Journal of Retailing and Consumer Services 82 (2025) 104087

Contents lists available at ScienceDirect

Journal of Retailing and Consumer Services

journal homepage: www.elsevier.com/locate/jretconser

AI assistant is my new best friend! Role of emotional disclosure,

performance expectations and intention to reuse
Sana Affandi a , Muhammad Ishtiaq Ishaq b , Ali Raza c,* , Qurat-ul-ain Talpur d, Rehan Ahmad e
a
School of Management Sciences, Quaid-i-Azam University, Islamabad, Pakistan
b
Léonard de Vinci Pôle Universitaire, Research Center, 92916, Paris, La Défense, France
c
Excelia Business School, CERIIM, France
d
ESSCA School of Management, France
e
University of the Punjab Lahore, Pakistan

A R T I C L E I N F O A B S T R A C T

Keywords: In this modern era, high-tech companies are launching their AI assistants, considering their role in shaping
AI assistant consumer behavior. This research sheds light by building a conceptual framework using dual process theory to
Immersion determine the impact of AI assistant advantage on user engagement and user emotional disclosure, subsequently
Emotional disclosure
influencing the intention to reuse AI with the moderating role of performance expectation. Using a random
Performance expectations
sampling method, a structured questionnaire was used for data collection from 644 consumers. The findings
User experience
show that AI assistant advantages positively impact user engagement, and user emotional disclosure increases
the intention to reuse AI assistants. This impact of user engagement and user emotional disclosure, along with
performance expectations, maximizes the reuse intention of AI. Managers and marketers in the AI environment
can insinuate the study methods to increase the intention of reuse and may transform their marketing strategies
to promote their businesses using AI more effectively.

1. Introduction this, it is crucial to identify the specific needs of the users (Maduku et al.,
2023). The perception of users regarding AI assistants in the marketing
With the innovation and massive development of AI, many devices field has been identified in the literature, which points out that the ca­
that support AIhave attained computer-human interaction, expanding pabilities of AI chatbots in data analytics and meaningful customized
via personalization, efficient responses, and various other functions in solutions impact the interaction process (Pantano and Pizzi, 2020).
the area of service delivery (Pelau et al., 2021; Moore et al., 2022; Humans process in two ways – cognitive and emotional under the dual
Saleem et al., 2024). In the modern era, many companies that are big process theory (Stanovich and West, 2000). Considering such aspects, it
giants in high-tech industries have launched numerous virtual assistants is crucial to study user engagement and emotional disclosure related to
such as Xiaodu, SISI, and Bixby, etc., that bid many advantages of AI the product derived from utilitarian and hedonic perspectives (Yuan
technology, which include entertainment in terms of audio-visual ac­ et al., 2021). Previous studies have explored users’ perceptions
cess, life benefits, and information queries (Guha et al., 2021; Mehmood regarding AI assistants based on the utilitarian perspective in marketing
et al., 2024). The global market is expected to touch eight billion dollars (Gursoy et al., 2019). AI chatbots’ data analytics capability affects
due to a surge in the use of AI assistants by retail and service businesses interaction through meaningful customized solutions (Pantano and
(Transparency Market Research, 2016). Therefore, such an advantage Pizzi, 2020). The perceived usefulness of AI devices has been focused on
forced consumers to adopt an immersive reality world using AI assistants and identified as the strongest antecedent of the purchases made by
in their daily lives (Xie et al., 2024). consumers in the field of fashion (Liang et al., 2020; Arachchi and
Using AI in the consumer decision-making process is not a temporary Samarasinghe, 2023).
attraction; instead, it has an upward trend, as depicted by multiple Marketers focus on products based on AI to identify willingness to
studies on AI usage (Kunduru, 2023; Gesk and Leyer, 2022). Based on accept such products and intention to reuse among consumers (Luo

* Corresponding author.
E-mail addresses: sanaaffandi@gmail.com (S. Affandi), ishaq.muhammadishtiaq@gmail.com (M.I. Ishaq), alirazaphd@hotmail.com (A. Raza), qurat-ul-
aintalpur@essca.fr (Q.-u.-a. Talpur), rahmad05@gmail.com (R. Ahmad).

