Randomized controlled trials determine the overall causal impact of a treatment or intervention program (e.g. marketing, medical, political, social, education). Uplift modeling takes a further step to identify individuals who are truly positively influenced by a treatment through statistical modeling or machine learning. This technique allows us to identify the "persuadables" and thus optimize target selection in order to maximize treatment benefits. This important area has gained significant attention in areas such as personalized marketing, personalized medicine, and political election with increasing publications and presentations in recent years from both industry practitioners and academics.
We will discuss multiple Uplift Modeling methodologies including some latest ones along with business applications. We will link predictive modeling to prescriptive analytics for a multiple treatment extension of Uplift Modeling. Examples from retail marketing and non-profit industries will be used to illustrate its application and methodologies. The same methodologies can be readily applied in other fields such as insurance, medicine, education, political, and social programs.