Step-by-Step Guide: Tracking Microsoft Copilot Usage by Department After the July 2025 Admin-Center Update

Introduction

The July 2025 Microsoft Admin Center update introduced a game-changing "Agents usage" report that finally gives IT administrators granular visibility into Copilot adoption across their organization. This new reporting capability addresses a critical gap that has plagued organizations trying to measure AI tool effectiveness without relying on user surveys. (Worklytics)

With over 50,000 businesses already integrating Copilot into their workflows, understanding departmental usage patterns has become essential for maximizing ROI and identifying adoption gaps. (Worklytics) The new admin center features allow you to segment usage by department, geography, licensing status, and even track user-created agents - providing the comprehensive analytics needed to drive strategic AI adoption decisions.

This guide will walk you through the complete process of accessing and leveraging these new reporting capabilities, from role prerequisites to data export options, ensuring you can deliver actionable insights to leadership about your organization's AI transformation journey.


Understanding the July 2025 Admin Center Updates

What's New in the Agents Usage Report

Microsoft's July 2025 update represents a significant leap forward in AI usage analytics. The new "Agents usage" report provides unprecedented visibility into how your organization is leveraging Copilot across different business units. (Microsoft 365 Blog)

Key features of the updated reporting include:

Departmental segmentation: Break down usage by organizational units
Geographic distribution: View adoption patterns across different locations
Licensing status tracking: Identify unlicensed usage that might be hidden
User-created agents monitoring: Track custom agent deployment and usage
Real-time usage metrics: Access up-to-date adoption data
Export capabilities: Download raw data for further analysis

The timing of this update aligns perfectly with the growing need for organizations to measure AI adoption scientifically. Research shows that to truly understand Copilot's impact, organizations should approach it as a scientific experiment or continuous improvement project. (Worklytics)

Integration with Autonomous Agents

The March 2025 release of autonomous agents in Copilot Studio adds another layer of complexity to usage tracking. These agents are designed to automate responses to critical events, eliminating the need for manual checks and scheduled workflows. (Microsoft Copilot Blog) The new admin center reports now capture usage data from these autonomous agents, providing a complete picture of AI utilization across your organization.


Prerequisites and Role Requirements

AI Administrator Role

To access the new Agents usage report, you must have the AI Administrator role assigned in your Microsoft 365 tenant. This role was specifically created to manage AI-related features and reporting across the Microsoft ecosystem.

Steps to verify your role:

1. Navigate to the Microsoft 365 Admin Center
2. Go to Users > Active users
3. Search for your account
4. Check the Roles section for "AI Administrator"

If you don't have the AI Administrator role:

1. Contact your Global Administrator
2. Request assignment of the AI Administrator role
3. Wait for role propagation (typically 15-30 minutes)
4. Sign out and back into the admin center

Additional Permissions

Beyond the AI Administrator role, ensure you have:

Reports Reader permissions for comprehensive data access
Exchange Administrator role if tracking email-based Copilot usage
Teams Administrator role for Teams-integrated agent monitoring

These additional permissions ensure you can access all relevant usage data across Microsoft's productivity suite, which is crucial for comprehensive AI adoption measurement. (Worklytics)


Accessing the Agents Usage Report

Navigation Path

1. Sign in to the Microsoft 365 Admin Center (admin.microsoft.com)
2. Navigate to Reports in the left sidebar
3. Select Usage from the Reports menu
4. Click on Microsoft 365 Copilot section
5. Choose Agents usage from the available reports

Initial Report View

Upon first access, you'll see a dashboard displaying:

Total active agents across your organization
Usage trends over the past 30 days
Top departments by agent utilization
Geographic distribution of usage
Licensing compliance overview

Common Access Issues and Solutions

Issue: "Access Denied" error when trying to view reports
Solution: Verify AI Administrator role assignment and wait for permission propagation

Issue: Empty or incomplete data in reports
Solution: Check that Copilot licenses are properly assigned and users have completed initial setup

Issue: Missing departmental breakdowns
Solution: Ensure Entra ID attributes are properly configured (covered in the next section)

Many organizations struggle with incomplete usage data because they haven't properly configured their directory attributes. This is where platforms like Worklytics excel, as they can integrate with over 25 collaboration tools to provide comprehensive usage analytics. (Worklytics)


Segmenting Usage by Department

Configuring Entra ID Attributes

To enable departmental segmentation, you must ensure your Entra ID (formerly Azure AD) user attributes are properly configured. The Agents usage report relies on these attributes to categorize users and their AI usage patterns.