https://doi.org/10.1016/j.jretconser.2024.104087
Received 30 July 2024; Received in revised form 6 September 2024; Accepted 13 September 2024
Available online 27 September 2024
0969-6989/© 2024 Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
S. Affandi et al. Journal of Retailing and Consumer Services 82 (2025) 104087

et al., 2019; Pillai et al., 2020). The underlying reason is that the suggests that unilaterally studying emotional and cognitive factors may
intention to reuse AI assistants is due to the consumers’ actual behaviors, leave unexplained valuable and meaningful differences (Sierra and
which are critical for organizations to carefully identify pre-purchase Hyman, 2011; Dewey, 2023). So, it is crucial to understand the aspects
and post-purchase marketing strategies (Gursoy et al., 2019). The ave­ from multiple angles. Furthermore, this theory is ideal to use that ex­
nues for future studies include the development of research models that plains the modeling for consumers’ engagement and their emotional
provide a holistic view for identifying and understanding the impact of disclosure, which are aspects of emotional and cognitive factors used as
AI on human life (Maduku et al., 2023), incorporating the variables of antecedents for exploring the user’s decision-making aspects, including
this study. The recent literature shows that the consumers’ perceived the intention to reuse AI assistants. Hence, it is important to understand
value is a crucial factor in adopting the new technologies as it imitates the user’s demand orientation and the advantages of AI assistants to
the product’s performance according to the perception of users (Lee formulate efficient and effective marketing strategies for performance
et al., 2022b; Jan et al., 2023), but they have been unable to explore it in improvement. This research suggests how AI assistant advantages affect
recent literature. Lastly, the existing literature used a single feature of AI user engagement and emotional disclosure, subsequently influencing
assistant, i.e., anthropomorphism, on which the researchers raised their the intention to reuse AI assistants. The conceptual framework is pre­
concerns about its generalizability (Xie et al., 2024; Liu et al., 2024; sented in Fig. 1.
Saleem et al., 2024).
Consequently, to fill the gap in the available literature, this research 2.2. Characteristics of AI assistants
study inspects the antecedents and precedents of user engagement and
user emotional disclosure towards AI assistants, including the advan­ AI assistants are viewed as innovative items and part of consumers’
tages of AI assistants for the former intention to reuse, respectively. novel experiences, while some users satisfy their curiosity and enjoy
Performance expectations have been incorporated as moderators in the communication with AI assistants (Gursoy et al., 2019; Hasan et al.,
study. With the interest of verifying the influence of marketing efforts 2021). AI technology powers the latest AI assistants, which are intelli­
that AI assistant has provided, measurement has been developed to gent products, and users need not take the initiative when using them as
determine the responsiveness, accuracy, interaction, affinity, and AI assistants have more humanlike methods of communication and ways
compatibility which affect user engagement and user emotional disclo­ to build boundaries effectively (Chakraborty et al., 2024; Saleem et al.,
sure with the moderating influence of performance expectation conse­ 2024). We used five characteristics of AI assistants, namely accuracy,
quently have an impact on the intention to reuse AI assistant in the high- responsiveness, compatibility, anthropomorphism, and affinity. Accu­
tech service context. User engagement, emotional disclosure, perfor­ racy refers to the correctness, desirability, and precision of services to
mance expectation, and intention to reuse are the variables that provide the users related to AI assistants (Chung et al., 2020). Such accurate
the key insights to the brands in identifying and fulfilling consumer services make users feel that demands are significant, and the solutions
needs. provided by such technology are ancillary to the best attempts at solving
This research study has significant theoretical and practical contri­ the problems (Cheng and Jiang, 2022). These aspects help promote the
