Chatbot Engagement

This guide provides deep insights into how users interact with your chatbot across sessions, messages, and resolutions.

ChatBot Engagement – Help Documentation

The ChatBot Engagement tab provides deep insights into how users interact with your chatbot across sessions, messages, and resolutions. These metrics help you understand engagement quality, user behavior, and where human intervention is required.

---

📊 Metrics & Charts Explained

1. Total Sessions

What it shows:

• Total number of chatbot sessions initiated by users within the selected time range.

How to use it:

• Identify traffic trends over time

• Measure chatbot adoption

Filters available:

• Date range

• Bot selector

• View type: Daily / Weekly / Monthly

---

2. Unique Users vs Returning Users

What it shows:

• Unique Users (Blue): New users interacting with the bot

• Returning Users (Green): Users who have interacted before

Why it matters:

• Helps measure user retention

• Indicates chatbot stickiness and repeat engagement

---

3. Session Duration

What it shows:

• Average duration (in seconds) of chatbot sessions

Insights you can draw:

• Longer sessions may indicate deeper engagement

• Very short sessions may indicate poor responses or confusion

---

4. Messages Per Session

What it shows:

• Average number of messages exchanged in a single session

Why it’s important:

• Higher values often mean stronger engagement

• Extremely high values may indicate looping or unresolved intents

---

5. Escalation Rate

What it shows:

• Percentage of sessions escalated from bot to human agent

How to interpret:

• High escalation rate → Bot needs better training or coverage

• Low escalation rate → Bot is resolving queries effectively

---

6. Resolution Rate

What it shows:

• Percentage of sessions resolved successfully without escalation

Best practice:

• Aim for a higher resolution rate to reduce support workload

---

7. Drop-off Rate

What it shows:

• Percentage of users who left the conversation before completion

What it indicates:

• High drop-off may mean:

  • Poor flow design

  • Long response times

  • Irrelevant bot replies

FAQs

What is considered a session?

A session starts when a user begins interacting with the chatbot and ends after inactivity or completion of the conversation.

What’s the difference between unique and returning users?

Unique users are first-time visitors, while returning users have interacted with the chatbot before.

Why is my escalation rate high?

This usually means the bot is unable to answer certain queries and is handing them over to a human agent.

How is resolution rate calculated?

Resolution rate is the percentage of sessions successfully completed by the bot without escalation.

What causes a high drop-off rate?

Users may drop off due to unclear flows, slow responses, or irrelevant answers.

Can I filter data for a specific bot?

Yes, use the Pick bots filter at the top of each chart.Yes. You can download conversation and contact data directly from the Inbox screens.