Required Attributes:

Attribute Purpose Example Values
Department Primary segmentation "Sales", "Marketing", "Engineering"
Office Geographic segmentation "New York", "London", "Tokyo"
JobTitle Role-based analysis "Sales Manager", "Developer", "Analyst"
CompanyName Multi-tenant organizations "Subsidiary A", "Division B"
Manager Hierarchical reporting Manager's UPN or Object ID

Steps to Configure Attributes:

1.

Access Entra ID Admin Center

• Navigate to entra.microsoft.com
• Sign in with Global Administrator credentials
2.

Bulk Update User Attributes

• Go to Users > All users
• Select Bulk operations > Bulk update
• Download the CSV template
• Populate department, office, and job title fields
• Upload the completed CSV file
3.

Verify Attribute Propagation

• Wait 2-4 hours for changes to propagate
• Check a sample user's profile to confirm updates
• Test the Agents usage report for departmental breakdowns

Best Practices for Department Tagging

Standardize Department Names: Use consistent naming conventions across your organization. Avoid variations like "Sales", "Sales Dept", and "Sales Department" - pick one format and stick to it.

Create Hierarchical Structures: Consider using parent-child relationships for large organizations:

• "Engineering - Frontend"
• "Engineering - Backend"
• "Engineering - DevOps"

Regular Maintenance: Schedule quarterly reviews of user attributes to ensure accuracy as employees change roles or departments.

Proper departmental segmentation is crucial for understanding AI adoption patterns. Research shows that many organizations segment usage by team, department, or role to uncover adoption gaps and optimize their AI investment. (Worklytics)


Geographic and Licensing Analysis

Geographic Segmentation

The July 2025 update enables geographic analysis of Copilot usage, helping global organizations understand regional adoption patterns and compliance requirements.

Key Geographic Metrics:

Usage by office location: Based on the "Office" attribute in Entra ID
Time zone analysis: Understanding peak usage hours across regions
Compliance tracking: Ensuring data residency requirements are met
Regional performance: Comparing adoption rates between locations

Setting Up Geographic Tracking:

1.

Configure Office Attributes

• Ensure all users have accurate "Office" values in Entra ID
• Use standardized location names (e.g., "New York, NY, USA")
• Include country codes for international compliance tracking
2.

Enable Location-Based Reporting

• In the Agents usage report, select "Geographic view"
• Choose your preferred geographic grouping (country, state, city)
• Apply filters to focus on specific regions

Licensing Status Monitoring

One of the most valuable features of the new reporting is the ability to identify unlicensed Copilot usage. This addresses a common blind spot where users access AI features through shared accounts or trial periods.

Licensing Categories Tracked:

License Type Description Tracking Method
Microsoft 365 Copilot Full commercial license Direct license assignment
Copilot Pro Individual subscription Microsoft account linking
Trial Access Temporary evaluation Time-limited usage tracking
Unlicensed Usage Unauthorized access Anomaly detection algorithms

Identifying Unlicensed Usage:

1.

Access License Compliance View

• Navigate to the Agents usage report
• Select "Licensing" tab
• Review "Unlicensed Activity" section
2.

Analyze Usage Patterns

• Look for users with high activity but no assigned licenses
• Check for shared account usage across multiple departments
• Identify trial users who should be converted to paid licenses
3.

Take Corrective Action

• Assign appropriate licenses to legitimate users
• Disable access for unauthorized usage
• Update compliance policies to prevent future violations

Understanding licensing compliance is crucial for budget planning and ROI calculation. With Copilot licenses costing approximately $240 per year per user, proper license management can result in significant cost savings. (Worklytics)


Tracking User-Created Agents

Understanding Agent Types

The March 2025 Copilot Studio updates introduced several new agent types that organizations need to track:

Autonomous Agents: These agents wait for specific events and execute actions automatically using natural language instructions rather than code. (Microsoft Copilot Blog)

Custom Copilots: User-created agents designed for specific business processes or departmental needs.