butions. Based on dual process theory, this study provides an improved degree of user engagement and emotional disclosure, which subse­
theoretical framework for AI assistants’ acceptance and reuse intention. quently influence the intention to reuse and willingness to accept AI.
The theme of this research has been developed by addressing the recent Chen et al. (2021) refer to responsiveness as the degree to which
call for papers in modern marketing. This study builds an improved AI consumers get immediate information through AI assistants. The ability
technology model that covers various user aspects. The results of this of AI assistants to deal with problems and answer the requests posed by
research study provide insights into the user’s cognitive and emotional users refers to responsiveness. This is driven by convenience, which
elements that are more attractive to the innovative capabilities of the AI provides effective and efficient solutions (Pihlströ;m and Brush, 2008)
assistant for its intention to be reused in the future. In addition, such that affect user engagement and emotional disclosure. Compatibility is
supplementation of the antecedents of user engagement and user related to the consumers’ perception of the current experience, needs
emotional disclosure provide the consequences for the acceptance of AI fulfilment, and innovativeness in using products/services (Zolkepli and
technology and the intention to use the technology repeatedly. Kamarulzaman, 2015). AI provides users customized services, and
The structure of the paper is organized as follows. The next section compatibility is crucial to technology use (Chiang, 2013).
contains the up-to-date literature review on the variables and a dedi­ AI users are significantly affected by anthropomorphism (Ashfaq
cated sub-section on theory and hypotheses development. Section 3 et al., 2020). Self-consciousness, human appearances, and feelings are
outlines the sample details, sampling strategy, data collection and key attributes of anthropomorphism (Sheehan et al., 2020). AI assistants
measuring instruments. Section 4 shows the results on reliability and have humanoid imagery and have subsequent interactions based on
validity, descriptive statistics, and hypotheses testing using structural users’ attributing features, behaviors, or emotions, so human beings are
equation modeling. Lastly, Section 5 contains a discussion of results prone to anthropomorphized objects (Pelau et al., 2021). Consumers
along with theoretical and managerial implications. recognize their strong relationship with AI assistants, making them feel
connected and close to such assistants via the pleasure they seek from
2. Literature review intelligent communication (Sung et al., 2021). A humanlike manner of
interaction among AI assistants is very courteous and cordial (Moriuchi,
2.1. Dual process theory 2021). Such interactions help consumers forget about social distancing,
enhance the pleasure of user experience, and instill the user’s knowledge
According to the framework of dual process theory (Stanovich and in an amiable environment (Sheehan et al., 2020). So, it is critical to
West, 2000), there are two paths to the process of human cognition: deliberate anthropomorphism in the current research framework.
system one and system two. The irrational part consists of the emotional Lastly, affinity refers to the communication between the consumer and
responses that relate to the human brain, which processes information an AI assistant with minimum distance and a humanlike tone that cre­
automatically, impulsively, and intuitively; this is called system one. ates relaxed and pleasant interactions (Yuan et al., 2022).
Rational is system two, composed of cognitive perceptions that process Various aspects of technical support through an interactive interface
the information thoughtfully and reasonably (Yuan et al., 2021). This are provided by an application of AI assistant (Pitardi and Marriott,
dual process contextualization describes the consumer decision pro­ 2021b). According to Han and Yang (2018), various forms are depicted
cesses using cognitive and emotional factors. via installation on hardware, such as watches and speakers, or
This theory believes that emotional and cognitive factors are essen­ embedded in operating systems. With significant technological ad­
tial in influencing decision-making at a personal level. This theory vancements, AI assistants now recognize human feelings and optimize

