Bot Analytics provides pre-built dashboards that offer immediate insights into your bot's performance, user engagement, and interaction patterns. These default metrics are organized into five dashboards.

Overview Dashboard

The Overview dashboard provides a quick snapshot of your bot's key performance indicators.

Active Users (Gauge)

Shows total active users compared to the total user base.

Metric	Description
Active Users	Users who interacted with or saw bot messages in the selected period
Total Users	Total users with access to the bot

Toggle: Include terminated users - Include/exclude users who have left the organization

Drilldown: Active Users Table → User's Sessions → Session Interactions

Most Used Policy

Displays the HR policy with the highest number of user queries.

Drilldown: Policy Usage Table → Policy Wise Users

Most Used Feature

Shows the most clicked quick reply, webview, or URL button.

Drilldown: Feature Usage Table → User Wise Feature Details

User Feedback (Pie Chart)

Distribution of thumbs up vs thumbs down feedback from users.

How it's tracked: Counts explicit feedback button clicks on bot responses.

Drilldown: Thumbs Up/Down Table → Feedback Request Details

Handled Requests (Gauge)

Ratio of successfully handled requests to total typed requests.

Metric	Description
Handled Requests	Queries successfully answered by bot or AI agent
Typed Requests	All plain text messages from users

Drilldown: Handled/Not Handled Table → Request Details

Total Session Count

Total number of user sessions in the selected period.

Drilldown: Active Users Table → User's Sessions

Average Session Time

Average duration of user sessions, shown as "Xm Ys" format.

How it's calculated:

Sum of all session durations ÷ Number of sessions
Sessions with zero duration (single message) are excluded

Drilldown: Average Session Time Table → Channel Wise Average Session Time

Average Interactions Per Session

Mean number of interactions per session.

How it's calculated: Total interactions (excluding "seen") ÷ Total sessions

Drilldown: Average Interactions Per Session Table

User Metrics Dashboard

Detailed insights into user acquisition, engagement, and distribution patterns.

Active Users (Gauge)

Same as Overview dashboard - shows active vs total users.

Key User Metrics (Stacked Bar)

Metric	How it's calculated
New Users	Users whose first ever message to the bot falls within the selected period
Active Users	Users with any interaction or seen message in the period
Repeating Users	Users with 2 or more sessions in the period
Logged-In Users	Users who have logged into the bot application

Drilldown paths:

New Users → New Users Table → User's Sessions
Active Users → Active Users Table → User's Sessions
Repeating Users → Repeating Users Table → User's Sessions
Logged-In Users → Logged-In Users Table → Sessions/Logins

Users Over Time (Area Chart)

Trend visualization of user metrics over time.

Toggle Options:

Daily/Monthly - Switch granularity
Include terminated users - Include/exclude terminated users

User Distribution (Stacked Bar)

Distribution of active users by channel or demographics.

View Option	Description
Channel Wise	Distribution by communication channel
Demographic Wise	Distribution by custom demographic fields

Drilldown: Channel Wise Active Users → User's Sessions

Bot Performance Metrics Dashboard

Focuses on bot effectiveness, response quality, and interaction patterns.

User Feedback (Pie Chart)

Same as Overview dashboard - distribution of thumbs up/down feedback.

Handled Requests (Gauge)

Same as Overview dashboard - handled vs typed requests ratio.

Bot Message Count

Total messages sent by the bot to users.

Drilldown: Bot Messages Table → Session Wise Interactions

Total Interactions

Total user interactions (typed queries + button clicks).

Note: "Seen" messages are excluded from this count.

Drilldown: Total Interactions Table → Session Wise Interactions

Total Requests

Combined count of Bot Flow and Autonomous Agent requests.

How it's calculated:

Count of interactions without AI agent involvement (Bot Flow)
Plus unique AI agent request IDs (Autonomous Agents)

Drilldown: Total Interactions Table

First Response Time

Average time between user message and bot's first response.

