Observação
GitHub Copilot usage metrics are currently in public preview with data protection and subject to change.
After you assign Copilot licenses across your enterprise, you can use the Copilot usage metrics dashboard and APIs to verify that licenses are active and monitor early usage trends. This helps you evaluate whether your rollout is reaching the right people and take quick action if adoption is slower than expected.
To get a wider view of adoption, you can combine dashboard insights with qualitative feedback, for example, short pulse surveys or team check-ins.
Prerequisite
"Copilot usage metrics" must be enabled on the AI control page. See Gerenciar políticas e recursos do GitHub Copilot em sua empresa.
Step 1: Access the usage metrics dashboard
- Go to the Enterprises page and select your enterprise.
- Click the Insights tab.
- In the left sidebar, click Copilot usage.
The dashboard displays aggregated telemetry from developer IDEs and updates daily, though data may be up to 3 days behind UTC.
Step 2: Track license activation
Use the "IDE active users" and "IDE daily active users" charts to confirm that developers are starting to use their assigned licenses.
| Metric | What it tells you | How to act |
|---|---|---|
| IDE active users | How many licensed developers have used Copilot at least once this calendar month. | Compare to your total assigned licenses. If the number is much lower, verify IDE configuration or communicate activation steps to your teams. |
| IDE daily active users | How many unique users are using Copilot each day. | Look for an upward trend in the first two weeks after rollout. A flat line may signal that users need additional enablement or setup guidance. |
| IDE weekly active users | Rolling 7-day total of active users. | Use this to track consistency. Steady or increasing WAU indicates successful activation and recurring use. |
Observação
Copilot usage metrics reflect activity in supported IDEs only. Usage in Copilot Chat on GitHub.com, GitHub Mobile, Revisão de código do Copilot, or CLI do Copilot is not included in dashboard data.
Step 3: Identify early adoption signals
Once licenses are active, focus on the metrics that indicate healthy early adoption.
| Signal | Where to find it in the dashboard | What to look for |
|---|---|---|
| Consistent DAU growth | “IDE daily active users” graph | Steady increase in daily users over the first month. |
| Feature variety | “Requests per chat mode” graph | Developers trying multiple chat modes (Ask, Edit, Agent) suggests curiosity and engagement. |
| Initial agent usage | “Agent adoption” card | Even small agent adoption (5–10%) early on is a positive signal of experimentation. |
Healthy early adoption usually looks like 60–80% of assigned users showing activity within the first month.
Step 4: Act on limited adoption signals
If your metrics show limited adoption, try one of the following strategies.
| Symptom | Possible cause | What to do |
|---|---|---|
| Low total active users | Users haven’t activated licenses or configured Copilot in their IDE. | Re-share onboarding materials or run a short “activation check” session. |
| Flat daily usage | Developers tried Copilot once but aren’t returning. | Provide enablement resources like Copilot prompts or internal success stories. |
| No agent usage | Teams may not know about the Agente do Copilot. | Share examples of advanced use cases to encourage exploration. |
You can also use Copilot Chat to help diagnose issues. For example:
What patterns might explain low adoption across some teams in the Copilot metrics export?
What patterns might explain low adoption across some teams in the Copilot metrics export?
Step 5: Track activation programmatically
Now that you have a sense of what adoption data is available to you and what the data tells you about your rollout, consider using the API to continue monitoring adoption over time.
- Use the Enterprise Usage endpoint to retrieve aggregated data for all organizations in your enterprise.
- Filter the export by
dayoruser_loginto identify newly activated users. - Compare
user_initiated_interaction_countandcode_acceptance_activity_countto see whether users are actively engaging after license assignment.
Step 6: Export user-level data for deeper analysis
In some cases, you may need user-level activity data for deeper analysis or to integrate with internal BI tools. Exports are most useful when you want to analyze long-term trends or correlate adoption with other metrics (for example, productivity or enablement activities).
-
In the dashboard, click Export JSON in the top-right corner.
-
Download the NDJSON file and use Copilot Chat to analyze the data. For example, ask:
Copilot Chat prompt * Summarize which organizations show the largest increase in `loc_added_sum` this month. * Identify users with high `user_initiated_interaction_count` but low `code_acceptance_activity_count`.
* Summarize which organizations show the largest increase in `loc_added_sum` this month. * Identify users with high `user_initiated_interaction_count` but low `code_acceptance_activity_count`.