Skip to main content

How to view your AI agent's performance

Updated over 3 weeks ago

Summary

Astra provides built-in analytics that help you understand how your AI agent is performing across your website and messaging channels such as WhatsApp. These analytics show how effectively the agent engages visitors, resolves conversations, captures leads, and supports your business outside regular working hours.

This guide explains the key metrics available in Astra agent analytics and how to interpret them to measure your agent’s performance and identify opportunities for improvement.

Instructions

Where to find Astra agent analytics

You can access your agent performance metrics directly from the Agent Analytics in Astra. These reports provide insights into conversation handling, resolution rates, lead generation, channel performance, and more.

To view your AI agent’s performance:

  • Log in to your Astra account.

  • Go to Agents.

  • Select the agent whose performance you want to review.

  • Open the Analytics section for that agent to view its performance metrics and reports.

Astra’s impact on your business

The Astra Impact section provides a quick snapshot of how your AI agent is contributing to your business within a selected time period.

Key metrics include

  • Conversations handled
    Total number of interactions handled by the agent across all supported channels, including both text and voice.

  • Time saved
    The estimated amount of time saved by automating conversations that would otherwise require human agents.

  • Resolution rate
    The percentage of conversations that the AI agent successfully resolved without human assistance.

  • Languages handled
    The number of languages in which the agent successfully interacted with users.

Overview report

The Overview Report provides a detailed summary of how conversations are handled by your agent.

Metrics available

  • Total conversations
    Total number of interactions handled by the agent across all channels during the selected time period.

  • Resolution rate
    Percentage of conversations successfully resolved by the agent without escalation to a human agent.

  • Average handle time
    Average time taken to resolve a conversation from start to finish.

  • Drop-off rate
    Percentage of users who leave the conversation before receiving a helpful answer or after a negative interaction.

Conversation funnel

The Conversation Funnel shows how conversations move through different stages:

  • Total conversations

  • Resolved

  • Escalated

  • Dropped

This helps you understand how effectively your agent handles queries before they require human support.

Conversation volume trend

This report shows how conversation volume changes over time and provides a breakdown by text and voice channels. It helps identify usage patterns and traffic growth.

Top FAQ categories

This section highlights the most frequently asked topics based on customer queries. It helps you identify the areas where users most often need assistance.

Knowledge gaps

Knowledge gaps highlight topics where the AI agent struggles to provide satisfactory answers. These insights help you improve your knowledge base by identifying missing or unclear information.

Escalation reasons

This report categorizes the reasons why conversations were escalated to human agents. Understanding these reasons helps you improve automation and reduce escalation rates.

Channels report

The Channels Report compares performance metrics across different communication channels such as text and voice.

Metrics available

  • Total conversations
    Total number of interactions handled by the agent across all channels.

  • Total duration
    Combined duration of conversations across selected channels.

  • Average time per session
    Average duration of a text-based chat session from start to end.

  • Average turns
    Average number of back-and-forth exchanges between the user and the AI agent before resolution.

  • Resolutions
    Percentage of conversations successfully resolved without human intervention.

  • Containment rate
    Percentage of conversations fully handled by the AI agent without requiring escalation.

You can also view a comparison table that shows channel performance side-by-side.

Availability report

The Availability Report shows how your AI agent supports users across different times, languages, and regions.

Metrics available

  • After-hours coverage
    Number of conversations handled outside defined business hours (evenings, nights, and early mornings).

  • Weekend coverage
    Number of conversations handled on Saturdays and Sundays.

  • Languages served
    Number of languages in which the agent successfully handled conversations.

Hourly query volume and resolution

This chart shows the distribution of conversations by hour of the day. It helps you identify peak traffic hours and see how much support your agent provides outside business hours.

Language distribution

This report shows the percentage of conversations by language, helping you understand which languages your customers prefer when interacting with your agent.

Language performance

This section breaks down resolution rates and satisfaction metrics by language. It helps identify languages where the agent may require additional training or improved knowledge base content.

Leads report

The Leads Report helps you track how effectively your AI agent converts conversations into potential sales opportunities.

Metrics available

  • Visitors
    Total number of potential leads who started a conversation with the agent.

  • Leads engaged
    Number of leads who interacted meaningfully with the agent and provided their contact information.

  • Qualified leads
    Number of captured leads that met your predefined qualification criteria (such as budget, timeline, or authority).

  • Meetings booked
    Number of demos, calls, or appointments successfully scheduled by the agent.

  • After-hours leads
    Leads captured outside business hours that might otherwise have been missed.

Lead funnel

The Lead Funnel shows how potential customers move through each stage of the lead journey, from initial conversation to qualification.

Lead sources

This section shows where your leads are coming from, broken down by channel or platform. It helps you understand which channels are driving the most valuable conversations.

Note: You may also see a Summary section within the charts in these reports. This is an AI-generated overview that highlights key insights and trends based on the data shown in the reports.

Tip: You can also view the overall performance of all your AI agents in one place. Just head to the Dashboard page to see the total number of visitors and leads generated across all your sales agents.

In addition to seeing insights on visitors and leads, the Dashboard also shows a breakdown of your agents by status - so you can easily see how many are live and how many are still in draft.

By regularly reviewing these analytics, you can continuously improve your AI agent’s performance, identify knowledge gaps, and optimize how your business engages and converts customers.

Did this answer your question?