Skip to main content

AI Performance Analytics

Updated over 2 months ago

The AI Performance tab inside Journey Analytics provides deeper insights into AI-driven conversations, helping businesses optimize automation and improve customer interactions.


Key Metrics:

  1. Unsatisfactory Responses – Identify where AI struggles and enhance customer interactions.

  2. Tickets Automated – Track AI's impact in reducing manual workload.

  3. Intent-wise Split – Analyze satisfactory vs. unsatisfactory messages for each intent.

  4. Customer Message Intent – Visualized using pie charts and bar graphs for quick insights.


How This Helps

With these insights, brands can:

  • Fine-tune AI performance

  • Reduce escalations

  • Enhance customer experience by improving automation efficiency.


Analytics Breakdown

A. Conversation & Automation Overview

  1. Unique Sessions – Number of distinct customers who interacted with AI.

  2. Tickets Automated – Number of tickets fully resolved without agent intervention.

  3. Total AI Messages Sent – Count of all AI-generated messages.

  4. Average Conversation Length – Average duration of AI-handled conversations.

  5. Average Interaction Time (sec) – Time spent per interaction between customers and AI.


B. Unsatisfactory Responses Overview

  1. Total Unsatisfactory Messages – Number of messages where AI response was deemed unsatisfactory.

  2. % of Unsatisfactory Responses – Percentage of responses classified as unsatisfactory.

C. Intent-wise Split of Responses

  • Detailed breakdown of satisfactory and unsatisfactory responses for each intent.

D. Customer Message Intent

  • Percentage of customer messages categorized by intent.


How to View Analytics

  1. Navigate to the BIK Dashboard.

  2. Go to Journey Analytics.

  3. Select the AI Performance tab to access the insights.


Frequently Asked Questions (FAQs):

1. What does "Unsatisfactory Responses" mean?

These are AI-generated messages that were marked as ineffective, unclear, or unhelpful to the customer.

2. Can I track AI’s impact on reducing agent workload?

Yes, the Tickets Automated metric shows how many tickets were resolved without agent intervention.

3. What visualizations are available?

Customer message intent is displayed as pie charts and bar graphs for easy analysis.

4. What is the significance of intent-wise response splits?

It helps businesses understand which intents lead to satisfactory or unsatisfactory responses, enabling targeted improvements in AI training.


For further assistance or to raise feature requests related to AI Performance Analytics, please contact [email protected].


Did this answer your question?