Research: Evaluating Incrementality in Digital Marketing Spend Using Marketing-Mix Mode...

Research question: How accurately can marketing-mix modelling combined with incrementality measurement distinguish the true causal impact of digital advertising spend on sales outcomes?

Background & Motivation: With the rise of digital channels, advertisers increasingly invest in digital marketing, demanding robust methods to quantify the causal impact of spend on sales and brand metrics. Traditional marketing-mix modelling (MMM) offers insights into channel effectiveness but often struggles to isolate incremental effects, especially in complex digital environments.

Research Gap: While MMM is widely used, its ability to measure incrementality—what would have happened without the digital spend—is limited, particularly given attribution challenges and confounding variables. Recent advances in causal inference and experimentation (e.g., geo-lift tests, synthetic controls) offer new ways to test and validate incrementality but have not been systematically integrated with MMM in real-world settings.

Approach & Expected Contribution: This thesis will critically review existing MMM and incrementality measurement methods, design an empirical study applying both to a real or simulated dataset (e.g., retail digital campaigns), and compare their effectiveness in distinguishing incremental versus non-incremental sales. The project will explore the integration of causal inference techniques (e.g., difference-in-differences, propensity score matching) with MMM and assess their impact on measurement accuracy.

Why It Matters: Accurate incrementality measurement is essential for optimizing digital marketing budgets and improving ROI, yet current practices may misattribute effects. This research aims to advance both academic understanding and practical application of marketing-mix and incrementality approaches, providing marketers with more reliable tools for decision-making.