Adoption metrics alone are insufficient for evaluating design system health — they only confirm a package is installed, not that it's working. A complementary engagement metric tracks the relationship between the design system team and product teams through three signals: office hours participation, contribution activity, and support volume. Each signal is normalized by team size and composited into a monthly score with a rolling average. A real example shows engagement dropping from 95% to 79% in May while adoption remained flat, enabling targeted intervention that brought June back to 91%. The post explains how to read adoption and engagement together using a quadrant model, warns against weaponizing the score or falling into Goodhart's Law, and provides a simple spreadsheet-based implementation guide.
