Timing customer reactivation initiatives

Abstract

Firms operating in non-contractual settings apply customer reactivation initiatives such as email messages to stimulate customers who have become inactive temporarily or permanently to resume their transaction activities. Thus, firms need to know which customers are inactive, and when a customer becomes inactive. Existing approaches struggle to distinguish active from inactive customers and do not provide time-scale estimates of when to send reactivation mails. To address these shortcomings, we develop an approach to target and time the sending of reactivation mails. Building on control chart methods, we introduce a gamma–gamma control chart, modelling the average customer interpurchase time and the variation therein to determine activity boundaries. Crossing these boundaries signals a potential change in a customer’s purchasing activity, providing a signal to initiate customer reactivation. A field experiment in the greetings and gifts industry, supported by several additional analyses, illustrates the improved performance of our approach when it comes to signaling customer activity against a wide range of competing models. The improved performance of our method occurs particularly in settings where customers vary strongly in purchase and inactivity patterns.