On this page

Conversation Analytics

Your analytics dashboard gives you a clear view of how your customer conversations are performing. Every conversation is analysed by AI, and the results are aggregated here so you can spot trends, understand your customers, and improve your support.

Key Metrics

Your analytics dashboard displays these metrics at the top of the page:

Total Conversations: The overall number of conversations in your selected date range. Track growth and seasonal patterns to plan resources.

Total Messages: The total messages exchanged during your selected date range. A growing message count alongside stable conversation counts may indicate longer, more complex interactions.

Human Handover Rate: The percentage of conversations where the AI was disabled and a human agent took over. A balanced rate (typically 15–30%) indicates healthy AI/human collaboration. Too high? Your AI needs more training data. Too low? Make sure customers can still reach humans when needed.

Average Lead Score: The mean lead score (0–100) across all conversations in the period. Higher scores indicate more engaged, high-intent visitors. This metric helps you gauge the quality of traffic your channels are attracting.

The Quality and Audience section also shows new conversations started during the period and returning conversations that became active again.

Visualisations

Intent Breakdown

See how conversations are distributed across all detected intent types:

  • RESOLVED — AI handled successfully
  • SALES_INQUIRY — Customer interested in purchasing
  • PRODUCT_INQUIRY — Questions about products
  • TECHNICAL_SUPPORT — Technical issues
  • GENERAL_INQUIRY — General questions
  • ORDER_INQUIRY — Order status questions
  • ORDER_PROCESSING — Placing or modifying orders
  • PURCHASE_INQUIRY — Ready to buy
  • COMPLAINT — Customer complaints
  • ESCALATION — Needs higher-level attention
  • FEEDBACK — Customer providing feedback
  • GREETING — Initial greetings
  • NEED_HUMAN — Customer explicitly asked for a human
  • OTHER — Doesn't fit other categories

Use this to understand what your customers actually need and prioritise your knowledge base accordingly.

Emotion Distribution

Every customer message is analysed for emotional tone. The dashboard shows the distribution across 10 emotion levels:

  • 😡 Furious / 😠 Angry — Immediate attention needed
  • 😟 Disappointed / 😢 Sad — Extra care required
  • 😐 Neutral — Standard interaction
  • 🙂 Content / 😊 Pleased — Positive experience
  • 😄 Cheerful / 😁 Joyful / 🤩 Ecstatic — Your biggest fans

Use emotion trends to measure whether your support is improving customer mood over time.

Lead Score Distribution

Lead scores are grouped into five ranges:

  • 0–20 — Cold leads, general browsing
  • 21–40 — Some interest shown
  • 41–60 — Engaged, considering options
  • 61–80 — Warm leads, likely to convert
  • 81–100 — Hot prospects, high purchase intent

Source Distribution

See which channels drive the most conversations: WhatsApp, website chat widget, Instagram, Facebook Messenger, Telegram, or SMS.

Top Topics

The AI extracts recommended topics from each conversation. This view shows up to the 12 most frequently discussed topics in a three-column layout, helping you identify:

  • Common questions that should be in your knowledge base
  • Product areas generating the most interest
  • Recurring issues that need attention

Line charts showing conversation volume and message volume per day, helping you identify:

  • Peak days and quiet periods
  • Growth or decline trends
  • The impact of marketing campaigns or product launches

Accessing Analytics

  1. Navigate to Monitoring in your sidebar
  2. Select the Analytics page
  3. Use Today, 7 days, 30 days, 90 days, or the custom date fields to narrow your view

Downloading a PDF Report

Select a reporting period and choose Download PDF. The generated PDF uses the currently selected range and includes:

  • Performance and audience metrics
  • Daily conversation and message trends
  • Intent and emotion breakdowns
  • Up to 12 top topics

The dashboard and report remain readable on mobile devices; cards and charts stack vertically on smaller screens.

Using Analytics to Improve

Knowledge Base Optimisation: Use intent breakdown and top topics to prioritise new article creation. High NEED_HUMAN rates for specific topics indicate knowledge gaps.

AI Performance Tuning: Monitor human handover rate and intent breakdowns. Review recurring handover intents and update your AI's training data.

Lead Prioritisation: Focus your sales efforts on conversations with high lead scores. Use the lead score distribution to understand your traffic quality.

Channel Strategy: Compare source distribution to see which channels drive the most valuable conversations, and allocate marketing resources accordingly.

Keywords

conversation analytics, customer support metrics, AI performance tracking, support analytics, intent breakdown, emotion distribution, lead score, conversation insights, analytics dashboard, performance monitoring