Skip to content

DAP — Data Analysis Process

The six-phase framework. Used to solve the right problem with the right data.

Not linear

The biggest mistake analysts make is looking for quick answers. Go back and forth between phases. Review each phase to improve.

Phases

# Phase What happens
1 Ask Define the problem, determine business challenge, identify objectives, and align with stakeholder expectations.
2 Prepare Decide what data is needed, locate sources, ensure ROCCC (Reliable, Original, Comprehensive, Current, Cited), and handle data generation, collection, storage, and management.
3 Process Clean, transform, and ensure integrity. Fix missing values, formatting errors, and handle duplicates to prepare data for analysis.
— Data Cleaning Detailed cleaning techniques (lives inside Process)
4 Analyze Perform EDA, identify trends and relationships, validate data, and apply analytical patterns to answer business questions.
— Metrics & KPIs Common business metrics and how to track them.
— Experimentation A/B testing workflow and statistical concepts.
5 Share Create visualizations, build presentations, communicate results, and make findings accessible to drive decisions.
6 Act Put insights to work by implementing new strategies, monitoring metrics, and iterating on the process.

PACE (alternative framework)

  • Plan — goals, scope, stakeholder needs
  • Analyze — acquire, clean, transform, EDA
  • Construct — build/revise models, statistical inference
  • Execute — present findings, address feedback

Data Life Cycle

  1. Plan — what data, who owns
  2. Capture — collect from sources
  3. Manage — care for and maintain
  4. Analyze — solve, decide, support goals
  5. Archive — keep for future reference
  6. Destroy — remove and delete shared copies

Ask the right question and understand the objective.
Don't worry about the right answer. Enjoy exploring the data to find the unknown insight.

References