2
S. Affandi et al. Journal of Retailing and Consumer Services 82 (2025) 104087

Fig. 1. Research model of the study.

themselves for user engagement that goes beyond the comfortable. Engagement experience is a critical consideration when the
information-oriented and utilitarian functions (Tsai et al., 2021). Huang level of intensity is seen in this (Alexander et al., 2018; Hollebeek et al.,
and Rust (2021) claimed that technology helps consumers learn about 2019). The type of interaction, which includes affective cognitive and
existing brands, which has key implications for organizations to engage behavioral aspects, which are the advantages of AI assistants and social
them with the brands. The literature showed that technology (such as influence, affects anticipated user engagement, affecting the intention to
voice assistants – Alexa, Siri, and Google) engages consumers with the reuse (Letheren et al., 2019; Saleem et al., 2024). Engagement has been
brands (Yuan et al., 2022). Therefore, we propose that. viewed as the state of mind of an individual or disposition, which is
reflected by the level of engagement of a user with a particular focal
H1. AI assistants positively influence user engagement.
object (Storbacka et al., 2016). Studies have added that various inter­
Emotional disclosure is another key aspect for consumers to use active technologies, such as AI assistants, require the engagement of
AIassistants (Skjuve et al., 2021), as it increases user engagement and users to utilize multiple services. The advantages of AI assistants, such as
experience and provides emotional support, leading to user satisfaction anthropomorphism, have been identified as those that can create posi­
(Adam et al., 2021). Using the computer-as-a-social actor paradigm, AI tive outcomes for user engagement and connectedness with them (Tam
assistants are considered entertainers, fostering emotional bonds, stim­ et al., 2013).
ulating consumer interaction with the technology (Rhim et al., 2022; In the presence of nonhuman agents, consumers are oriented toward
Klein and Martinez, 2023), and increasing their reliability and usability engagement with machine learning AI and robots (Yuan et al., 2022).
(Jenneboer et al., 2022). Epley et al. (2008) concluded that anthropo­ Based on social cues, consumers automatically respond to
morphism can be activated through emotional aspects such as emotional technological-oriented agents as mindless, friendly, or social processes.
disclosure and social dialogue. Additionally, Araujo (2018) claims When talking to an AI, consumers tend to make tangible interactions in
anthropomorphism is critical in human-AI assistant interactions, while conversation (Nass and Moon, 2000). Social interactions tend to be
Huh et al. (2023) suggested that AI assistants provide emotional con­ increased when nonhuman agents exhibit human traits. There is more
nections. Therefore, we propose that. engagement when consumers interact with nonhuman agents naturally
(Rosenthal-Von Der Pütten and Krämer, 2014). Language style, a lin­
H2. AI assistants positively influence user emotional disclosure.
guistic cue, is one of the human traits shown by nonhuman agents that
AI can solve personalized and complex problems through data pro­ affect anthropomorphic perceptions. This is the case for other AI assis­
cessing and algorithm capabilities (Tsai et al., 2021; Guha et al., 2023). tant advantages that could influence how humans identify themselves
DeLone and McLean (2003) designed a framework which suggested the with AI (Verhagen and Wagenmakers, 2014; Araujo, 2018; Xu and
criteria for user evaluation for the technical information system and Lombard, 2017).
illustrated the high-performance demands of the users concerning User engagement is a context-dependent consumer behavior with
compatibility, accuracy, and responsiveness. Accurate information and varied intensity levels (Brodie et al., 2013) that goes beyond participa­
trustworthiness induce users’ reliance on the AI assistant and enhance tion and involvement (Hari et al., 2022). User engagement is among the
positive attitudes related to the users’ perceptions that make users use AI strong influencers of loyalty (So et al., 2016) and psychological and
assistants when in need and face problems repeatedly (Kumar et al., behavioral intentions such as repurchasing a brand (Nazir et al., 2023).
2024; Cheng and Jiang, 2022). Responsiveness amongst the leading Brodie et al. (2013) argued that a genuinely engaged consumer has a
drivers that provide accurate and timely service increases the overall strong psychological association (with AI assistants) that cannot be
evaluation of the utility of products (Pelau et al., 2021; Kim and Baek, measured through behavioral intentions. AI assistants are designed to
2018). Consistency in the user experience and the technology is ensured take users’ queries and provide solutions, leading to future reuse
by compatibility, and it is generally regarded as the factor with the intention (Nazir et al., 2023). The literature confirmed that the AI as­
highest correlation with the personalization aspect of the user that sistants’ engagement with users affects their intention to reuse and
drives user adoption of modern AI (Zolkepli and Kamarulzaman, 2015). willingness to accept AI, which are the cognitive and emotional con­
nections with products or brands (Kamoonpuri and Sengar, 2023). Other
H3. AI assistants positively influence intention to reuse among
features of AI assistants also increase user engagement, which has an
consumers.
impact on the intention to reuse, so the following hypotheses are
formed.
2.3. User engagement as a mediator
H4. User engagement has a positive influence on intention to reuse.
The worth of emotions may escalate the desire for increased in­
H5. User Engagement positively mediates the relationship between AI
teractions, leading to more engagement from the users (Babin et al.,
assistant advantages and the intention to reuse AI.
1998). The current literature demonstrated that the drastic technolog­
ical changes allow consumers to engage when and where they are

