Tactical use of rewards to enhance loyalty program effectiveness

Research output: Contribution to journalArticleResearchpeer-review

Abstract

A key member benefit for participating in a loyalty program (LP) is the rewards earned for points accrued. One popular reward structure is a catalog of many diverse items. The rewards among this broad selection are likely to differ in their appeal due to their intrinsic differences and customer heterogeneity. Prior research has shown that after redeeming a reward, LP members are more motivated to increase their purchase volume/frequency and share-of-wallet within the program, thereby becoming more active. In this study, we fit a hidden Markov model to a 4½ year longitudinal data set of points accrual and reward redemption activity for about 4500 members of a large coalition LP. Our analysis reveals three latent states — active, hyperactive and inactive. We then investigate the likelihood of LP members transitioning between these states across successive time periods, and examine the reward categories and marketing effort associated with these transitions. Subsequently, we use our model to optimally promote particular reward categories to encourage LP member migration to managerially desirable states or prevent them sliding into a less desirable state. Our proposed optimal reward strategy potentially increases the estimated proportion of LP members in the hyperactive latent state from 35.7% to 40.1%, with a resultant increase in sales revenue for retailers and service providers in the LP of 7.7%. We find that rewards which are more fungible have the strongest influence on increasing points accrual activity.

Original languageEnglish
Number of pages16
JournalInternational Journal of Research in Marketing
DOIs
Publication statusAccepted/In press - 2020

Keywords

  • CRM
  • Hidden Markov model
  • Loyalty programs
  • Multinomial logit model
  • Random effects

Cite this