GitHub Copilot Usage Metrics API Now Tracks AI Adoption Cohorts
Key Takeaways
- ▸GitHub Copilot's usage metrics API now classifies users into four adoption phases based on their 28-day usage patterns
- ▸Organizations can track progression from code-first to agent-first to multi-agent adoption workflows
- ▸New granular per-phase metrics enable data-driven enablement strategies and deeper understanding of AI adoption beyond simple active user counts
Summary
Microsoft has enhanced GitHub Copilot's usage metrics API with new cohort classification capabilities that go beyond simple active user counts. The update introduces an "ai_adoption_phase" field that automatically categorizes engaged users into four distinct phases based on their Copilot product usage over a rolling 28-day window. This enables enterprise administrators and organization owners to understand not just who is using Copilot, but how they are adopting and evolving their usage patterns.
The four adoption phases classify users based on the Copilot surfaces they've engaged with on at least two days in the past 28 days: Phase 0 (no engagement), Phase 1 (code-first usage focused on code completion and IDE agents), Phase 2 (agent-first usage with a single GitHub-based agent), and Phase 3 (multi-agent adoption involving multiple agent surfaces or the new GitHub Copilot app). Each phase includes a version field to allow the classification logic to evolve as Copilot's capabilities expand.
New granular metrics are now available at the enterprise and organization level, including total engaged users per phase, interaction averages, code generation and acceptance metrics, lines of code changes, pull request activity, and median time-to-merge. These per-phase metrics enable organizations to tell a richer adoption story, track how users graduate from code-first to multi-agent workflows, and target enablement programs where they see the greatest opportunity for adoption.
