Embedded Files
Introduction
Introduction
The Insights section helps you understand how users interact with your AI agents and how well your automation is performing.
It provides visibility into conversations, intents, automation outcomes, user behavior, and overall usage patterns. Insights are designed to help teams continuously improve AI performance using real data.
Plan availability
The Insights section is not included in the Starter plan.
Insights are available on higher plans only.
What Insights Help You Understand
Insights help answer questions such as:
How many conversations are being handled?
What are users asking most often?
Which intents dominate conversations?
How effective are AI Actions like leads, appointments, and tickets?
Where are users located geographically?
Where does the AI need improvement?
Where to Access Insights
Insights are available at two levels:
Organization-level Insights:
Accessible from the main Insights section in the left navigation.Agent-level Insights:
Accessible inside an individual AI agent under Insights.
Both views use the same data but differ in scope.
Organization-Level Insights
Organization-Level Insights
Organization-level insights provide a consolidated view across all AI agents in your organization.
Data Included
Data Included
Organization-level insights aggregate data from:
Website chat
Website voice
Shopify storefront
VoIP (toll-free inbound calls)
Playground testing
All conversations across all channels appear in one place.
Key Metrics Overview
At the top of the dashboard, you’ll see summary metrics:
Total Interactions
Shows the total number of conversations handled within the selected time range.
Average Session Time
Represents the average duration of conversations.
Longer sessions often indicate deeper or more complex interactions.
Average Messages
Shows the average number of messages exchanged per conversation.
Message Count
Displays the total number of AI-generated messages.
Copilot Actions (AI Actions Performance)
The Copilot Actions section shows how effectively your AI is automating operations.
It includes counts for:
Chats answered
Leads captured
Appointments scheduled
Service tickets generated
This helps you measure how much work the AI is handling without human involvement.
Intent Distribution
The Intent chart shows why users are reaching out.
Examples include:
Enquiry
Support
Understanding intent distribution helps you:
Improve training for high-volume intents
Optimize AI Actions triggers
Identify new automation opportunities
Feedback Received
This section shows feedback collected during conversations.
It includes:
Likes
Dislikes
Feedback helps identify response quality issues and areas that need improvement.
Top Keywords
The Top Keywords section highlights the most frequently used terms in conversations.
This helps you:
Identify recurring issues
Discover missing training topics
Improve documents, website training, or text rules
Usage by Country
Usage by Country
The Usage by Country section shows where users are interacting from.
It includes:
Country-wise user count
A visual geographic map
This data is useful for:
Understanding regional adoption
Deciding language and accent for voice agents
Market and expansion analysis
Time Range Filter
You can change the time range (for example, Last Year) from the top-right of the dashboard.
This allows you to:
Track trends over time
Measure impact after changes
Compare performance across periods
Agent-Level Insights
Agent-level insights focus on the performance of a single AI agent.
Where to Access
Where to Access
Go to My Agents
Open an AI agent
Click Insights
What Agent-Level Insights Show
Agent-level insights include:
Conversations handled by the agent
Intent and topic distribution
Fallback frequency
Channel-wise usage
Sentiment trends
This helps teams fine-tune individual agents.
Using Insights to Improve Performance
Insights are most effective when used for action.
You can use them to:
Reduce fallback responses
Improve training coverage
Adjust AI Persona & tone
Refine AI Actions triggers
Decide when to scale or upgrade plans
Common Mistakes to Avoid
Reviewing insights without taking action
Ignoring recurring keywords or intents
Forgetting to check insights after training updates
Assuming insights are available on all plans
Insights are meant for continuous improvement, not passive reporting.
What’s Next?
Once you understand Insights, the next step is optimization.
Recommended next guide:
Page updated
Report abuse