Voice Assistant vs. Chatbot – Examining the Fit Between Conversational Agents’ Interaction Modalities and Information Search Tasks

Abstract

Owing to technological advancements in artificial intelligence, voice assistants (VAs) offer speech as a new interaction modality. Compared to text-based interaction, speech is natural and intuitive, which is why companies use VAs in customer service. However, we do not yet know for which kinds of tasks speech is beneficial. Drawing on task-technology fit theory, we present a research model to examine the applicability of VAs to different tasks. To test this model, we conducted a laboratory experiment with 116 participants who had to complete an information search task with a VA or a chatbot. The results show that speech exhibits higher perceived efficiency, lower cognitive effort, higher enjoyment, and higher service satisfaction than text-based interaction. We also find that these effects depend on the task’s goal-directedness. These findings extend task-technology fit theory to customers’ choice of interaction modalities and inform practitioners about the use of VAs for information search tasks.

Voice bots on the frontline: Voice-based interfaces enhance flow-like consumer experiences & boost service outcomes

“…In contrast, prior work focusing on voice as an output examined predominantly questions related to how much people trust a recommendation that is delivered verbally (vs. in written form) by a voice-based interface (Qiu & Benbesat, 2005). Only few studies examined a fully conversational interaction that required both the voice input by the user and a verbal response by the system (see Rzepka et al, 2021). Second, the type of boundary conditions typically examined either appearance-related factors (e.g., predominantly the presence of avatars; see Hess et al, 2009;Qiu & Benbasat, 2009) or broad consumer characteristics (e.g., consumers' general privacy concerns or emotional attachment tendencies; see Pagani et al, 2019;Son & Oh, 2018).…”

“…Second, the type of boundary conditions typically examined either appearance-related factors (e.g., predominantly the presence of avatars; see Hess et al, 2009;Qiu & Benbasat, 2009) or broad consumer characteristics (e.g., consumers' general privacy concerns or emotional attachment tendencies; see Pagani et al, 2019;Son & Oh, 2018). Except for emerging work examining the task characteristics as possible boundary conditions (Rzepka et al, 2021), no prior work we are aware of examined the unique conversational design characteristics during an interaction such as the specific semantic properties of the language or how the conversation itself is structured. Third, the type of outcomes studied in earlier work focused predominantly on perception outcomes (such as trust or satisfaction measures; Hess et al, 2009;Pagani et al, 2019;Qiu & Benbasat, 2009;Qiu & Benbesat, 2005) as opposed to the evaluation of the firm or downstream behavioral outcomes (see Son & Oh, 2018 for an important exception).…”

“…Third, the type of outcomes studied in earlier work focused predominantly on perception outcomes (such as trust or satisfaction measures; Hess et al, 2009;Pagani et al, 2019;Qiu & Benbasat, 2009;Qiu & Benbesat, 2005) as opposed to the evaluation of the firm or downstream behavioral outcomes (see Son & Oh, 2018 for an important exception). Finally, the type of mechanisms examined in earlier work illuminated important processes such as the extent of perceived social presence (Hess et al, 2009;Qiu & Benbasat, 2009) as well as fundamental cognitive processes (such as the extent of deliberation or how efficient the process was perceived overall; Berger et al, 2021;Rzepka et al, 2021). However, whether affordances unique to a voice-based interaction also change how consumers experience a service interaction, such as whether the interaction is perceived as more immersive or absorptive, and how these effects in turn shape downstream firm-related behavior is unclear.…”

“…As summarized in the review of Table 1, the current work examines a fully conversational service interaction experience (voice both as input and output), examines key Our literature review revealed that voice-based interfaces have been associated with a more intuitive, positively valenced experience (Rzepka et al, 2021) that can impact preference construction (Klesse et al, 2015) and the type of content that consumers reveal about themselves . These effects raise the question about the underlying properties that make speaking versus writing truly unique.…”

How do you Converse with an Analytical Chatbot? Revisiting Gricean Maxims for Designing Analytical Conversational Behavior

“…Text and voice interaction modalities are particularly relevant to our work. A comparative study of voice and text-based interaction with chatbots [57] found that voice was generally preferred in terms of cognitive effort, enjoyment, efficiency, and satisfaction, but this was influenced by goal-directedness of the task.…”