Skip to main content

AI Employee Analytics

Table of Contents

  1. Accessing Employee Analytics
  2. Key Performance Metrics
  3. Daily Activity Analysis
  4. Status Distribution Charts
  5. Hourly Distribution Patterns
  6. Analysis Tab and Conversation Review
  7. Filtering and Searching Conversations
  8. Evaluation Tracking
  9. Using Analytics for Optimization
  10. Reporting and Insights

Accessing Employee Analytics

Each AI employee has a dedicated analytics view providing detailed insights into their performance.
  1. Go to AI Employees from the main navigation
  2. Click on the AI employee name you want to analyze
  3. The default view shows the Overview tab with summary analytics
  4. Additional depth is available in the Analysis tab

Analytics Overview

The analytics page presents:
  • Key metric cards with current performance data
  • Visual charts showing trends and distributions
  • Comparative data to understand patterns

Date Range

By default, analytics show data from a recent period (typically the last 30 days). This provides meaningful trend data while keeping focus on current performance.

Key Performance Metrics

The top of the analytics page displays essential metrics in card format.

Total Calls

The total number of conversations this AI employee has handled during the analysis period. What this tells you:
  • Overall activity level
  • Workload compared to other employees
  • Capacity utilization
Context matters: A new AI employee will have lower total calls than an established one. Compare against similar timeframes and use cases.

Inbound vs Outbound Breakdown

Understanding the direction of calls provides important context: Inbound Calls: Conversations initiated by contacts calling in. High inbound volume indicates:
  • The AI employee is serving as a primary contact point
  • External awareness of the phone number
  • Reactive communication patterns
Outbound Calls: Conversations initiated by the AI employee calling contacts. High outbound volume indicates:
  • Active campaign involvement
  • Proactive outreach strategies
  • Lead generation or follow-up activities

Success Rate

The percentage of calls that achieved a positive outcome. Calculation: (Successful calls + Voicemail left) / Total calls x 100 Interpreting success rates:
  • 80%+ for inbound is typically strong
  • 20-40% for cold outbound is common
  • Compare against your own historical benchmarks

Average Duration

The mean length of conversations, measured in seconds or displayed as minutes:seconds. What duration indicates:
  • Very short calls might suggest quick resolutions or early disconnects
  • Very long calls could indicate complex issues or inefficiencies
  • Optimal duration depends on your use case

Total Duration

Cumulative time spent on calls, showing total investment in conversations. Use this metric to:
  • Understand time commitment
  • Plan capacity
  • Calculate efficiency ratios

Daily Activity Analysis

The daily activity chart visualizes call patterns over time.

Reading the Chart

X-axis: Dates within the analysis period Y-axis: Number of calls per day Visual elements: Bars or lines representing daily call volumes

Identifying Patterns

Consistent volume: Similar call counts each day suggest stable operations. Growth trends: Increasing daily volumes may indicate successful marketing, seasonal demand, or expanding use. Decline patterns: Decreasing volumes warrant investigation - are there configuration issues, or has demand shifted? Weekly cycles: Many businesses see patterns by day of week:
  • Higher volume Monday-Wednesday
  • Lower volume Thursday-Friday
  • Minimal weekend activity (for B2B)
Spikes and anomalies: Sudden peaks may correlate with:
  • Campaign launches
  • Marketing activities
  • External events affecting caller behavior

Using Daily Data

Capacity planning: Identify your busiest days to ensure adequate coverage. Campaign timing: Launch outbound campaigns on days with historically better answer rates. Performance investigation: Correlate daily performance drops with specific events or changes.

Status Distribution Charts

Visual breakdown of call outcomes helps identify where improvements are needed.

Understanding Status Categories

Success: Calls that completed with positive outcomes. The AI employee fulfilled the conversation objective. Voicemail: Calls where voicemail was detected. Depending on configuration, a message may have been left. No Answer: Calls that rang but were not answered. No voicemail system was detected. Busy: The contact’s line was engaged when the call was attempted. Failed: Technical issues prevented call completion. This could include network errors, system issues, or audio problems. Canceled: Calls that were terminated before completion, either programmatically or by user intervention.

Visual Representation

Status distribution is typically shown as:
  • Pie charts showing proportional breakdown
  • Bar charts comparing category counts
  • Color-coded segments for quick recognition

Analyzing Distribution

Healthy inbound profile:
  • High Success percentage (70%+)
  • Minimal Failed calls
  • Low Busy rates
Typical outbound profile:
  • Moderate Success (20-40%)
  • Significant Voicemail and No Answer
  • Some Busy during business hours
Warning signs:
  • High Failed percentage suggests technical investigation needed
  • Very low Success on inbound indicates configuration issues
  • Unexpected status changes from historical patterns

Hourly Distribution Patterns

Understanding when calls occur helps optimize operations.

The Hourly Chart

This visualization shows call volume distributed across hours of the day. X-axis: Hours (0-23 or displayed as AM/PM) Y-axis: Number of calls during each hour

Identifying Peak Hours

Most businesses have predictable peak periods: B2B typical peaks:
  • Late morning (10-11 AM)
  • Early afternoon (1-3 PM)
B2C typical peaks:
  • Lunch hours
  • Early evening (5-7 PM)
  • Weekend afternoons

Using Hourly Insights

Optimize outbound timing: Schedule campaigns for hours when answer rates are highest. Plan human backup: Ensure transfer destinations are staffed during peak hours. Identify opportunities: Off-peak hours might offer opportunities for scheduled callbacks.

