Extend analytics with Power BI data model customization

  • 10 minutes

The out-of-the-box dashboards cover the most common reporting needs, but organizations often have more unique reporting requirements. Power BI data model customization can help solve for these requirements by exposing the underlying data models used by the historical and real-time analytics reports, letting report authors extend them and publish the results back into Copilot Service workspace.

The process follows four steps: selecting a Microsoft Fabric workspace, provisioning the data models, granting permissions, and embedding the customized reports back into Dynamics 365.

Prerequisites

Before you begin, confirm the following prerequisites are in place:

  • Licenses: Report viewers, authors, and administrators all have specific licensing requirements. Review the Microsoft Product Terms and Power Platform Pricing pages to confirm the right licenses before starting.
  • Analytics features enabled: For historical data model customization, at least one historical report must be enabled. For real-time data model customization, Omnichannel real-time analytics must be enabled.
  • Microsoft Entra ID security group: Your Microsoft Entra ID administrator must create a security group and add the Dynamics 365 Analytics service account as a member. This service principal deploys data models and manages the Microsoft Fabric workspace. If the Dynamics 365 Analytics service account isn't available in your organization, use the Dynamics CCA Data Analytics service account instead.
  • Power BI tenant settings: Your Power BI administrator must enable several settings in the Microsoft Fabric Admin portal, including workspace creation, service principal access to Fabric APIs, DirectQuery connections, XMLA endpoints, and content embedding. See Customize data models of historical and real-time analytics reports for the full list.

Enable Power BI data model customization

Once prerequisites are met:

  1. In Copilot Service admin center, navigate to Operations > Insights.
  2. In the Report settings section:
    • For historical: select Embedded Power BI extensibility—Historical data model customization, then Manage.
    • For real-time: select Embedded Power BI extensibility—Real-time data model customization, then Manage.
  3. Switch the Enable embedded Power BI data model customization toggle to On.

Step 1: Select a Microsoft Fabric workspace

Specify the customer workspace where Dynamics 365 deploys the data model copies and reports.

  1. From the Insights page, go to the data model configuration for historical or real-time.
  2. Select Create new workspace to create a new Microsoft Fabric workspace, or select an existing workspace from the dropdown list.
  3. Select Save to start provisioning.

Note

If you're using an existing workspace, the user enabling model customization must be a workspace administrator of that workspace. Microsoft automatically creates a separate managed workspace.

You can use the same customer workspace for both historical and real-time analytics, or separate workspaces.

Step 2: Provision the data models

Provisioning deploys copies of the out-of-the-box data models and reports to your specified workspace. This can take up to 24 hours. Select Refresh on the settings page to check the current status.

Step 3: Grant permissions for semantic models and reports

After provisioning completes, select the Power BI Data model link to open the managed workspace. Use the Share dialog to grant access by entering user email addresses. Both authors and viewers need Allow recipients to build content with the data associated with this semantic model access.

Optionally, select the Customized Power BI Report link to open the customer workspace and grant access based on role:

Role Access level
Report authors Contributor
Report viewers Viewer
Supervisors viewing embedded reports in Dynamics 365 Member

Step 4: Embed customized reports back to Dynamics 365

After a report author publishes a customized report to the Microsoft Fabric workspace, you can make it available in Copilot Service workspace.

  1. From the data model customization page, go to Step 4.
  2. Select Add report.
  3. Enter a Report name. This is the label users see in the workspace site map.
  4. From the Select Power BI report dropdown, select the report from your workspace.
  5. Select Add, then Save.

The dropdown lists all reports in the workspace configured in Step 1. You can add up to 40 reports. Reports from historical and real-time customization share the same site map section in Copilot Service workspace, and you can reorder them across both pages.

Note

For custom reports, manually update the semantic data model to receive product updates, then refresh and redeploy the custom reports.

Manage visual customizations across environments

Organizations with separate development, test, and production environments can use the built-in export and import workflow to promote visual customizations without manually recreating them in each environment.

From the data model customization page, report authors with the appropriate role can download the current report version to store offline for audit and rollback purposes, then import it into another environment's workspace. This supports an end-to-end lifecycle management pattern, moving visual updates from development to production with version control and governance.

Note

Role-based access controls who can promote visual updates. Only users with the required permissions can import or overwrite report versions in a given workspace.

Enable automatic page refresh for real-time reports

For real-time customized reports, you can configure automatic page refresh so supervisors always see current data.

  1. In Microsoft Fabric Admin portal, enable Automatic page refresh for your capacity.
  2. Download the real-time report from Copilot Service workspace.
  3. In the report, go to Visualizations > Format page and switch Page refresh to On.
  4. Set Auto page refresh as the refresh type, with a 5-second or 10-second interval.
  5. Publish the updated report back to the customized workspace.

Note

Automatic page refresh is only available for real-time analytics reports, not historical.

Monitor capacity and performance

Two tools help you track how your customized reports perform under load.

Microsoft Fabric Capacity Metrics app shows how your customized data models and reports consume shared capacity in Microsoft Fabric. Use views like Compute, Timepoint, and Item Detail to identify whether complex measures or high query volumes are driving compute demand. This helps you catch scaling issues before they affect report performance.

Performance Analyzer in Power BI Desktop measures how long each visual takes to load, broken down across DAX queries, DirectQuery calls, and rendering time. For customized semantic models, this helps identify the impact of complex measures and composite model relationships. Capture performance data during real interactions like applying slicers, export the logs, and use them to validate whether model optimizations produce measurable improvements.