The value creation of artificial intelligence customer service in E-self service

Hye Young Hah, Purdue University

Abstract

The purpose of this study is to conceptualize service value and usage intention in the context of Artificial Intelligence usage. The current research demonstrates two research objectives: (1) to investigate the relationships between perceived benefits/sacrifices and service value and (2) to explore the moderating role of consumer innovativeness in this Artificial Intelligence (AI) domain. Since service value has regarded as a domain specific construct, it is worthwhile to scrutinize the value of an innovative service online. To do so, conceptual framework from Cronin et al.’s value added model (1977) was introduced to measure the consumers’ perception of get and give during transactions in self service environment. To delineate those questions, this study utilizes Confirmatory Factor Analysis and Structural Equation Modeling with AMOS 17.0 to specify the unidimensionality of each construct and hypothesized paths in the research model. Throughout a paper-and-pencil survey in a Midwestern university, the results reflect that (1) each construct satisfied construct validity, (2) perceived usefulness and compatibility of innovative service delivery were significant indicators for service value creation, and (3)strong association between service value and usage intention in an online self service shopping environment. For the moderating role of consumer innovativeness, multigroup analysis with Chi-square difference test was introduced in the SEM structure. In each relationship of perceived benefits – service value, perceived sacrifices – service value, and service value – continued usage intention, there was no significant Chi-square difference to capture how innovative consumers can create more value differently than non-innovator groups. Finally, this study proposes future research to collect more consumer sample with a variety of age groups and reexamine this study. In a sense that relatively small sample size (n=215) may hinder sacrificial aspects of service value and moderating effect to be revealed, more large sample size will increase statistical power to catch the differences. In addition, for those consumers who are not familiar with this innovative Artificial Intelligence, experimental approach is suggested. Within an experiment setting, researchers can control other factors in order to capture the value perception and continued usage intention.

Degree

M.S.

Advisors

Park, Purdue University.

Subject Area

Marketing|Artificial intelligence

COinS