nonBot AI

Sentiment Analysis

nonBot AI scores the sentiment of brand mentions across AI responses. This helps you see how AI assistants view your brand and spot risks or chances to improve.

Sentiment Types

Positive - AI responses that portray your brand well. This includes:

  • Praise and endorsements
  • Pointing out your strengths
  • Good comparisons to rivals
  • Praise for features or value

Example: "Brand X is known for great customer support and fair pricing."

Neutral - Factual mentions with no opinion. This includes:

  • Listing your brand next to others
  • Saying what your product does
  • Giving details without judgment

Example: "Brand X offers project management software with task tracking features."

Negative - Bad mentions. This includes:

  • Pointing out weak spots
  • Warnings or cautions
  • Bad comparisons
  • Mentions of problems or complaints

Example: "Brand X has been criticized for its steep learning curve."

Mixed - Mentions with both good and bad parts. This includes:

  • Balanced pros-and-cons reviews
  • Conditional picks
  • "It depends" responses

Example: "Brand X offers strong features but may be too complex for beginners."

Viewing Sentiment Data

The dominant sentiment KPI card shows your most common sentiment type at a glance.

The detailed view breaks down the share of each of the four types.

In the Results tab, each response shows its sentiment label. You can see how AI described your brand, read specific statements, and grasp the context.

Reading Sentiment

Most brands should aim for mostly positive or neutral mentions with few negative ones. Some mixed sentiment is normal. It can even look more genuine than all-positive responses.

Warning signs to watch for:

  • Negative sentiment growing over time
  • The same negative mentions across many AI providers
  • Negative sentiment focused on certain topics

Good signs:

  • Growing positive sentiment
  • Positive mentions in head-to-head comparisons
  • Picks in key use cases

Reacting to Sentiment

If you see negative sentiment:

  1. Review the specific AI responses in the Results tab
  2. Spot the concerns being raised
  3. Decide if the critique is valid
  4. Take action: update your website content, fix cited product areas, or create content to correct wrong claims

If you see positive sentiment:

  • Note which strengths AI brings up
  • Back up those messages in your content
  • Make sure the claims are correct and current
  • Build on areas where you're seen in a good light

If you see mixed sentiment:

Mixed results often mean your product works well for some uses but not others. There may be a known trade-off. Respond by:

  • Stressing ideal use cases in your content
  • Facing limits head-on
  • Giving resources for common hurdles

Sentiment Trends

Track changes over time. Rising sentiment may mean your tuning is working. Falling sentiment may point to new issues. Some industries see seasonal shifts too.

Limits

Sentiment scoring is not perfect. Keep in mind:

  • AI providers phrase things in different ways
  • Context shapes how text is read
  • Sarcasm or subtle tone may be tagged wrong
  • Small sample sizes cut into accuracy

Use sentiment as one signal among many. Don't treat it as your only measure of brand health.

Next Steps

Compare your sentiment to rivals in Competitor Detection, create content to shift views with Content Ideas, or review tips in Best Practices.

Previous
Trend Charts