3
S. Affandi et al. Journal of Retailing and Consumer Services 82 (2025) 104087

2.4. User emotional disclosure as mediator engagement that results in intention to reuse AI. Performance expec­
tancy, along with social influence, positively affects user engagement
Under the framework of emotional disclosure, AI assistants provide and user emotional disclosure and promotes emotional boundaries,
advantages similar to those of human beings. Ho et al. (2018) claimed which increases the willingness of users to accept innovative technolo­
that consumers feel comfortable with virtual assistants and disclose gies and the intention to reuse them (Gursoy et al., 2019; Guo and Luo,
personal details due to past positive experiences. Subsequently, users 2023). Along with the theory of dual process, as explained above, it is
experience more positive outcomes (Lee et al., 2022b). People with argued that the advantages of AI assistants must be explored and
negative judgment or social influence prevent individuals from analyzed expansively, which influences engagement and emotional
disclosing deeply to others. So, it is easier for people to disclose the disclosure of the user. Thus, the following hypotheses are formed.
chatbots than for human beings. This refrains individuals from worrying
H8. Performance expectation has a positive influence on intention to
about being burdened, rejected, or the listener and restrains disclosing
reuse.
to people seeking the potential benefits (Ho et al., 2018). As individuals
know, AI assistants have no judgments and perceptions about the users; H9. Performance expectation positively moderates the relationship
hence, disclosure intimacy exists that reduces the management of im­ between user engagement and intention to reuse AI.
pressions in situations where fears of negative evaluation are obvious
H10. Performance expectation positively moderates the relationship
(Lucas et al., 2014). If this happens in all situations, not just the negative
between user emotional disclosure and intention to reuse AI.
ones, users may disclose more confidentially to chatbots than human
beings (Ho et al., 2018).
3. Methodology
The more intimate disclosures are made to AI assistants, the greater
the psychological benefits that can accrue, resulting in increased
3.1. Sample and data collection
intention to reuse AI in everyday life. This also creates room for per­
formance expectations that affect future intentions and willingness to
We formulate the list of the most used and renowned AI assistants:
use (Ho et al., 2018). The cognitive processing model describes that the
Siri, Amazon Alexa, Google Assistant, Bixby, and Cortona. We developed
critical component of the link between beneficial outcomes resulting
a questionnaire containing demographic questions, contact information,
from cognitive changes involves the process through which disclosing
consent to participate in this research, and the name of the AI assistant
the undisclosed aspects eliminates negative influence and induces
they used the most in October 2023–December 2023. The initial data
reappraisal (Pennebaker and Chung, 2007).
was collected from metropolitan cities of Pakistan. After gathering these
The existing literature is insufficient to analyze the significance of AI
responses, we randomly selected respondents and sent the self-
assistant characteristics as these assistants are adopting themselves with
administered questionnaire from January 2024 to March 2024, along
more humanlike traits (Sheehan et al., 2020) and can fulfill consumers’’
with a cover letter that contained the purpose of this research and
needs (social and emotional) (Song et al., 2022). Additionally, Cheng
ensured anonymous data collection. Considering the context of this
and Jiang (2022) claimed that consumers show greater acceptance due
research, extracting an accurate sample size has significant value. For
to humanlike responses. Choi and Drumwright (2021) argued that
instance, Comrey and Lee (1992) proposed that a sample size of 500 is
consumers use AI assistants not only for productivity reasons but also to
very good, whereas 1000+ respondents in a survey-based study are
motivate themselves with empathic and friendly relationships. This
considered excellent. Tinsley and Tinsley (1987) proposed that ten re­
provides new insights into such situations and eliminates cogitation,
spondents against each item are acceptable for the generalizability of the
which is bothersome and confusing. This results in increased acceptance
results. Similarly, Everitt (1975) suggested a 10 to 1 ratio, and Hair et al.
and reuse intentions among users who rely on performance expecta­
(2010) suggested a 20 to 1 ratio. Therefore, the minimum sample should
tions. Based on this, we propose.
be in the range between 350 and 700.
H6. User emotional disclosure has a positive influence on intention to In the first part of the questionnaire, we asked screening questions
reuse. related to the name of the AI assistant they used, followed by de­
mographics and scales of all variables. After a few reminders, we
H7. User emotional disclosure positively mediates the relationship
received 729 questionnaires, from which 23 questionnaires were deleted
between AI assistant advantages and intention to reuse.
due to incomplete responses, 29 responses were deleted due to extreme
value selection, and 33 responses were deleted after finding the differ­
2.5. Performance expectations as moderator
ence between the pre-data and post-data collection phases. Hence, we
correctly received 644 responses (response rate: 64%) and used them in
Performance expectation has been defined as the outcomes, results,
further analyses.
and behaviors expected from the AI assistant (Buschmeyer et al., 2023).
The demographic profile contains gender, education level, name of
Outlining performance expectations of AI is a complex and challenging
AI assistant, age, and monthly income. As shown, 56% of participants
task since such expectations diverge among applications and individuals
were male, and 44% were female. The age was measured on four options
(Shimizu et al., 2021). The available literature has emphasized the
– up to 25 years (24% of participants), 26–35 years (30% of partici­
requirement for a performance expectation framework to balance con­
pants), 36–45 years (28% of participants), and above 45 years (18% of
sumers’ performance expectations and AI systems’ increased accep­
participants). Most people use Google Assistant AI (23%) and Siri (23%),
tance, trust, and reuse intention (Kaur et al., 2022).
followed by Alexa (20%), Bixby (19%), and Cortona (15%). The edu­
Performance expectation fulfills the criteria of an individual’s desire
cation level was measured in three categories – undergraduate qualifi­
to remain in control and the ability to foretell regarding the surround­
cation (39% of participants), postgraduate qualification (35% of
ings. Preceding studies have concentrated on evaluating a service or
participants), and diploma/certification (26% of participants).
product evaluation, which is mandatory for forming a satisfactory
response (Grimes et al., 2021). Satisfaction is an experience’s positive
outcome, a function of the perceived performance (Moriuchi, 2021). 3.2. Scales
When the performance of the service or product is favorable or above
expectations, users will have positive experiences that are the attributes We used established scales with solid reliability and validity results,
of the performance expectation (Deng et al., 2010). Many studies (Zhang and responses were collected on a 5-point Likert on strongly disagree
et al., 2021; Buschmeyer et al., 2023) have seen performance expecta­ and strongly agreed points.
tion as a moderator along with user emotional disclosure and user AI Assistant Characteristics: We asked about five characteristics of AI

