Turning the wheels of engagement: Evidence from entertainment live streaming

Xiaofei Song, Mengyao Fu, Jie Fang, Zhao Cai, Chee Wee Tan, Eric Tze Kuan Lim, Alain Yee Loong Chong

Research output: Journal PublicationArticlepeer-review

Abstract

Influencers (e.g., live streamers) earn gratuities from engaging customers, a phenomenon that we label as engagement monetization. Although scholars have advocated a process view of customer engagement, they tend to accentuate unidirectional transitions from low- to high-engagement-level states. This constrains influencers’ understanding of customer engagement, which in turn inhibits their ability to both trigger profitable states of engagement as well as prevent deteriorating states of engagement from manifesting. To this end, we extend extant literature by conceptualizing customer engagement as an emergent process that embodies engagement states and engagement transitions. Empirically, we conduct an illustrative study in the context of entertainment live streaming and subscribe to the Markov chain method to showcase how customer engagement transition can be modeled. Based on 91,148 engagement records, we scrutinize the effects of influencers’ scheduling strategy on engagement transitions and viewers’ gratuities using the Multilevel Linear Model (MLM). Findings yield practical implications for influencers in terms of how live streaming sessions can be strategically scheduled to bolster customer engagement and monetization opportunities.

Original languageEnglish
JournalJournal of the Academy of Marketing Science
DOIs
Publication statusPublished - 16 May 2024

Keywords

  • Customer engagement
  • Emergent process
  • Engagement transition
  • Gratuity
  • Live streaming
  • Markov chain
  • Scheduling strategy

ASJC Scopus subject areas

  • Business and International Management
  • Economics and Econometrics
  • Marketing

Fingerprint

Dive into the research topics of 'Turning the wheels of engagement: Evidence from entertainment live streaming'. Together they form a unique fingerprint.

Cite this