How it's calculated:

For each message: Bot response timestamp − User message timestamp
Averaged across all interactions, displayed in seconds

Drilldown: Channel Wise First Response Time → Request Details → Interaction Details

Channel Wise Interactions (Bar Chart)

Distribution of interactions across communication channels.

Drilldown: Channel Interactions Table

Peak Usage Time

Hour of day with maximum interactions.

How it's calculated:

Group all interactions by hour (0-23)
Find hour with highest count
Display in 12-hour format with timezone adjustment

Drilldown: Usage Hour Table

Requests Over Time (Area Chart)

Trend of handled vs not handled requests.

Session Metrics Dashboard

Detailed insights into user session behavior and patterns.

Effective Sessions (Gauge)

Ratio of effective sessions to total sessions.

Metric	Description
Effective Sessions	Sessions where at least one message was successfully handled
Abandoned Sessions	Sessions where no message was successfully handled

How it's calculated: A session becomes "effective" the moment any interaction within it gets handled. The status is updated in real-time as messages are processed. "Seen" messages don't affect effectiveness.

Drilldown paths:

Effective Sessions → Effective Sessions Table → User Wise Effective Sessions
Abandoned Sessions → Abandoned Sessions Table

Average Session Time

Same as Overview dashboard - average duration of sessions.

Drilldown: Average Session Time Table → Channel Wise Average Session Time

Average Time Between Sessions

Average interval between consecutive sessions by the same user.

How it's calculated:

For each user, list all sessions chronologically
Calculate gap between consecutive session start times
Average all gaps (excluding zero-duration gaps)

Drilldown: Average Time Between Sessions Table → Details

Average Interactions Per Session

Same as Overview dashboard - mean interactions per session.

Drilldown: Average Interactions Per Session Table

Sessions Over Time (Stacked Column Chart)

Trend of effective vs abandoned sessions over time.

Channel-Wise Sessions (Stacked Bar)

Distribution of sessions by channel, showing effective vs abandoned breakdown.

Drilldown: Channel Wise Session Table → Channel User Wise Session Table

Channel-Wise Average Session Time (Bar Chart)

Average session duration by channel.

Drilldown: Channel Wise Average Session Time Table

Sessions & Interactions Over Time (Area Chart)

Comparison of session count vs interaction count trends.

Queries & Interactions Dashboard

Analysis of user queries, feature usage, and interaction patterns.

Most Used Policy

Same as Overview dashboard - policy with highest user interactions.

Drilldown: Policy Usage Table → Policy Wise Users

Most Used Feature

Same as Overview dashboard - most clicked feature.

Drilldown: Feature Usage Table → User Wise Feature Details

Most Used Skill

Most frequently invoked skill by the autonomous agent.

Drilldown: Tool Call Table → User Tool Call Table → Tool Call Details

Total Button Clicks

Total button click interactions.

Included message types:

Button, Quick Reply, Webview, Webview Open
URL Click, Menu Button, FAQ Response, Related Intent

Survey responses are excluded.

Drilldown: Button Clicks Table → User Wise Button Clicks → Button Click Details

Queries Handled by Smart Answers

Queries answered using knowledge base/helpdesk articles.

How it's tracked: Counts interactions where NLU response category is "Smart Answer"

Drilldown: Smart Answers Table → Smart Answer Details

Usage (Trend Line Chart)

Usage trends for features, policies, and skills over time.

View	Description
Features	Top features and usage over time
Policies	Top policies and access over time
Skills	Top AI agent skills and invocation over time

Drilldown paths:

Feature → Feature Wise Table → User Wise Feature Details
Policy → Policy Details Table → Policy Wise Users
Skill → Tool Call Table → User Tool Call Table

Interactions Type (Bar Chart)

Distribution of user message types.

Drilldown: Message Type User Table → Message Type User Details

Intents (Bar Chart)

Distribution of user queries by detected intent/category.

Drilldown: Intent Details Table → Intent Wise User Table

Voice Calls Dashboard

The Voice Calls dashboard provides analytics for voice-based bot interactions. This dashboard is only visible when voice calling is enabled for your bot.