4
S. Affandi et al. Journal of Retailing and Consumer Services 82 (2025) 104087

assistants, namely accuracy (4 items: “AI assistant produces accurate Table 1

information”), responsiveness (3 items: “I obtain the information from Summary of measurement model.
AI assistant in a timely manner”), compatibility (4 items: “AI assistant Variable Items Factor CR AVE Cronbach’s
fits into my lifestyle”), anthropomorphism (4 items: “AI have con­ Loading Alpha
sciousness”), and affinity (3 items: " I feel as though this AI assistant AI Assistant Usage ACC1 0.659 0.846 0.580 0.840
really understands me”). These scales were derived from Prentice and ACC2 0.840
Nguyen (2020), Lu et al. (2019), Kim and Baek (2018), and ACC3 0.757
Thorbjørnsen et al. (2002). ACC4 0.781

User Engagement: A three-item scale by Saiful Bahry et al. (2021) was RES1 0.693 0.789 0.557 0.784
used to collect responses on user engagement. A sample item is “I feel RES 2 0.710
RES 3 0.828
that the engagement I have with AI is very human like".
User Emotional Disclosure: Malloch and Zhang (2019) proposed a COM1 0.790 0.816 0.528 0.812
four-item scale on user emotional disclosure. The sample item is “I feel COM2 0.753
COM3 0.695
more optimistic now that I have talked with the AI." COM4 0.660
Performance Expectation: Performance expectation is measured on a
ANTH1 0.621 0.836 0.564 0.822
four-item scale proposed by Venkatesh et al. (2007). The sample item is
ANTH2 0.838
“I believe that using AI to answer my question." ANTH3 0.843
Intention to Reuse: Vijayasarathy and Jones (2000) developed a ANTH4 0.677
five-item scale for intention to reuse measurement. A sample item is “If I AF1 0.739 0.780 0.544 0.768
had access to AI, I plan to use it all the time." AF2 0.825
AF3 0.637
3.3. Common method bias and endogeneity User Experience UE1 0.753 0.801 0.574 0.800
UE2 0.778
Using a random sampling method, we used self-report measures to UE3 0.740
collect the data from AI assistant users, but common method bias (CMB) User Emotional UED1 0.745 0.774 0.536 0.770
could be a problem. Using Podsakoff et al. (2003) guidelines to minimize Disclosure UED2 0.805
the CMB using ex-ante and ex-post methods. First, we carefully mixed all UED3 0.636

the questions (ex-ante approach) of all variables, which helped us in Performance PR1 0.768 0.754 0.507 0.750
obtaining unbiased responses. For the ex-post approach, we used Har­ Expectations PR2 0.699
PR3 0.665
man’s one-factor method by running an exploratory factor analysis
using the maximum likelihood method. The results revealed that the Intention to Reuse ITR1 0.767 0.917 0.788 0.908
ITR2 0.933
single factor explained 34% of the variance. Hence, we confirm that
ITR3 0.951
CMB is absent in the data (Podsakoff and Organ, 1986). On the other
hand, endogeneity is another common issue in survey-based studies that Fit Indices (χ 2/df = 4.53; p < 0.01, SRMR = 0.065, RMSEA = 0.074, CFI
= 0.901)
leads to inconsistent results due to inappropriate statistical prediction
and inference. Using Hult et al. (2018) method, we run a regression ACC = accuracy, RES = responsiveness, COM = compatibility, ANTH =
analysis using gender, age and education as control variables on inten­ anthropomorphism, AF = affinity, UE = user experience, UED = user emotional
tion to reuse. All the paths are insignificant, confirming that endogeneity disclosure, PR = performance expectations, ITR = intention to reuse.
is absent in this research.
(SEM) on AMOS v. 24. The results showed that AIassistant has a positive
4. Results impact on user engagement (β = 0.564, p < 0.01), user emotional
disclosure (β = 0.272, p < 0.01), intention to reuse ((β = 0.359, p <
4.1. Measurement model 0.01) supporting H1, H2, and H3. User engagement has been found to
positively influence intention to reuse (β = 0.303, p < 0.01), supporting
The analyses were conducted using a two-step process. In the first H4. User emotional disclosure positively influences intention to reuse (β
step, the reliability and validity of data were determined using AMOS v. = 0.118, p < 0.01) supporting H6.
24. The hypothesized model contains nine variables, namely AI char­ The mediating roles of user engagement (β = 0.171, LLCI – ULCI =
acteristics (accuracy, responsiveness, compatibility, anthropomor­ 0.108–0.250, p < 0.01) and user emotional disclosure (β = 0.032, LLCI –
phism, and affinity), user engagement, user emotional disclosure, ULCI = 0.010–0.066, p < 0.01) between AIassistant and intention to
intention to reuse, and performance expectations. The results in Table 1 reuse are also confirmed, hence providing support to our H5 and H7.
showed that the nine-factor model fulfills the minimum model fitness Lastly, the moderating role of performance expectation was also deter­
(χ2/df = 4.53; CFI = 0.901, SRMR = 0.065, and RMSEA = 0.074). The mined. The results indicated that performance expectation has a positive
factor loadings of a few items are less than the minimum threshold value impact on intention to reuse (H8 supported) and positively moderate the
of 0.70, but Hair et al. (2010) proposed that the factor loading between user engagement-intention (H9) to reuse relationship (β = 0.071, LLCI –
0.40 and 0.70 is also acceptable if the model fits along with composite ULCI = 0.052–0.189, p < 0.05), and user emotional disclosure-intention
reliability (CR) and average variance extraction (AVE) meets the mini­ to reuse (H10) relationship (β = 0.044, LLCI – ULCI = 0.017–0.215, p <
mum requirement. All items are retained for the structural model since 0.05). All the results are shown in Table 3, whereas slope diagrams are
Cronbach’s Alpha, AVEs, and CR exceed their minimum values. There­ shown in Figs. 2 and 3, respectively.
fore, the results support the reliability and validity of the data (Fornell
and Larcker, 1981). Lastly, HTMT analysis (Table 2) was performed, and 5. Implications & conclusion
all the correlations were less than 0.85, confirming the discriminant
validity of the data. 5.1. Discussion and theoretical implications

