Review Article

A Study on the Improvement Direction of Artificial Intelligence Speakers Applying DeLone and McLean’s Information System Success Model

Table 2

Summary of analysis of existing research.

AuthorsMain used variableResults

Brill et al. (2019) [3](A) Perceived trust, information privacy
(D) User’s satisfaction
User value of Siri and Alex (+)
Lee et al. (2019) [4](A) Hedonic motivation, compatibility, perceived security
(D) User’s satisfaction
Both the antecedent variables have a positive (+) effect on the dependent variable which is satisfaction. However, negative (-) influence on intention to continue use
Hsieh and Lee (2021) [5](A) Perceived usefulness, perceived ease of use, and trust, anthropomorphic
(D) Continuous use
As for the antecedent variable, three factors had a positive (+) effect, but only anthropomorphic factors such as people had a negative (-) effect on the intention to continue use
Farah et al. (2020) [6](A) Affective experience with brands, impulsiveness, word of mouth, continued interaction with Amazon’s purchasing solutions
(D) Continuous purchase
If the existing impulsive buyer used the smart speaker, the impulsive purchase continues. However, market share is not sustainable
Han and Yang (2020) [7](A) Device design, user interface, personal information, and security
(D) Intention to continue use
Among the antecedent variables, personal information and security have a negative (-) effect on the dependent variable
Lee (2020) [8](A) IoT is vulnerable to personal information and security
(D) Stability
The risk of privacy infringement of IoT affects the stability (-)
Lee and Choi (2017) [9](A) Self-disclosure and reciprocity, intimacy, trust, and interactional enjoyment
(D) Satisfaction, continuous use
Among the antecedent variables, only intimacy has a negative (-) effect on satisfaction
Lin et al. (2019) [10](A) Social impact, pleasure motive, anthropomorphism, performance, and effort expectation, emotions
(D) User’s acceptance intention
Although anthropomorphism had a positive (+) effect on the acceptance intention dependent variable, it is difficult to expect a level like that of humans
Mamonov and Koufaris (2020) [11, 12](A) Techno-coolness includes factors of modern, futuristic, advanced feel, and fun
(D) Perceived enjoyment of information, services, and systems
Techno-coolness has a positive (+) effect on the intention to accept technology in smart home service.
Mulcahy et al. (2020) [13](A) Optimism, innovation, discomfort, and insecurity
(D) Perceived risk and trust, user’s adoption intention
Discomfort and insecurity negatively affect user adoption intention
Nguyen et al. (2019) [14](A) Voice user interface (VUI) information quality, system quality, information satisfaction, and system satisfaction
Perceived usefulness, perceived ease of use, mobile self-efficacy, trust, perceived enjoyment, and perceived risk
(D) Attitude toward VUI use, VUI use continuance intention
Mobile self-efficacy and trust influence negative (-)
Park et al. (2018) [15](A) Perceived usefulness, perceived ease of use, perceived enjoyment, and perceived privacy risks
(D) Intention to adopt smart speakers
Perceived ease of use negatively (-) affects.
Intention to adopt smart speakers
Roy et al. (2019) [16](A) Aesthetics, superior functionality, social presence, perceived interactivity, and perceived personalization
(D) Smart experience cocreation, consumers’ service brand equity, and word of mouth intentions
All antecedent variables have a positive (+) effect on smart consumer experience cocreation, brand equity, and word of mouth
Shin et al. (2018) [17](A) Compatibility, perceived ease of use, and perceived usefulness
(D) Purchase intention
For smart home/product, both antecedent variables have a positive (+) effect on purchase intention
Wu et al. (2017) [18](A) Interaction style (friend-like and engineer-like communication)
(D) Consumers’ brand perception
Friend-like interaction style has a greater influence on brand intimacy. User’s brand intimacy is mediated by brand warmth and competence
Flavian et al. (2021) [19](A) Technology readiness (optimism, innovativeness, discomfort, insecurity, and awareness)
(D) Intention to use AI (robo-advisors)
Innovativeness and discomfort had a negative (-) effect on the dependent variable
Vitezić &Perić (2021) [20](A) Anthropomorphism, hedonic, social influence, performance expectancy, effort expectancy
(D) Emotion, willingness to use AI devices
Only anthropomorphism has a negative (-) effect on the dependent variable
Belanche et al. (2020) [21](A) Humanoid anthropomorphism
(D) Intention to use humanoid service robot
In a group with high-risk aversion among customers, if a humanoid feels like a machine rather than a human, the intention to use it is low
Gaur et al. (2021) [22](A) Artificial intelligence and robotics, cleanliness and hygiene, health care and wellness
(D) Research direction of AI robotics
By integrating AI robots to secure customer trust in the hotel industry, an alternative means to a pandemic such as COVID-19
Gaur et al. (2021) [22]&Romero and Lado (2021) [23](A) Health history, health importance, perceived susceptibility, prevention efficacy, anthropomorphism, and social presence
(D) Attitude, booking intentions
Antecedent variables A positive (+) effect on the dependent variable.
Anthropomorphism positively affects prevention efficacy and attitude

(A): antecedent variable; (D): dependent variable.