Voice Calls Completed (Gauge)

Shows completed voice calls compared to total connected calls.

Metric	Description
Completed Calls	Voice calls where user sent at least one message
Connected Calls	Total calls that were successfully connected

How it's calculated:

A call is connected when it has a successful connection event OR a disconnection event
A call is completed when the user sends at least one message during the call

Drilldown tabs:

Completed → Completed Calls Table → Call Details
Non-completed → Non-Completed Calls Table

Voice Call Users (Gauge)

Shows users who made voice calls compared to total active users.

Metric	Description
Voice Call Users	Users who completed at least one voice call
Total Active Users	All active users in the selected period

Drilldown tabs:

Voice call users → Voice Call Users Table
Non-voice call users → Non-Voice Call Users Table

Effective Voice Calls (Gauge)

Shows effective voice calls compared to total completed calls.

Metric	Description
Effective Calls	Completed calls with at least one handled query
Completed Calls	All calls with user messages

How it's calculated: A call becomes "effective" when the bot successfully handles at least one user query during the call.

Drilldown tabs:

Effective → Effective Calls Table
Non-Effective → Non-Effective Calls Table

Total Call Duration

Sum of all call durations in the selected period.

How it's calculated:

If both connected and disconnected timestamps exist: Disconnected time − Connected time
If only disconnected exists: Disconnected time − First message time
If only connected exists: Last message time − Connected time

Drilldown: Call Duration by User Table

Voice Calls Connected

Percentage of calls successfully connected after initialization.

How it's calculated: (Calls with successful connection ÷ Total initialized calls) × 100

Drilldown: Connection Rate Table

Average Voice Call Duration

Average duration of voice calls.

How it's calculated: Total call duration ÷ Number of calls with duration > 0

Drilldown: Call Duration by User Table

Average User Messages Per Voice Call

Average number of messages users send per completed call.

How it's calculated: Total user messages in completed calls ÷ Number of completed calls

Drilldown: User Messages Per Call Table

Voice Calls Over Time (Line Chart)

Trend showing connected and completed voice calls over time.

Lines displayed:

Connected calls - Calls that were successfully connected
Completed calls - Calls where user sent at least one message

Drilldown tabs:

Completed → Completed Calls Table
Non-completed → Non-Completed Calls Table

Key Concepts

What is a Session?

A session represents a continuous conversation between a user and the bot.

How sessions are calculated:

A session starts when a user sends their first message to the bot
The session continues as long as the user keeps interacting within 30 minutes of their last message
If there's more than 30 minutes of inactivity (no user message), the current session ends
Any new message after this 30-minute gap starts a new session

Session timing:

Start Time: Timestamp of the first message in the session
End Time: Timestamp of the last message in the session
Session Duration: Time difference between end time and start time

Example: If a user messages at 10:00 AM, then again at 10:15 AM, and finally at 10:45 AM, all three messages are part of one session. But if their next message is at 11:30 AM (45 minutes later), that starts a new session.

What is Request Status?

Request Status indicates the outcome of a user's request. The values differ based on whether the request was processed by Bot Flow or Autonomous Agents.

For Bot Flow Requests

Status	Description
Handled	Bot successfully understood and responded to the query
Not handled	Bot couldn't process the request (fallback response triggered)

For Autonomous Agents Requests

Status	Description
Handled Successfully	AI agent completed the task successfully
Handled Unsuccessfully	AI agent attempted but couldn't complete the task successfully
Handled	Request was processed (used when no specific sub-status applies)
Unhandled	AI agent couldn't handle the request
Failed	AI agent encountered an error during processing
User Aborted	User cancelled the AI agent interaction
User Abandoned	User left the conversation before the AI agent completed

What is Request Type?

Request Type indicates which system processed the user's query:

Request Type	Description
Bot Flow	Query handled by standard bot conversation flows (pre-configured responses)
Autonomous Agents	Query processed by AI-powered autonomous agent (can perform complex tasks)

What is Content Details?