4.2. Hypotheses testing AI assistants have been widely used, and every player in the mobile
phone market and other technology leaders have introduced intelligent
The hypotheses were analyzed using structural equation modeling AI assistants (Saleem et al., 2024; Hasan et al., 2021). As the future of AI

5
S. Affandi et al. Journal of Retailing and Consumer Services 82 (2025) 104087

Table 2
HTMT results.
ACC RES COM ANTH AF UE UED PR ITR

ACC ​ ​ ​ ​ ​ ​ ​ ​ ​
RES 0.739 ​ ​ ​ ​ ​ ​ ​ ​
COM 0.617 0.641 ​ ​ ​ ​ ​ ​ ​
ANTH 0.372 0.319 0.529 ​ ​ ​ ​ ​ ​
AF 0.371 0.312 0.455 0.722 ​ ​ ​ ​ ​
UE 0.150 0.023 0.327 0.536 0.702 ​ ​ ​ ​
UED 0.015 0.006 0.162 0.265 0.236 0.264 ​ ​ ​
PR 0.043 0.041 0.002 0.039 0.104 0.025 0.149 ​ ​
ITR 0.274 0.168 0.462 0.328 0.476 0.523 0.182 0.154 ​

is debated, this research has wider implications. There have been found
Table 3
positive relationships among the variables of the study as AI that has a
Hypotheses testing.
reduction in barriers to the use of technology can expand the adoption of
Path beta LLCI ULCI p-value technology as the user is engaged and makes emotional disclosures to AI,
AIAD → UE 0.564 – – 0.001 resulting in greater intention to reuse (Rahman et al., 2023). Perfor­
AIAD → UED 0.272 – – 0.001 mance expectation also positively moderates the relationship between
AIAD →ITR 0.359 0.001
– –
user engagement, emotional disclosure, and intention to reuse. This
UE → ITR 0.303 – – 0.001
UED → ITR 0.118 – – 0.009 raises questions about brand loyalty, education, and employability sta­
AIAD → UE → ITR 0.171 0.108 0.250 0.001 tus, which may also contribute to such reasoning.
AIAD → UED → ITR 0.032 0.010 0.066 0.007 This research has a significant contribution to marketing literature in
PR → ITR 0.114 – – 0.050 information technology. First, the existing literature used anthropo­
UE x PR → ITR 0.071 0.052 0.189 0.050
UED x PR → ITR 0.044 0.017 0.215 0.050
morphic feature as singular (Liu et al., 2024; Munnukka et al., 2022),
and others used perceived usefulness and perceived ease of use (Silva
et al., 2023), examining the AI assistant effectiveness using multidi­
mensional features are rare in existing literature. However, anthropo­
morphism has been investigated in an online service context (Benbasat
et al., 2020), but using it as a sole feature is a major issue. Hence, this
research explores the impact of five key dimensions of AI assistants,
namely accuracy, responsiveness, compatibility, affinity, and anthro­
pomorphism, on the intention to reuse and extend the existing studies of
Munnukka et al. (2022), Liu et al. (2024) and Xie et al. (2024).
Second, our study is consistent with dual process theory and
expanded it in marketing literature by empirically testing the AI assis­
tants and intention to reuse the application. As Dhar and Gorlin (2013)
explained, dual process theory postulates that consumers process in­
formation using two distinctive but synergistic systems. Although the
dual process theory has been extensively tested in various fields, the
cognitive and emotional scope has been studied in our context. There­
fore, our study advanced the literature that explains preferences and
choice decisions of using AI assistants as deliberate and rational.
Third, performance expectation acts as a catalyst and increases the
Fig. 2. Moderating role of PR between UE and ITR. intention to reuse AI assistants. Consumers have been using AI assistants
to ask basic questions, search, browse, and listen to music (Hasan et al.,
2021). The critical issue reinforced by this study includes accruing
greater benefits through AI assistant advantage. Although companies are
careful regarding consumers’ privacy, it leads to more user engagement,
user emotional disclosure, and performance expectations impacting
their reuse intention. Public apologies have been issued by companies
who have breached users’ privacy and made consumers revisit and
rethink using such devices and software (Hasan et al., 2021). This
research study empirically tested the role of AI assistants in boosting
user engagement and user emotional disclosure that influences intention
to reuse among consumers.
We found the positive impact of AI assistant characteristics on user
engagement. Also, a positive relationship has been found between AI
assistant advantages and user emotional disclosure under the umbrella
of dual process theory. Limited research has analyzed the benefits of AI
assistants from the perspective of user emotional disclosure and
engagement, which impact the intention to reuse AI assistants. Although
the literature used both utilitarian and hedonic perspectives to explore
Fig. 3. Moderating role of PR between UED and ITR. the relationship between AI assistants and customer experience (Gursoy
et al., 2019), this research extends it by adding user emotional disclo­
sure. These findings are aligned with previous literature that considers

