2. Build a clean marketing dataset
Covers the rows, columns, time stamps, IDs, and grain choices behind campaign, customer, sales, and media data. You practice spotting leaks, duplicates, missing values, and mismatched time windows before modeling begins.