Adoption numbers say how often the system is used. Health says whether what ships still looks like what was approved. About a quarter of teams measure consistency: in the zeroheight Design Systems Report 2026, 41% of teams measure adoption and 26% measure product consistency (zeroheight).
The signals behind it
One score, with the list behind it
Leadership needs one number, and teams need the findings behind it. A simple score weights findings by reach. For example (illustrative): if 2,400 usages of tokens and components are tracked and 120 of them are affected by open divergences or bypasses, health is 100 × (1 − 120 ÷ 2,400) = 95.
In this model, time to fix stays outside the score and is tracked next to it as a trend, such as the median number of days a finding stays open.
How FISYCO scores health
FISYCO builds its score differently: a weighted average of six parts, from 0 to 100.
A part with no data drops out together with its weight instead of counting as zero. Each component also gets its own component health score on the same 0–100 scale, from parts of its own: design match, accessibility, tokens, variant axes, API and adoption.
What to avoid
A score nobody can open is a vanity metric. Every point lost should lead to a finding someone can fix. And the score should be recomputed automatically, on every push or on a schedule, so the trend is real rather than a quarterly guess.
Related
- Design system assurance
- How to measure design system health
- Design system health in FISYCO: what each part counts and what would raise the score.