Content Details shows what the bot identified or used to respond to the user's query. The value depends on how the request was processed:

Scenario	Content Details Shows
Query matched a knowledge article	The article name(s) that were used to respond
Query triggered a skill/tool	The skill or tool name(s) that were invoked
Query couldn't be matched	The actual user message text

Examples:

"Leave Policy, Attendance Policy" - Bot found relevant HR policies
"create_ticket, get_ticket_status" - Autonomous agent used these skills
"how do I reset my password" - User's original message (when no specific match)

What is Interaction Type?

Interaction Type describes how the user sent their message:

Interaction Type	Description
Plain Text	User typed a free-form message
Button	User clicked a button in a card or message
Quick Reply	User selected a quick reply option
Webview	User interacted through a webview form
URL Click	User clicked a URL/link
Menu Button	User clicked a menu button
FAQ Response	User selected from FAQ suggestions
Postback	User clicked a button that triggers an action

Handled vs Not Handled

Handled: The bot successfully understood and responded to the user's query
Not Handled: The bot couldn't process the request - typically results in a fallback response or escalation

Active Users vs Interactions

Active Users includes anyone who:

Sent a message to the bot, OR
Clicked a button, OR
Simply saw/viewed a bot message

Interactions counts only active engagements:

Typed messages
Button clicks
Quick replies
Form submissions

Note: "Seen" messages (where user only viewed but didn't respond) are included in Active Users but excluded from Interaction counts.

Available Channels

Channel	Description
MS Teams	Microsoft Teams
Web	Web chat interface
Slack	Slack workspace
WhatsApp	WhatsApp
Workplace	Meta Workplace
Google Chat	Google Chat

Voice Call Terms

Term	Description
Connected Call	A call that has a successful connection event or disconnection event
Completed Call	A connected call where the user sent at least one message
Effective Call	A completed call where at least one user query was successfully handled
Call Duration	Time between call connection and disconnection

Understanding Drilldown Tables

When you click on a metric, you can drill down into the data for more details. Here's what the common columns mean:

User Table Columns

Column	Description
User ID	User's display name or identifier
Employee ID	Employee ID from HR system
Is Terminated	Whether user has left the organization (Yes/No)
Session Count	Number of sessions by this user in the selected period
First Active Time	When user first accessed the bot

Session Table Columns

Column	Description
Session ID	Unique identifier for the session
Channel	Communication channel (MS Teams, Web, etc.)
Start Time	When session began (first message)
End Time	When session ended (last message)
Interaction Count	Number of interactions in this session
Is Effective	Whether at least one message was handled

Request/Interaction Table Columns

Column	Description
Content Details	What was matched - article names, skill names, or user message
Request Type	Bot Flow or Autonomous Agents
Request Status	Outcome of the request (Handled, Not handled, Failed, etc.)
Interaction Type	How user sent the message (Plain Text, Button, etc.)
Request Count	Number of requests with this content
Interaction Count	Total interactions for this content
Timestamp	When the interaction occurred
User Message	The actual message sent by user
Bot Response	The bot's reply

Quick Reference

Session Rules

New session starts: After 30+ minutes of inactivity
Session duration: Last message time − First message time
Effective session: At least one handled message

What's Counted Where

Metric Type	Includes "Seen" Messages?
Active Users	✅ Yes
Interactions	❌ No
Session Count	✅ Yes (session still counts)
Handled Requests	❌ No (only typed messages)

Request Type & Status

Request Type	Processed By
Bot Flow	Standard bot conversation flows
Autonomous Agents	AI-powered autonomous agent

Request Status (Bot Flow)	Meaning
Handled	Successfully responded
Not handled	Fallback triggered

Request Status (Autonomous Agents)	Meaning
Handled Successfully	Task completed successfully
Handled Unsuccessfully	Attempted but failed
Unhandled	Couldn't process
Failed	Error occurred
User Aborted	User cancelled
User Abandoned	User left mid-conversation