How to Build a Data Metric System for E‑commerce, Community, and Finance Apps

Data metric systems form the foundation of business analysis, enabling quick anomaly detection, operational monitoring, and health assessment.

Typical workflow includes clarifying business goals, defining metric dimensions and measurement methods, establishing data collection and processing pipelines, designing storage and query solutions, and implementing visualization and reporting.

What Is a Data and Metric System?

Metrics are objective quantitative standards—also called "measures"—that evaluate business performance, such as PV, UV, session duration, and bounce rate.

Because a single metric rarely captures the full picture, multiple metrics are combined into a metric system for multidimensional analysis, linking various levels and dimensions to reflect business status and enable rapid response.

Why Build a Data Metric System?

Relying on a single indicator cannot fully represent complex, multi‑dimensional operations; for example, production cost alone misses sales insights, while sales amount alone omits profit context. A systematic metric system helps identify core issues, formulate strategies, and support informed decision‑making for sustained growth.

Key benefits include:

Decision‑makers gain a comprehensive view of business health and can set effective North Star metrics.

Front‑line staff receive guidance for strategy formulation and business expansion.

Data analysts reduce routine data extraction, freeing time for exploratory analysis and deeper insights.

Unified definitions ensure metric reliability and adaptability as the business evolves.

How to Build a Data Metric System?

General steps:

Determine the North Star metric.

Break it down into sub‑metrics.

Further decompose sub‑metrics into process metrics.

Add classification dimensions.

Below are industry‑specific examples.

1. E‑commerce Scenario

North Star Metric: Gross Merchandise Volume (GMV).

Sub‑metrics: GMV = Number of Transactions × Average Order Value.

Further breakdown: Number of Transactions = Users × Conversion Rate.

Process metrics follow the funnel model (visit → add‑to‑cart → purchase), covering metrics such as site visits, conversion rates, and repeat purchase rates.

Classification dimensions may include product category, sales channel, region, and user attributes.

2. Community Scenario (e.g., Zhihu)

North Star Metric: User interaction count.

Sub‑metrics: Interaction = Readers × Key‑action Conversion Rate × Interaction Rate.

Process metrics use the "People‑Content‑Place" model or AARRR framework to capture various stages of user engagement.

People: producers vs. consumers.

Content: articles, videos, ads, paid consultations.

Place: homepage, search, recommendation, member slots.

3. Financial App Scenario

North Star Metric: Daily active users (DAU).

Sub‑metrics: DAU = New Users + Active Users + Returning Users + Lost Users.

Process metrics follow the AARRR model (Acquisition, Activation, Retention, Revenue, Referral) to break down each stage.

Classification dimensions are similar to the previous scenarios, tailored to financial contexts.

Metric systems must evolve with business changes; regular reviews and updates ensure they remain accurate and continue to support effective decision‑making.