriptions to improve.

## What to check next

1. Compare scores and mention counts using the same prompts, models, and period conditions.
2. Open [Answers](https://orchestor.io/docs/visibility/answers.md) for the scope that changed and read the wording about your brand.
3. Check the features being evaluated, such as price, functionality, or reliability, and the cited sources.
4. If you find incorrect information or outdated descriptions, consider updating the relevant primary information.

Sentiment is an AI assessment of text. For irony, conditional recommendations, and comparisons of multiple brands, review the text alongside the score. It is not a customer survey or a measure of actual customer satisfaction.

## How API values correspond

The API's `sentiment_score` ranges from −1 to 1. Use the conversion above when comparing it with the displayed number. An answer's sentiment label classifies its description and is distinct from the aggregate score.

- [Main metrics](https://orchestor.io/docs/visibility/metrics-overview.md)
- [Share of voice](https://orchestor.io/docs/visibility/share-of-voice.md)
- [Average position](https://orchestor.io/docs/visibility/position.md)

---

[Documentation index](https://orchestor.io/docs/llms.txt)