We've updated the ServiceNow Community Code of Conduct, adding guidelines around AI usage, professionalism, and content violations. Read more

Sharon_Barnes
ServiceNow Employee

 

Crafting Custom Dashboards and Reporting for Now Assist in ServiceNow

As organizations begin adopting AI capabilities within ServiceNow, one of the biggest questions quickly becomes: How do we measure the impact of AI?

ServiceNow provides Now Assist Analytics out of the box to track usage and adoption, but many organizations need deeper insights tailored to their internal KPIs. In this Platform Academy session, Sharon Barnes walks through how to extend those analytics and create custom dashboards and reports to better understand AI performance.

 


Understanding Now Assist Analytics

Now Assist Analytics provides a built-in framework that allows administrators to track AI activity across the platform without building custom reports from scratch.

These dashboards provide visibility into:

  • AI skill usage

  • Adoption trends

  • Self-service performance

  • Deflection metrics

  • User interaction insights

These dashboards are accessible through the Now Assist Admin Console and are integrated with Performance Analytics starting in the Zurich release.

The key benefit is that ServiceNow not only provides prebuilt dashboards but also exposes the underlying indicators, tables, and analytics data, allowing organizations to extend reporting beyond the default views.


The Key Table for Custom Reporting

When creating custom reports for Now Assist, the most important data source is:

sys_gen_ai_usage_log (Gen AI Usage Logs)

This table contains critical information such as:

  • Which AI skill was used

  • Application source (ITSM, Creator, etc.)

  • Number of assists consumed

  • User information 

Because the table references user records, you can dot-walk into user attributes such as:

  • Department

  • Location

  • Cost center

  • Business unit

This allows organizations to answer questions like:

  • Which departments use Now Assist the most?

  • Which teams are getting the most value?

  • Where is adoption growing?


Reusing Existing Now Assist Dashboard Visualizations

One interesting challenge with the out-of-box dashboards is that their visualizations are not automatically available in the Data Visualization Library.

To reuse them:

  1. Navigate to Platform Analytics

  2. Open the Now Assist Analytics dashboard the report you want is locate.

  3. Duplicate the Dashboard

  4. Open the dashboard in edit mode.

  5. Select the visualization you want.

  6. Use the “Add to Library” option.

  7. Once saved, the visualization becomes available in Platform Analytics dashboards.

This allows you to reuse existing analytics components when building custom dashboards for different personas.


Building Persona-Based Dashboards

Instead of giving stakeholders access to a complex analytics dashboard with dozens of widgets, a better approach is to create focused dashboards for specific roles.

Examples include:

Platform Owner Dashboard

  • Overall Now Assist usage

  • Skill distribution

  • Time saved metrics

IT Operations Dashboard

  • ITSM skill usage

  • Deflection trends

  • Ticket resolution times

Executive Dashboard

  • AI adoption growth

  • ROI indicators

This approach reduces information overload and allows stakeholders to quickly see the metrics that matter to them.


Leveraging Performance Analytics Indicators

The real power of customization comes from Performance Analytics indicators.

You can filter indicators using: Application = Now Assist Analytics

 
This reveals all the out-of-box indicators used by the Now Assist Performance dashboards.

These indicators can then be used to build new visualizations


Now Assist Readiness Evaluation

Before implementing Now Assist or Agentic AI features, ServiceNow provides a helpful tool called the Now Assist Readiness Evaluation.

This tool analyzes your instance and checks:

  • Data quantity
  • Configuration deviations

It runs entirely within your instance and does not send data externally.

The tool is available through the ServiceNow Store and can be installed on Yokohama or later releases.

A best practice is to:

  1. Clone production to a development environment

  2. Run the readiness evaluation

  3. Investigat any findings before deploying Now Assist. Remember the goal is not 100% perfect. This is a tool to improve the quality of AI results and speed of implementation.

For more information 


Introducing Assistant Analytics

ServiceNow recently introduced Assistant Analytics, which provides insights into conversational AI performance.

This includes analytics for:

  • Virtual Agent

  • Now Assist pPanel

  • Voice Assistants

Key insights include:

  • Conversation volumes

  • Channel usage (Slack, Teams, etc.)

  • Escalation to human agents

  • Customer sentiment

  • CSAT trends

This helps organizations understand whether AI conversations are helpful, frustrating, or improving over time.

Assistant Analytics requires the Zurich release and is included with Now Assist for Platform.
For more information 


What’s Coming: Now Assist Center

ServiceNow is also developing a new tool called Now Assist Center, currently available in Innovation Lab.

The goal is to simplify AI adoption by centralizing tools related to:

  • Setup

  • Insights

  • Analytics

  • Adoption tracking

Think of it as a single hub for managing Now Assist and Agentic AI capabilities, similar to how ServiceNow Studio unified multiple developer tools.

For more information 


Key Takeaways

When working with Now Assist analytics, keep these best practices in mind:

  1. Define your baseline metrics first – understand KPIs before introducing AI.

  2. Use out-of-box dashboards when possible – then extend them.

  3. Leverage the GenAI usage logs table for custom reporting.

  4. Create persona-based dashboards instead of large generic ones.

  5. Use Performance Analytics indicators to unlock deeper insights.

By combining these techniques, organizations can move beyond simple usage tracking and begin measuring the real value of AI within ServiceNow.