Analysis Tab and Conversation Review

The Analysis tab provides detailed access to individual conversations handled by the AI employee.

Conversation List

A table displays all conversations with key information: Title: The conversation subject or automatically generated title Status: Outcome of the conversation Duration: Length of the call Date: When the conversation occurred

Accessing Conversation Details

Click any conversation row to navigate to the full conversation detail page, where you can:
  • Read the complete transcript
  • Listen to the audio recording
  • View conversation properties
  • See evaluation results

Conversation Counts

The header shows the total number of conversations matching your current filters, helping you understand the scope of data you are reviewing.

Filtering and Searching Conversations

Powerful filtering helps you find specific conversations or analyze subsets of data.

Search Functionality

The search bar allows text-based searching across:
  • Conversation titles
  • Contact names
  • Phone numbers
Search tips:
  • Use partial matches to find related conversations
  • Search for specific topics discussed
  • Combine with filters for precise results

Filter Options

Status filter: Select one or more status types to include:
  • Success
  • Failed
  • Voicemail
  • No Answer
  • Busy
  • Canceled
Direction filter: Focus on inbound or outbound calls:
  • All directions
  • Inbound only
  • Outbound only
Channel filter: Filter by communication channel if multiple are used. Date filters: Narrow results by time period:
  • Preset ranges (Last 3 days, Last 7 days, etc.)
  • Custom date range selection
Duration filters: Find calls within specific length ranges:
  • Minimum duration
  • Maximum duration
Phone number search: Find calls involving specific numbers.

Property Filters

If you have configured custom properties, these appear as additional filter options:
  • Select property type
  • Choose values to filter by

Active Filter Display

When filters are applied, badges appear showing your active selections:
  • Click badges to remove individual filters
  • Use clear options to reset all filters

Evaluation Tracking

If evaluation criteria are configured for this AI employee, the Analysis tab includes tracking capabilities.

Evaluation Criteria Display

A sidebar or panel shows:
  • All evaluation criteria assigned to this AI employee
  • Which criteria are actively being tracked
  • Links to configure or modify criteria

Understanding Evaluations

Evaluation criteria are AI-powered assessments that automatically analyze conversation quality:
  • Did the AI employee achieve the call objective?
  • Were required disclosures made?
  • Was the tone appropriate?
Viewing evaluation results: Individual conversation details include evaluation scores when available.

Tracking Properties

Alongside evaluation criteria, custom properties can be tracked and analyzed:
  • Properties auto-extracted during conversations
  • Manual property assignments
  • Property-based filtering and reporting

Using Analytics for Optimization

Analytics exist to drive improvement. Here is how to turn data into action.

Identifying Issues

Low success rates: Listen to sample conversations to understand why calls are not succeeding:
  • Is the prompt inadequate for common questions?
  • Are there technical audio quality issues?
  • Is the target audience appropriate?
High failed rates: Technical investigation priorities:
  • Check phone number configuration
  • Review network and connectivity
  • Examine error logs if available
Unexpected patterns: Sudden changes warrant investigation:
  • What changed in configuration recently?
  • Are there external factors affecting calls?
  • Is there a technical issue?

Configuration Refinement

Use analytics to guide configuration changes: Prompt improvements: If common questions are not being handled well, update the system prompt to address them. Timing adjustments: Schedule campaigns for hours with better historical success. Transfer optimization: If transfers are high, consider whether the AI employee needs expanded capabilities.

Comparative Analysis

If you have multiple AI employees:
  • Compare performance across similar roles
  • Identify what top performers do differently
  • Apply successful patterns to underperformers

Trend Monitoring

Track metrics over time to identify:
  • Gradual improvements from optimization efforts
  • Degradation that needs attention
  • Seasonal patterns to anticipate

Reporting and Insights

Exporting Data

While the platform may not offer direct data export in all areas, you can:
  • Screenshot charts for presentations
  • Use the command palette to ask the AI assistant for specific data summaries
  • Navigate to individual records for detailed information

Building Reports

For periodic reporting, develop a standard approach:
  1. Capture key metrics at consistent intervals
  2. Document trends and significant changes
  3. Note actions taken and their results
  4. Plan next optimization steps

Sharing Insights

Use analytics to communicate with stakeholders:
  • Volume metrics show scale of operations
  • Success rates demonstrate effectiveness
  • Trend data illustrates improvement over time

AI-Assisted Analysis

Throughout the platform, “Ask AI” links provide access to intelligent analysis:
  • Request explanations of patterns
  • Ask for optimization recommendations
  • Get help interpreting complex data

Summary

AI employee analytics provide the visibility needed to continuously improve your voice communication operations. By understanding what each metric means, how to interpret charts, and how to filter for specific insights, you can make data-driven decisions that enhance performance. Regular analytics review, combined with action on insights, creates a cycle of continuous improvement that maximizes the value of your AI employees.