---
title: Analyzing AI effectiveness
description: Measure the performance and adoption of AI capabilities on your instance to identify what is working and where to improve.
locale: en-US
canonical_url: https://www.servicenow.com/docs/r/intelligent-experiences/platai-analyzing-ai-effectiveness.html
release: brazil
topic_type: concept
last_updated: "2026-09-10"
reading_time_minutes: 2
keywords: [AI analytics, Now Assist analytics, Conversation Insights, Conversation Improvement Themes, Inferred CSAT, adoption metrics]
breadcrumb: [Enable AI Experiences]
---

# Analyzing AI effectiveness

Measure the performance and adoption of AI capabilities on your instance to identify what is working and where to improve.

Deploying AI capabilities is only the beginning. Measure performance and adoption continuously to confirm that your AI investments deliver value and to identify opportunities for improvement.

## AI analytics

Effective AI measurement spans several dimensions: adoption \(how many users engage with AI features\), usage \(which features are used and how often\), sentiment \(whether users find AI outputs helpful\), accuracy \(how well AI responses meet quality standards\), and value \(quantifiable outcomes such as time saved or cases deflected\). Establishing baselines for these metrics before launching AI capabilities enables you to assess impact accurately.

AI Admin Center provides built-in analytics dashboards that track usage, adoption, skills performance, and user feedback across your instance.

For information about built-in analytics dashboards, see [AI analytics](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/brazil/markdown/intelligent-experiences/platai-ai-analytics-overview.md).

## Conversation Insights

Conversation Insights delivers Inferred CSAT scores and explanatory factors for conversations in Virtual Agent and agentic workflows. The application uses AI to analyze conversations in real time and provides insights that help improve Virtual Agent interactions, live agent interactions, and agentic workflows.

Inferred CSAT is a score from 1 \(least satisfied\) to 5 \(most satisfied\) predicted entirely from conversation transcripts without requiring input from the user. Along with the score, the model predicts contributing factors such as resolution, confusion, effort, empathy, next steps, frustration, and transfers and escalations. These factors explain user satisfaction or dissatisfaction, helping you target specific improvements.

To measure Inferred CSAT and the factors behind it across Virtual Agent and agentic workflow conversations, see [Conversation Insights](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/brazil/markdown/intelligent-experiences/conversational-intelligence/conversation-insights.md).

## Conversation Improvement Themes

Conversation Improvement Themes automatically surfaces common topics and patterns across conversations, helping you prioritize which enhancements will have the greatest impact on the user experience. Use these themes to identify recurring issues, understand where conversations succeed or fail, and focus improvement efforts where they matter most.

**Note:** Analytics capabilities across the platform are consolidating. Monitor the AI Admin Hub console for the latest analytics features added in each release.

To surface recurring themes across conversations and prioritize the improvements with the greatest impact, see [Conversation Improvement Themes](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/brazil/markdown/intelligent-experiences/conv-impr-themes-landing.md).