Gains: An R package for gains tables and lift charts

Abstract

Gains tables and lift charts are commonly used in direct marketing analytics to evaluate the performance of a predictive model on a target sample. A gains table shows how the average (and cumulative average) response values in the sample change with the scores from the predictive model. These numbers can be used to estimate and maximise a model’s effectiveness for direct marketing campaigns. This paper introduces the R package known as ‘gains’: a free, publicly available tool to construct gains tables and lift charts. Instructions for installing and loading the package in R software are provided, and the package’s use is illustrated using real data from an apparel retailer’s e-mail marketing campaign. The use of gains tables raises important questions for model selection, and the paper concludes by discussing future research in this area.