Key concepts – Analytics in Copilot Studio

Important

Power Virtual Agents capabilities and features are now part of Microsoft Copilot Studio following significant investments in generative AI and enhanced integrations across Microsoft Copilot.

Some articles and screenshots may refer to Power Virtual Agents while we update documentation and training content.

Copilot Studio has a comprehensive set of analytics that show you the key performance indicators for your copilot.

Note

Conversation transcripts in Dataverse are unavailable for download on the Microsoft Copilot Studio app in Teams. To review and export transcripts in Dataverse, you'll need to sign up for the Microsoft Copilot Studio web app. Session chat transcripts can be downloaded via the Microsoft Copilot Studio app. For more details, refer to Download copilot session transcripts.

Multiple charts show you trends and usage for your copilot's topics. These charts use AI to highlight the topics that have the greatest impact on your copilot's performance.

As a copilot maker, your job isn't done when you publish your copilot for use. You'll want to know how well your copilot is performing its intended duties. The comprehensive analytics in Microsoft Copilot Studio keep your copilot's key performance indicators front and center. Multiple charts show trends and usage for your copilot's topics. These charts use AI to highlight the topics that have the greatest effect on your copilot's performance.

Analytics in Microsoft Copilot Studio

In Microsoft Copilot Studio, in the navigation menu, select Analytics.

The analytics section is divided into a number of pages to give you multiple ways to understand copilot performance.

The following terms are used in this section.

  • A conversation is an ongoing interaction between a specific user, or group of users, on a channel and your copilot. Conversations can pause and resume later, or be transferred to a live agent. The conversation may be one-way, either from the end user to the copilot, or from the copilot to the end user, but is more commonly a back-and-forth interaction between the end user and the copilot.
  • Analytics sessions track user engagement for your copilot. Analytics sessions try to capture how well your copilot handles user tasks.
  • Billed sessions are used to measure resource consumption. Billed sessions are independent of analytics sessions.

Transcripts are available for download about 20 minutes after an analytic session ends, as conversation transcripts in Dataverse via the Power Apps portal and as session chat transcripts via the Microsoft Copilot Studio app. It can take up to an hour after the analytic session ends before the appearing in the analytics dashboard.

In the classic version of Microsoft Copilot Studio, conversation transcripts in Dataverse are available approximately 6 hours after the analytic session starts.

Note

Copilot analytics features aren't yet available for Microsoft Copilot Studio US Government Cloud plans. In the meantime, you can use a custom analytics solution to create a Power BI dashboard to get analytics in these plans.

View analytics

In the side navigation pane, select Analytics.


Article Description
Analyze copilot performance and usage Review overall analytics for your copilots.
Analyze customer satisfaction Identify overall trends in customer satisfaction, and where satisfaction is falling behind.
Analyze topic usage See how individual topics are doing and how well they're working at resolving issues.
Analyze session information Review and monitor how sessions are doing, including analytics from the topics triggered during a session.
Analyze billed session information Understand how sessions are being billed, based on their usage.
Work with conversation transcripts Export conversation transcripts from Dataverse and create custom reports.