Workflow Agents: Agents that integrate with existing business processes and automate routine tasks.

Monitoring Agent Creation and Usage

Agent Creation Metrics:

1. Creation Rate: Number of new agents created per department per month
2. Creator Profiles: Which roles and departments are most active in agent creation
3. Agent Complexity: Measuring the sophistication of user-created agents
4. Adoption Rate: How quickly new agents are adopted by other team members

Usage Analytics:

Interaction Volume: Total interactions with user-created agents
Success Rate: Percentage of successful agent interactions
User Satisfaction: Implicit feedback through continued usage
Business Impact: Correlation with productivity metrics

Accessing Agent Analytics:

1.

Navigate to Agent Details

• In the Agents usage report, select "User-Created Agents"
• Choose a specific agent or view aggregate data
• Review creation date, creator, and usage statistics
2.

Analyze Department-Specific Patterns

• Filter by department to see which teams are most innovative
• Compare agent creation rates across different business units
• Identify successful agent patterns for replication
3.

Export Detailed Data

• Use the export function to download agent-specific usage data
• Analyze trends in external tools like Power BI or Excel
• Create custom dashboards for executive reporting

Tracking user-created agents provides valuable insights into organizational innovation and AI adoption maturity. This data helps identify power users who can serve as AI champions and departments that may need additional training or support.


Data Export and Integration Options

CSV Export Functionality

The July 2025 update includes robust export capabilities that allow you to extract raw usage data for further analysis.

Export Options Available:

Export Type Data Included Best Use Case
Summary Report High-level metrics by department Executive dashboards
Detailed Usage Individual user interactions Compliance auditing
Agent Analytics Custom agent performance data Innovation tracking
Licensing Report License compliance status Budget planning
Geographic Data Location-based usage patterns Regional analysis

Steps to Export Data:

1.

Select Export Type

• In the Agents usage report, click "Export"
• Choose your desired data granularity
• Select date range for the export
2.

Configure Export Parameters

• Choose file format (CSV, Excel, JSON)
• Select specific departments or users
• Apply any necessary filters
3.

Download and Verify

• Wait for export processing (typically 2-5 minutes)
• Download the generated file
• Verify data completeness and accuracy

Power BI Integration

For organizations using Microsoft's business intelligence platform, the new admin center provides direct Power BI connectivity.

Setting Up Power BI Integration:

1.

Enable API Access

• In the admin center, navigate to Settings > API Access
• Enable "Microsoft Graph reporting APIs"
• Configure authentication credentials
2.

Create Power BI Connection

• Open Power BI Desktop
• Select "Get Data" > "Microsoft Graph"
• Authenticate with your AI Administrator credentials
• Choose "Copilot Usage" data source
3.

Build Custom Dashboards

• Create visualizations for departmental usage trends
• Build executive summary reports
• Set up automated refresh schedules

Worklytics Integration Benefits

While Microsoft's native reporting provides valuable insights, platforms like Worklytics can enrich this data with additional productivity signals and cross-platform analytics. Worklytics integrates with over 25 collaboration tools and uses machine learning to clean, de-duplicate, and standardize datasets. (Worklytics)

Enhanced Analytics with Worklytics:

Cross-platform correlation: Connect Copilot usage with email, calendar, and project management data
Productivity impact measurement: Correlate AI usage with actual output metrics
Privacy-compliant analysis: Leverage data anonymization and aggregation for GDPR compliance
Advanced segmentation: Create custom user cohorts based on multiple data sources

Worklytics' approach to measuring AI adoption without user surveys aligns perfectly with the new Microsoft admin center capabilities, providing a comprehensive view of how AI tools impact actual work patterns. (Worklytics)


Common Misconfigurations and Troubleshooting

Hidden Unlicensed Usage Issues

Based on preview rollout feedback, several common misconfigurations can hide unlicensed usage from admin reports:

Issue 1: Shared Account Usage

Problem: Multiple users sharing a single Copilot license, making individual usage tracking impossible.

Solution:

• Implement conditional access policies requiring individual authentication
• Monitor for unusual usage patterns (24/7 activity, multiple simultaneous sessions)
• Use the "Concurrent Sessions" metric to identify shared accounts

Issue 2: Trial License Confusion

Problem: Users continuing to access Copilot features after trial periods expire through cached credentials.

