Feedback

The Feedback page shows all visitor feedback collected through the Feedback tool, including thumbs up/down ratings on bot responses and detected interest signals.

Filtering & Search

Feedback Types

The Feedback tool collects two types of feedback:

Thumbs Ratings

When enabled, visitors see thumbs up/down buttons on chatbot messages. This provides direct quality ratings for individual responses:

• Thumbs Up — The visitor found the response helpful
• Thumbs Down — The visitor found the response unhelpful or incorrect

Use low-rated responses to identify which answers need improvement in your data sources.

Interest Signals

The AI automatically analyzes conversations and detects visitor interest signals such as:

• Purchase intent
• Feature requests
• Complaints or frustration
• Specific product/service interests

These signals are extracted from the natural conversation flow without requiring explicit visitor input.

Using Feedback to Improve

1. Filter by thumbs down — Focus on negatively rated responses first
2. Review the conversation context — Understand why the response was unhelpful
3. Update data sources — Add or improve content to fix incorrect or incomplete answers
4. Monitor interest signals — Use detected interests for follow-up marketing or product development insights

Best Practices

• Enable thumbs feedback — Direct ratings are the fastest way to find quality issues
• Review feedback weekly — Regular reviews help you improve systematically
• Look for patterns — If multiple visitors rate the same topic negatively, that's a priority fix
• Use interest signals for sales — Detected purchase intent or specific interests can be valuable for sales outreach