AI-powered chatbot communication with customers: Dialogic interactions, satisfaction, engagement, and customer behavior

Hua Jiang, Yang Cheng, Jeongwon Yang, Shanbing Gao

Research output: Contribution to journalArticlepeer-review

119 Scopus citations

Abstract

The present study is grounded in social exchange theory and resource exchange theory. By exploring customers' satisfaction with chatbot services and their social media engagement, it examined the effects of responsiveness and a conversational tone in dialogic chatbot communication on customers. To test the proposed mediation model, we surveyed a representative sample of customers (N = 965) living in the U.S. After examining the validity and reliability of our measurement model, we tested the hypothesized model using structural equation modeling (SEM) procedures. All proposed hypotheses were supported, indicating the significant direct effects of (1) responsiveness and a conversational tone on customers' satisfaction with chatbot services, (2) customers' chatbot use satisfaction on social media engagement, (3) customers’ social media engagement on price premium and purchase intention, and (4) purchase intention on price premium. In addition, we examined satisfaction, social media engagement, and purchase intention as significant mediators in the proposed model. Theoretical and practical implications of the study were then discussed.

Original languageEnglish (US)
Article number107329
JournalComputers in Human Behavior
Volume134
DOIs
StatePublished - Sep 2022

Keywords

  • Chatbot use satisfaction
  • Dialogic chatbot communication
  • Engagement-facilitating AI technologies
  • Price premium
  • Purchase intention
  • Social media engagement

ASJC Scopus subject areas

  • Arts and Humanities (miscellaneous)
  • Human-Computer Interaction
  • General Psychology

Fingerprint

Dive into the research topics of 'AI-powered chatbot communication with customers: Dialogic interactions, satisfaction, engagement, and customer behavior'. Together they form a unique fingerprint.

Cite this