How Baidu’s Light Design System Uses PATS Metrics to Boost Efficiency

Why Measure a Design System

Metrics provide a reliable way to evaluate experience, turning qualitative insights into quantifiable data. As Peter Drucker said, "If you can’t measure it, you can’t improve it." Baidu’s Light Design System needed measurable, sustainable monitoring to guide improvements and allocate resources effectively.

How to Build Design System Metrics

2.1 Research Existing Metric Models

We reviewed industry and academic models such as HEART+GSM, PTECH, and cloud‑product experience quality models, asking:

What metric models already exist in the industry?

Can design‑system metrics be derived from these models?

Analysis showed most models focus on specific products and users, with limited coverage of design‑system evaluation. However, common traits—dimension division, indicator derivation via GSM (Goal‑Signal‑Metric), and a mix of objective and subjective methods—can guide our own framework.

2.2 Constructing the Design System Metric

We defined the metric to assess the design system’s workflow and business impact, establishing a four‑step process:

Step 01 – Inventory Current State

Identify existing content, pain points, and standards (principles, assets, tools, support) to define what should be measured, forming the PATS metric system.

Step 02 – Define Metric Dimensions

We focused on five dimensions:

Generality – coverage of business needs and adaptability.

Usability – ease of understanding and use.

Efficiency – improvement of work speed.

Reliability – stability and maintainability.

Satisfaction – overall user satisfaction.

Step 03 – Set Goals, Signals, and Indicators

Using the GSM method, we derived specific targets, observable signals, and concrete indicators for each dimension, ensuring alignment with system objectives.

Step 04 – Choose Measurement Methods

We created questionnaires for subjective scoring, conducted qualitative interviews to explore weak areas, and added objective data tracking for component usage to reflect flexibility.

PATS Metric in Practice

Initial measurements revealed low scores in usability and reliability, prompting targeted improvements such as switching to Figma for asset management, which significantly raised usability scores. Establishing a maintenance team and bi‑weekly updates boosted reliability.

Conclusion

Implementing the PATS metric gave Baidu a comprehensive, data‑driven view of the Light Design System, guiding decisions that enhanced experience and maturity. As the system evolves, the metric framework will continue to expand, providing richer insights for ongoing improvement.