Understanding Data, Metrics, and Indicator Systems: From Basic Concepts to Pirate Metrics and First‑Key‑Metric Method

1. Data

Data refers to raw, unprocessed records.

The essence of data is the mathematical observation, recording, and understanding of the world; data analysis is the human process of moving from qualitative to quantitative, from vague to precise.

People prefer to make decisions based on data rather than lengthy textual descriptions.

2. Metrics

Metric = data + business scenario; it guides the formulation of the next action plan.

Example: Weight is a data point; 120 kg does not automatically mean overweight, nor does 60 kg mean underweight, because other factors such as height are needed. Body‑fat percentage, however, directly reflects the proportion of fat in the body and can guide health‑related actions, making it a true metric.

A good metric should answer the 5 W questions.

1. Usage scenario (who, when, where)

Define dimensions to clarify the analysis contexts a metric supports, e.g., body‑fat percentage can be broken down by gender, age group, region, etc.

2. Metric definition (what)

Clarify the calculation scope to avoid “same name, different meaning” or “different name, same meaning” issues, as illustrated by the sales‑amount and stock‑quantity metrics.

3. Metric purpose (why)

Define logical relationships between metrics, e.g., Sales Profit = Sales Revenue – Purchase Cost – First‑leg Tax – Tax Refund Difference; Gross Profit = Sales Profit – Obsolete Inventory Provision – Capital Occupation Interest.

3. Metric Systems

After clarifying the problems a metric should solve, the next step is to build a metric system. Two common methodologies are presented: the Pirate Metrics (AARRR) model and the First‑Key‑Metric method (also known as the North Star metric).

Pirate Metrics (AARRR)

In 2007, Dave McClure of 500 Startups introduced the PirateMetrics framework, dividing startup‑relevant metrics into Acquisition, Activation, Retention, Revenue, and Referral.

The model is useful for internet companies focusing on traffic‑to‑revenue conversion, but may not cover all scenarios for traditional e‑commerce businesses.

First‑Key‑Metric Method

The core idea is that at any point in time a company has one most critical metric, which may change as the business evolves. Different business models (e‑commerce, SaaS, mobile apps, two‑sided markets, media, UGC) and stages (MVP, growth, revenue) require different metric systems.

Implementation steps (illustrated with a cross‑border e‑commerce company):

Identify the first‑key metric – in this case, sales revenue, despite being a “vanity metric”.

Divide metrics into modules (user conversion, retention, procurement, warehousing, logistics, cost, timeliness, etc.).

Map logical relationships from the first‑key metric downwards.

Summary: Different industries and development stages produce varied “metric trees”. While the Pirate Metrics and First‑Key‑Metric methods suit different scenarios, both aim to turn metrics into actionable business decisions; combining methodologies as needed is recommended.

Welcome to like, bookmark, and share this post!