Objective / Business Problem
To identify and quantify the relationships between sales and diverse marketing activities, enabling the client to understand how different elements of the marketing mix collectively drive sales in a competitive market landscape.
Approach / Methodology:
-
Utilized the client’s transactional database incorporating retail audit data: volume sales, Weighted Average Retail Price (WARP), weighted distribution, and promotion volumes.
-
Integrated media data including Gross Rating Points (GRP) and Net GRP (NGRP) on a monthly basis for premium brands/SKUs.
- Observed complex, non-intuitive uni-variate relationships; applied domain expertise to engineer transformed variables such as Share of Voice (SOV), Adstock effects, and Price Index to better capture marketing dynamics.
- Built a multiplicative market mix model within an Empirical Bayesian framework to estimate the interactive effects of marketing variables on sales.
- Model performance evaluated using Mean Absolute Percentage Error (MAPE).
- Estimated price elasticity and quantified the contribution of each marketing activity to sales outcomes.
Outcomes and Impact:
-
Delivered actionable insights on the impact of both the client’s and competitors’ marketing activities on sales performance.
-
Empowered the client to optimize marketing budget allocation by focusing investment on high-impact activities.
-
Enabled maximization of Return on Investment (ROI) through evidence-based marketing planning and resource deployment.
Business Value:
This modeling approach enhanced the client’s ability to make data-driven marketing decisions, improving sales effectiveness in a competitive environment and driving greater financial efficiency.
