Designing from experience: on the adoption of chatbot conversational design practices by practitioners

One of the essential aspects of chatbot management is user retention, which is a consequence of good service provision. In order to reach this goal, there needs to be a well-thought-out conversational design. This process involves everything from understanding the domain to designing and conceiving how the conversations between machine and user can develop, aiming for greater acceptance of the chatbot through the most spontaneous and natural dialog possible. Related literature on general design revealed that personal experiences affect software design decisions. This work aims to evaluate how participants’ built-up perceptions and experiences with chatbots and their previous experiences assuming specific positions in a software development team influence their adoption of conversational guidelines proposed by a guide. Invited practitioners were asked to develop conversations without and with the conversational guidelines, and then they were interviewed about this process. We ran objective, narrative, and thematic analyses on interview transcripts and participants’ conversations. Most participants had negative experiences with chatbots, which greatly influenced how they viewed some guidelines. Furthermore, their experiences in specific software development positions also influenced their design and adoption of guidelines. Results revealed that personal experience is a significant moderating variable in conversational design. This work’s main contribution is providing knowledge on how practitioners enter the world of chatbot design and what concerns should exist in preparing introductory material considering their personal experiences.