Solution:

• Set up automated alerts for trial expiration dates
• Implement grace period policies with clear communication
• Use the "License Transition" report to track trial-to-paid conversions

Issue 3: Department Attribute Gaps

Problem: Users without proper department attributes appearing as "Unassigned" in reports.

Solution:

• Run regular audits of Entra ID user attributes
• Implement automated workflows to populate missing attributes
• Create default department assignments for new users

Data Accuracy Validation

Verification Steps:

1.

Cross-Reference License Assignments

• Compare admin center usage data with license assignment reports
• Identify discrepancies between assigned licenses and reported usage
• Investigate users with high usage but no assigned licenses
2.

Validate Department Segmentation

• Spot-check individual user assignments against HR records
• Verify that department totals match expected organizational structure
• Test filter functionality with known user groups
3.

Monitor Data Freshness

• Check report timestamps to ensure data is current
• Understand the reporting delay (typically 24-48 hours)
• Set up alerts for data processing failures

Performance Optimization

Large Organization Considerations:

For organizations with thousands of users, report generation can be slow. Optimize performance by:

Using date range filters to limit data scope
Applying department filters before generating reports
Scheduling exports during off-peak hours
Breaking large exports into smaller, manageable chunks

Proper configuration and regular maintenance of these reporting systems is essential for accurate AI adoption measurement. Organizations that take a scientific approach to measuring Copilot's impact see better results and can make data-driven decisions about their AI investments. (Worklytics)


Building Your Repeatable Tracking Checklist

Weekly Monitoring Tasks

Every Monday:

• [ ] Review previous week's usage trends by department
• [ ] Check for new unlicensed usage alerts
• [ ] Verify data freshness and report accuracy
• [ ] Update stakeholders on key metrics changes

Every Wednesday:

• [ ] Analyze user-created agent adoption rates
• [ ] Review geographic usage patterns for anomalies
• [ ] Check license compliance status
• [ ] Export data for external analysis if needed

Every Friday:

• [ ] Prepare weekly summary report for leadership
• [ ] Identify departments needing additional AI training
• [ ] Plan follow-up actions for low-adoption areas
• [ ] Archive reports for historical trend analysis

Monthly Strategic Reviews

Department Performance Analysis:

• Compare adoption rates across business units
• Identify high-performing departments for best practice sharing
• Calculate ROI by department using productivity metrics
• Plan targeted interventions for low-adoption areas

License Optimization:

• Review license utilization rates
• Identify candidates for license reassignment
• Plan budget for additional licenses based on demand
• Analyze cost per active user trends

Innovation Tracking:

• Catalog successful user-created agents
• Identify innovation leaders within the organization
• Plan knowledge sharing sessions
• Measure business impact of custom agents

Quarterly Business Reviews

Executive Reporting:

• Prepare comprehensive AI adoption dashboard
• Calculate overall ROI and productivity gains
• Benchmark against industry standards
• Present strategic recommendations for AI expansion

Compliance Auditing:

• Conduct thorough license compliance review
• Validate data privacy and security measures
• Update policies based on usage patterns
• Plan for regulatory reporting requirements

Strategic Planning:

• Assess need for additional AI tools or features
• Plan training programs based on usage data
• Identify opportunities for process automation
• Set targets for next quarter's adoption goals

This systematic approach to tracking and analysis ensures that organizations can maximize their AI investment while maintaining compliance and driving continuous improvement. The key is consistency in monitoring and responsiveness to the insights generated. (Worklytics)


Measuring Business Impact and ROI

Productivity Correlation Analysis

The real value of tracking Copilot usage by department lies in correlating this data with actual business outcomes. Research shows that developers with Copilot completed coding tasks 55% faster and with higher completion rates. (Worklytics)

Key Productivity Metrics to Track:

Department Primary Metrics Copilot Impact Indicators
Engineering Code commits, PR merge rate, bug resolution time Faster feature delivery, fewer bugs
Sales Deal velocity, proposal creation time, follow-up response rate Shorter sales cycles, higher close rates
Marketing Content creation speed, campaign launch time, A/B test frequency More campaigns, faster iteration
Customer Support Ticket resolution time, first-call resolution rate, customer satisfaction Faster resolutions, happier customers
Finance Report generation time, audit preparation speed, analysis depth More timely insights, better accuracy

