Understanding and Building Data Indicator Systems: JD’s Practices and Methodologies

The article introduces JD’s insights and practices on building data indicator systems, organized into six main parts.

1. Understanding Indicator Systems – It defines basic concepts of metrics and indicator systems, using everyday examples (e.g., health check‑ups) and explains how business metrics are derived from data collection and analysis.

2. Metrics vs. Tags – It clarifies the differences in meaning, processing methods, and application scenarios between quantitative metrics (statistical calculations) and descriptive tags (entity attributes, manual labeling, or AI‑generated predictions).

3. Position in the Data Chain – It maps the indicator system to the four‑layer data pyramid (Data, Information, Knowledge, Wisdom) and describes how indicators link upstream business goals with downstream data models and analytics.

4. Traffic Indicator System – It outlines how to design traffic‑related metrics (exposure, clicks, PV, UV, conversion rates, etc.), emphasizing the need to separate dimensions from metrics, limit the number of core metrics, and classify them as atomic or derived.

5. Effective Landing of Indicator Systems – It presents key guarantees such as scientific, complete, and goal‑oriented design, standard definitions, and development processes, including the OSM model for aligning metrics with business strategies.

6. OSM Model and Business Alignment – It describes how to extract a north‑star metric, combine it with the OSM framework, and use DuPont analysis to decompose business objectives into measurable indicators.

The article also shares JD’s standardized metric definition (4W1H + dimension) and low‑code development platform that enables non‑technical users to define, compute, and monitor metrics consistently.