By Cardarello Fierro, M. & Canielli Scibelli, W.
Advanced Microeconomics final project
Marketing actions have been adopted as the main strategy that firms have to incentive the propensity of buying of their consumers. This paper introduces uplift modelling as a novel approach to estimate the causal effect of a marketing treatment, which facilitates targeting actions to responsive customers and efficient allocation of marketing resources. Uplift modelling has received increasing interest in the data analytics community as an improved framework for predictive analytics for data-driven decisions. A simulated dataset of a promotion campaign with two kinds of discounts was used to assess the impact on incremental sales, comparing the purchase uplift score between them, and to highlight the advantages of this approach over response models that only estimate the net buying propensity of customers.