Broken Aggregates and Messy Media Information

Learning Objectives

Align data from different granularities and temporal formats

Transform media data from long to wide format for model readiness

Handle missing values while preserving business meaning

Ensure proper joins between media and sales data to avoid leakage or misalignment

Overview

You're part of GlobalMart’s analytics team building a Market Mix Modeling (MMM) pipeline to analyze how media investments drive sales. However, the preprocessing steps taken by the team have led to misaligned data, wrong aggregations, and an unpivoted mess — severely affecting model results.

Your job: Debug these upstream issues before modeling begins.

Prerequisites

Time series granularity (daily, weekly, monthly)
Experience with Pandas operations
Understanding of MMM data structures