ROI Calculation Framework

Time-Value Analysis:

If Copilot saves a user even 2 hours per week, that equals approximately 100 hours per year. At $60/hour, that represents $6,000 in productivity value gained annually, while the Copilot license costs about $240/year - a 25x payoff in pure time-value terms. (Worklytics)

Quality Improvement Benefits:

• Reduced error rates in code, documents, and analysis
• Faster identification and resolution of issues
• Improved consistency across team outputs
• Enhanced creativity and innovation in problem-solving

Employee Satisfaction Impact:

• Higher job satisfaction from reduced mundane tasks
• Improved retention rates for AI-enabled teams
• Increased engagement with modern, efficient tools
• Better work-life balance through productivity gains

Advanced Analytics Integration

For organizations seeking deeper insights, integrating Copilot usage data with comprehensive workplace analytics platforms provides a more complete picture. Worklytics specializes in leveraging existing corporate data to deliver real-time intelligence on how work gets done, analyzing collaboration, calendar, communication, and system usage data without relying on surveys. (Worklytics)

**Enhanced Measurement Capabilities

Frequently Asked Questions

What new reporting capabilities were introduced in the July 2025 Microsoft Admin Center update?

The July 2025 update introduced a game-changing "Agents usage" report that provides IT administrators with granular visibility into Copilot adoption across their organization. This addresses the critical gap that previously forced organizations to rely on user surveys to measure AI tool effectiveness, now offering detailed departmental usage analytics.

How can organizations measure Copilot success beyond basic adoption metrics?

According to Worklytics research, high adoption metrics are necessary but not sufficient for achieving downstream benefits. Organizations should segment usage by team, department, or role to uncover adoption gaps and measure efficiency gains. Many successful companies like Dow anticipate saving millions of dollars through strategic Copilot implementation in their operations.

What role requirements are needed to access the new departmental Copilot usage reports?

The new "Agents usage" report requires specific administrative permissions within the Microsoft Admin Center. Organizations need to ensure their IT administrators have the appropriate roles assigned to access these granular departmental analytics and export capabilities introduced in the July 2025 update.

Can the new Copilot usage data be exported for further analysis?

Yes, the July 2025 Admin Center update includes data export options for the "Agents usage" report. This allows organizations to integrate Copilot usage metrics with their existing analytics tools and data warehouses for comprehensive productivity analysis across departments.

How does departmental Copilot tracking help improve AI adoption in organizations?

Departmental tracking helps identify adoption gaps and enables targeted interventions to boost AI proficiency. By analyzing usage patterns across different teams, organizations can provide focused training, address specific barriers, and optimize their AI investment ROI by ensuring consistent adoption across all departments.

What are the key benefits organizations have seen from effective Copilot measurement and tracking?

Organizations with effective Copilot tracking have reported significant operational improvements. For example, Dow anticipates millions in savings from shipping operations in the first year, while GitHub Copilot has grown to over 1.3 million developers across 50,000+ organizations. Proper measurement enables organizations to identify high-impact use cases and scale successful implementations.

Sources

1. https://www.microsoft.com/en-us/microsoft-365/blog/2024/12/12/the-latest-on-ai-at-work-december-2024/
2. https://www.microsoft.com/en-us/microsoft-copilot/blog/copilot-studio/whats-new-in-copilot-studio-march-2025/
3. https://www.worklytics.co/blog/adoption-to-efficiency-measuring-copilot-success
4. https://www.worklytics.co/blog/introducing-worklytics-for-ai-adoption-measure-benchmark-and-accelerate-ai-impact-across-your-organization
5. https://www.worklytics.co/blog/measuring-github-copilots-impact
6. https://www.worklytics.co/blog/the-roi-of-github-copilot-for-your-organization-a-metrics-driven-analysis
7. https://www.worklytics.co/blog/tracking-employee-ai-adoption-which-metrics-matter
8. https://www.worklytics.co/get-started
9. https://www.worklytics.co/integrations
10. https://www.worklytics.co/privacy-policy