JultikaUniversity of Oulu repository

Abstract

The service industry keeps growing these years. Artificial intelligence (AI) has started to be used in the service industry gradually, and the service chatbot is an excellent example of this phenomenon. Many giants have applied chatbots to handle their consumer services, such as LATTJO from IKEA, Stylebot from Nike, and Siri from Apple.

Understanding the advanced chatbot service experiences can help companies to optimize their chatbot services and improve their consumers’ satisfaction, which can bring them positive word-of-mouth, customer loyalty, re-purchase behavior, etc. However, chatbot services is an edge research area with limited studies about it. Thus, having the most advanced understanding of chatbot service experiences becomes particularly important. This study intends to fill this gap from chatbot service encounters’ perspective by understanding consumers’ satisfactory and unsatisfactory experiences with chatbots.

Due to this study focuses on chatbot service encounters and online customer service experiences, a qualitative research method be applied because it enables data to be explainable and justifiable. Data collection methods consist of the critical incident technique (CIT) and the online focus group. In the end, 22 validity incidents were collected.

Through data analysis, the author developed an incident sorting process and concluded eight types of chatbot service encounters within three groups by this process. The three groups are chatbot response to after-sales services, chatbot response to consumers’ needs, and unprompted chatbot actions. Moreover, 16 sources of different types of chatbot service encounters were found. Based on all the findings stated above, this study created an integrated framework for chatbot service encounters in online customer service experiences.

In conclusion, this study develops theoretical contributions by developing the integrated framework, creating an incident sorting process, and finding the sources for different service encounters. Based on these findings, this study also provides some managerial implications that companies could use to manage their chatbot services.