How to Build a Data Metric System: Purpose, Methodology, and Application

As economic growth slows, businesses need a scientific and effective way to drive continuous growth, making a data metric system essential.

The talk, presented by Xu Yao from Zhiguan Technology, outlines the purpose of a metric system: monitoring business changes, enabling problem tracing, and providing feedback for solutions.

It introduces the OSM model—Object (goal), Strategy (actions), Measure (metrics)—and details four steps: define the core (North Star) metric, identify related factors, assign metrics to each factor, and promote, archive, and implement the metrics.

Examples include determining a ride‑hailing driver’s core metric (rating) and using the model to break down revenue components and influencing factors.

Further sections explain how the metric system helps answer why a metric changed, what factors are involved, and how to resolve issues, supported by diagrams.

The Q&A covers the impact of focusing on a single core metric, the role of product and business teams in building the system, promotion strategies, and the distinction between metric and tag systems.

Overall, a well‑designed metric system provides global, systematic information to drive business decisions, describe the current state, uncover causes, predict the future, and improve team efficiency.