A Practical Framework for Data Analysis

Start with the decision#

A metric system should describe what matters to a business and help people decide what to do next. The source argues that useful indicators should reflect current priorities, include both outcome and process measures, and connect to actions that a team can take.

For example, GPA can summarize one aspect of a student’s academic performance, but it cannot describe the whole picture. Project and research experience may provide additional process evidence. Neither type should be treated as a complete measure on its own.

Build a metric system with OSM#

The article introduces OSM as a way to connect a goal to measurable work:

• Objective: state the goal clearly.
• Strategy: break the goal into strategies and intermediate steps.
• Measure: choose indicators for those steps, aiming for a set that is distinct enough to interpret and broad enough to cover the intended work.

The point is to make the path from an objective to a practical intervention visible. A collection of numbers without an associated decision process is unlikely to guide improvement.

A general analysis workflow#

The source recommends clarifying the question, checking the data, breaking down the problem, selecting an analysis method, and translating findings into actions. It also stresses engineering skills, product understanding, business context, modeling knowledge, and project execution as complementary parts of data work.

These are broad reflections rather than a reproducible case study. The original Chinese article contains the full discussion of indicator design, common mistakes, and its example workflow.