Sooner or later, the same question comes up in almost every tenant I work with. Customers want to know how to measure whether Copilot for Microsoft 365 is actually being used, and whether the money spent on those licenses is justified. Copilot licenses aren't cheap, so people want hard numbers on who's enabled, who's active and which apps see real engagement. That's why Copilot usage reports matter well beyond license assignment. They give you transparency and control, and they form the basis for ongoing governance.

The admin center and Viva Insights both provide Copilot usage reports

I always start in the Microsoft 365 admin center. Under Reports → Usage → Copilot for Microsoft 365 you'll see which users have Copilot licenses assigned and how many of them actively use Copilot in each application. The date range filter is useful for comparing the phases before and after rollout. One detail matters a lot here. If the "concealed user names" feature is enabled (Microsoft 365 admin center → Settings → Org settings → Reports), usernames in these reports are anonymized by default. In Germany this often isn't optional, because works councils typically require it. Check your settings before you plan any analysis at the person level.

The Copilot Dashboard in Viva Insights gives you far more to work with. If you have Copilot for Microsoft 365 and your user groups are large enough for aggregation, the dashboard shows usage rates by department or function, so you can see which teams are adopting Copilot most eagerly. Keep in mind that Viva Insights aggregates groups smaller than ten for privacy reasons, which means you won't get granular results for very small teams. Prerequisites and features are documented under Copilot Dashboard in Viva Insights.

Group-based licensing in Entra ID keeps license assignment under control

Copilot licenses are expensive, and not everyone needs one on day one. That's why I always recommend managing assignment with group-based licensing in Entra ID. With Entra ID P1 or higher you can create dynamic groups and assign Copilot only to pilot users or specific functions. This keeps you flexible, since you can move licenses around quickly when pilot projects change or when you notice licenses nobody uses. Microsoft's documentation explains how group-based licensing in Entra ID works. I covered the practical side in an earlier post, Copilot for Microsoft 365: The real challenge is your tenant.

Pro tip: Assigning licenses is simple, and that simplicity can mislead you. If you hand out lots of Copilot licenses and see little real usage, get into the habit of checking the usage reports and reclaiming the ones sitting idle. A monthly review works well during pilot phases and lets you adjust as the rollout matures.

This is the workflow I typically use for group-based license management:

flowchart TD
    start["Define group for Copilot users"]
    assign["Assign Copilot license to group"]
    sync["Automatic license assignment to members"]
    report["Track usage via reports/dashboard"]
    reassign["Reassign unused licenses"]
    start --> assign --> sync --> report --> reassign

Figure: Typical flow for group-based Copilot license assignment and monitoring in Entra ID and Microsoft 365.

Concealing usernames protects privacy in Copilot reports

Reporting on Copilot usage is often a sensitive topic in German tenants, because user analytics fall under works council co-determination. Microsoft accounts for this with a setting that conceals usernames in all reports, located under Org settings → Reports. With it enabled, analytics stay at group level and nothing can be attributed to individual people. Many works councils I've dealt with insist on this before they approve Copilot or other Microsoft 365 analytics tools. If you genuinely need user-level detail for adoption work, bring legal and the works council on board early.

Usage reports reveal unused licenses you can reclaim

Once the initial rollout is done, you'll usually find that not everyone with a Copilot license is using it. The usage report in the Microsoft 365 admin center makes these cases easy to spot. I suggest treating them as a reason to talk to the teams involved. Often a bit of enablement is all it takes, and sometimes licenses turn out to have been assigned to admins or service accounts by mistake. Cleaning up regularly keeps the investment where it actually delivers value.

With group-based assignment, removing a user from the Entra ID group frees their license immediately so you can reassign it. The change syncs quickly and appears in the next reporting cycle.

Usage tracking is more than admin work in my view

In my experience, tracking Copilot usage goes well beyond optimizing licenses. It's the main lever you have for steering adoption and getting real value out of Copilot. If you use reporting well, you can target enablement, put licenses where people actually want them and refine your Copilot program based on real data. What matters most is treating reporting as a way to keep expanding Copilot's impact rather than as a stick. If you're planning a pilot, my article Copilot pilot: Planning a successful first few weeks has tips for that. Whatever else you do, track usage as you roll out Copilot, if only because those licenses are far too valuable to waste.