> ## Documentation Index
> Fetch the complete documentation index at: https://faii.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Understanding Sentiment Scores

> How FAII measures whether AI platforms speak positively, neutrally, or negatively about your brand - and what drives sentiment changes.

<Info>
  **Quick Answer:** Sentiment is scored from -100 (strongly negative) to +100 (strongly positive). It measures the tone and recommendation strength of AI mentions - not just whether you're mentioned, but how enthusiastically.
</Info>

## The Sentiment Scale

|    Range    |      Rating     | Example Language                                                |
| :---------: | :-------------: | --------------------------------------------------------------- |
| +70 to +100 | Strong Positive | "highly recommend," "excellent choice," "industry leader"       |
|  +30 to +69 |     Positive    | "good option," "worth considering," "solid reputation"          |
|  -29 to +29 |     Neutral     | "is available," "one option," "has been around since"           |
|  -69 to -30 |     Negative    | "has some issues," "may want to research more," "not ideal for" |
| -100 to -70 | Strong Negative | "avoid," "not recommended," "significant problems"              |

## What Drives Sentiment

### Recommendation strength

The strongest sentiment signal is how forcefully AI recommends you:

| Language                             | Sentiment Impact |
| ------------------------------------ | :--------------: |
| "I'd definitely recommend..."        |        +90       |
| "They're an excellent choice for..." |        +80       |
| "A strong option if you need..."     |        +60       |
| "Worth considering alongside..."     |        +40       |
| "One option you could look at..."    |        +10       |
| "Has some limitations including..."  |        -40       |
| "You might want to avoid..."         |        -80       |

### Comparative context

How AI positions you relative to alternatives:

* **"Better than most"** → Strong positive
* **"Similar to \[competitor]"** → Neutral
* **"Not as strong as \[competitor]"** → Negative

### Qualifier words

Subtle language cues affect sentiment:

| Qualifiers                            |       Direction      |
| ------------------------------------- | :------------------: |
| "definitely," "absolutely," "clearly" | Strengthens positive |
| "might," "could," "perhaps"           |   Weakens positive   |
| "however," "but," "although"          |  Introduces negative |
| "despite," "unfortunately"            |   Negative context   |

### Context of mention

Where your brand appears in the response structure:

* **Featured solution** (discussed in detail) → Higher sentiment
* **One of several options** (listed briefly) → Moderate sentiment
* **Mentioned with caveats** (pros AND cons) → Mixed sentiment
* **Afterthought** ("you could also try...") → Low positive

## Sentiment on the Dashboard

### Overall sentiment score

Your aggregate sentiment across all mentions, all platforms, all keywords. This is the number that feeds into your AI Authority Rank.

### Sentiment distribution

A breakdown showing what percentage of your mentions are:

* Strong positive
* Positive
* Neutral
* Negative
* Strong negative

### Sentiment by platform

Each AI platform may have different sentiment for your brand. Common patterns:

* Claude tends to be more moderate (fewer extremes)
* ChatGPT tends to be more positive when mentioning
* Grok reflects social sentiment (more volatile)
* Perplexity is factual (more neutral)

### Sentiment trends

How your sentiment changes over time. Watch for:

* Gradual improvement (content strategy working)
* Sudden drops (potential reputation issue)
* Platform-specific changes (one platform shifting)

## Improving Sentiment

### Content strategies

* Publish case studies with measurable results
* Create comparison content that honestly positions you
* Address known weaknesses transparently
* Highlight awards, certifications, and recognition

### Reputation management

* Monitor and respond to negative reviews
* Build a strong portfolio of positive testimonials
* Engage with industry communities
* Publish thought leadership content

### What doesn't work

* Overly promotional content (AI sees through it)
* Fake reviews or testimonials
* Ignoring legitimate criticism
* Keyword stuffing brand mentions

<Note>
  Sentiment reflects what the web says about you. You can't game it directly - you need to genuinely improve your reputation and make that improvement visible to AI systems through quality content and third-party validation.
</Note>

## FAQ

<AccordionGroup>
  <Accordion title="Can sentiment be negative even if I'm mentioned?">
    Yes. Being mentioned with negative context ("has some quality issues" or "not the best choice for enterprise") scores negatively. It's possible to have high mention rate but poor sentiment.
  </Accordion>

  <Accordion title="Is neutral sentiment bad?">
    Not necessarily. Neutral means you're mentioned without strong opinion. It's better than negative, but positive is obviously preferred. Neutral often means AI lacks enough signal to form a strong opinion about your brand.
  </Accordion>

  <Accordion title="Why does sentiment vary between platforms?">
    Each platform has different training data, different response patterns, and different tendencies. Grok may reflect recent social controversy while Claude remains neutral because it's less influenced by social media.
  </Accordion>
</AccordionGroup>

## Related Articles

<CardGroup cols={2}>
  <Card title="Position Tracking" icon="list-ordered" href="/docs/chat-intelligence/position-tracking">
    Position + sentiment = complete picture
  </Card>

  <Card title="Improving Your Score" icon="arrow-up" href="/docs/ai-authority-rank/improving-your-score">
    Strategies to improve all components
  </Card>
</CardGroup>