Course contentsShow
← Data Science
Lesson 1171 of 2,14530. Exploratory Data Analysis (EDA) ProcessFree lesson

Creating an Analysis Plan

Outline the key questions, required data, and analytical steps needed to address the business problem.

Creating an Analysis Plan

What you'll learn: How to translate a business problem into a structured, step-by-step analysis roadmap that guides your exploratory data analysis.

Why You Need a Plan

Now that you've defined your business question, scoped your analysis, and clarified constraints, it's time to create a concrete plan. Think of this as your analytical roadmap—a document that answers three critical questions before you dive into code: What questions must I answer? What data do I need? What steps will I take?

Without a plan, EDA becomes aimless wandering. With one, you work efficiently toward actionable insights.

The Three Components of an Analysis Plan

1. Key Questions

List the specific questions your analysis must answer. If your business problem is "Why are sales declining?", your key questions might be:

  • Has decline occurred across all regions or specific ones?
  • Are certain product categories affected more than others?
  • Does the decline correlate with seasonal patterns or external events?

2. Required Data

Identify exactly what data you need and where it lives. For the sales example:

  • Sales transactions (database table orders)
  • Product categories (reference table products)
  • Regional data (column region in customers)
  • Date ranges (past 24 months for trend analysis)

3. Analytical Steps

Outline your approach in ordered steps:

  1. Load and validate sales data
  2. Check for missing values and outliers
  3. Aggregate sales by region, category, and time period
  4. Visualize trends over time
  5. Calculate percentage changes
  6. Compare against previous year

This isn't code—it's a logical sequence you'll translate into Python later.

Key Takeaway: An analysis plan transforms vague exploration into purposeful investigation by documenting what questions to answer, what data you'll use, and what analytical steps you'll follow—acting as your guide before writing a single line of code.