GitHub adds opt-in peer benchmarks to its Copilot impact dashboard

GitHub added industry comparisons to its Copilot impact dashboard on October 22, allowing eligible enterprises and standalone organizations to compare AI coding-agent adoption against aggregated benchmarks from participating peers. The feature is disabled by default, and organizations that belong to an enterprise participate through their parent enterprise rather than independently.

The new dashboard section measures agentic adoption, defined as the percentage of active Copilot users classified as agent-first or multi-agent, and leverage, which compares pull-request merge rates for agentic and nonagentic users. Administrators can select a company-size group and one of ten industries, with comparisons based on a 28-day reporting period. GitHub says rankings may initially be unavailable or use a broader peer group until enough organizations have opted in.

Enabling the comparison requires the Copilot usage metrics policy and a separate opt-in setting that enterprise or eligible organization owners can control. Organizations that opt out stop seeing comparisons immediately, and their contributed data is removed from peer cohorts in the next daily update. GitHub says aggregate benchmarks omit identifying information and are deleted after 30 days. Customers using GitHub Enterprise Cloud with data residency are compared only with participants in the same region.

For now, industry comparisons are displayed only in the Copilot impact dashboard, not in its usage-metrics API, NDJSON exports or downloads. GitHub's accompanying documentation explains that the underlying measure tracks pull-request merge activity, so these comparisons describe adoption and repository activity rather than directly measuring software quality.