When the platform measures team performance, is it based only on actual activity inside the system, like model updates, issue resolution, and coordination work, or are managers still relying on manual judgment outside the platform?