JMIR Publications

ABSTRACT

Background:

Engagement with smartphone applications (apps) for smoking cessation tends to be low. Chatbots (i.e., software that enables conversations with users) offer a promising means of increasing engagement.

Objective:

We aimed to explore smokers’ experiences of a quick-response, AI-driven chatbot (‘Quit Coach’) implemented within a popular smoking cessation app and identify factors that influence users’ engagement with the Quit Coach.

Methods:

In-depth, one-to-one, semi-structured qualitative interviews were conducted with adult, past-year smokers who had voluntarily used the Quit Coach in a recent smoking cessation attempt (n=5) or current smokers who agreed to download and use the Quit Coach for a minimum of two weeks to support a new cessation attempt (n=9). Verbal reports were audio recorded, transcribed verbatim, and analysed within a constructivist theoretical framework using inductive thematic analysis.

Results:

Three higher-order themes were generated to capture users’ experiences of and engagement with the Quit Coach: 1) anthropomorphism of and accountability to the Quit Coach (i.e., users ascribing human-like characteristics and thoughts to the chatbot, which helped foster a sense of accountability to it); 2) the Quit Coach’s interaction style and format (e.g., a positive and motivational tone of voice, quick and easy-to-complete check-ins); and iii) users’ perceived need for support (i.e., chatbot engagement was motivated by seeking distraction from cravings or support to maintain motivation to stay quit).

Conclusions:

Anthropomorphism of a quick-response, AI-driven chatbot implemented within a popular smoking cessation app was enabled by its interaction style and format and users’ perceived need for support, which led to feelings of accountability and increased engagement.