Causal Inference in A/B Testing : In collaboration with Euroconsumers

Details

Supervisors

Caelen, Olivier

Faculty

Faculté des sciences

Degree label

Master [120] en science des données, orientation statistique, à finalité spécialisée

Abstract

The thesis was conducted in collaboration with Eurroconsumers. The company provided data from a marketing campaign designed to evalutate the impact of financial incentives on the client retention. The campaign was divided in different randomized group: one serving as a control baseline and receive nothing, while two treatment groups received an email containing a 30-euro voucher or a 50% discount depending on their assigned group. The main objective of the thesis is to quantify the causal effect of using the incentive on the expected retention time of the clients. To quantify this effect two complementary approaches are developed: the Random Controlled Trial framework, and the Causal Machine Learning framework.