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

# Analytics: Measure AI Agent and Customer Satisfaction

> Track CSAT scores, analyze message-level satisfaction, and monitor ZynfoAI agent performance to identify gaps and optimize your knowledge base content.

# Analytics and Satisfaction

Measuring the quality and accuracy of your AI agent's responses is critical for continuously improving your business knowledge base. ZynfoAI provides built-in analytics to track user satisfaction effortlessly.

At the top of the **Chat Sessions** page, you will find real-time global statistic cards:

## CSAT (Customer Satisfaction)

When a chat session ends, users can rate their overall support experience on a 5-star scale. The CSAT metric shows the aggregate percentage of total sessions that received a positive rating of **4 stars or higher**.

## Message Satisfaction

Beyond the overall session rating, users can give a direct **Thumbs Up** or **Thumbs Down** to individual AI agent responses during the chat flow. This granular metric shows the percentage of specific messages that received positive feedback.

> **Optimization Best Practice:** Routinely filter your chat sessions to find conversations with low CSAT scores. Review the transcript to see exactly which questions the AI agent struggled to answer, and update your Knowledge Base documentation accordingly to prevent future failures.

## Related Topics

* [Inbox and Session Management: Track AI Agent Conversations](/inbox/session-management)
* [Human Handoff: Seamless Escalation from AI Agents](/inbox/human-handoff)
* [Exporting Data: Download AI Agent Chat Transcripts](/inbox/exporting-data)