6
S. Affandi et al. Journal of Retailing and Consumer Services 82 (2025) 104087

AI assistant users to have the intention to reuse due to the accuracy, enjoyment and germane to adopt that brand. For instance, using AI
fastness, ease, and compatibility when they help the consumers com­ technology in beauty and fashion gives consumers first-hand experience
plete their regular tasks (Huang and Rust, 2021; Cheng and Jiang, before buying. As retailers are now actively engaging consumers
2022). through AI assistants, relationship marketing should design strategies to
There has been minimal attention to the area of research considering differentiate themselves and provide an opportunity for a unique
users’ perspectives, such as engagement and emotional disclosure, to tap customer journey. As Ramadan (2021) and Saleem et al. (2024) claimed,
the intention to reuse AI assistants in the future. This research addresses brand-consumer relationships would strengthen if retailers aggressively
this gap by empirically verifying the relationship between consumers’ used technology.
engagement and emotional disclosure to reuse (Ramadan, 2021). In the Lastly, people have increased performance expectations commen­
area of AI devices, dual process theory has been applied in the area of surate with the growing dependency on electronic devices such as AI
marketing, which is bringing novelty to this research. Regarding the use assistants and smartphones (Prizant-Passal et al., 2016). Results of this
of AI assistants, it has been seen that high user engagement and high user study show that AI assistant advantages led to increased user engage­
emotional disclosure, along with high-performance expectations of AI ment and user emotional disclosure among users with high-performance
assistants, have increased intention to reuse among consumers since expectations that influence the intention to reuse AI assistants among
they evaluate and give more importance to such factors when engaging users. Marketers should emphasize the marketing strategies that prac­
with AI assistants. However, performance expectation has been identi­ tically benefit the users.
fied as a boundary condition that moderates customer engagement and
emotional disclosure relationship with reuse intentions. It has been 5.3. Study limitations
found in previous studies that users with high-performance expectations
have higher expectations in social interactions, which are virtual, as Limitations of this research include the need for improvement in the
with AI assistants (Guerreiro and Loureiro, 2023; Okazaki et al., 2012). adaptability and availability of the testing models in various back­
This is attributable to less user tolerance for inaccuracies and errors grounds related to AI. The proposed model of this study has been
when using AI assistants. This involves a trade-off between the inter­ examined in the framework of AI assistants, which provide importance
action between real human beings and the expectation of users to AI technology in human life. Future research should consider diverse
regarding performance, which consists of meeting social needs and be­ samples to analyze the conceptual model in AI areas. Also, using the
longings through AI assistants. Technological advancements entirely Delphi method to stimulate experts’ opinions on related topics will add
changed the consumers’ emotional responses to high-performance ex­ qualitative support to this area of research. This study has examined
pectations (Okazaki et al., 2012). limited variables that affect users’ intention to reuse AI. Other factors
may also affect users’ intentions, including product, performance, and
5.2. Managerial implications emotional perspective. There is a dire need to pay attention to the in­
fluence of AI technology on the behaviors of consumers since AI tech­
There is massive potential for AI assistants, which reduce costs and nology has been increasingly used in the retail industry. This study has
provide operational efficiency, prompting a wide investigation of such only investigated high-tech technology products such as AI assistants;
products (Chakraborty et al., 2024). Still, the focus is mainly on however, future studies can focus on studying the influence of such AI
adopting a technical perspective, and user emotional disclosure and products on shopping and other experiences and performance outcomes
engagement remain tentative in such studies. In addition, there has been in less technological and highly diverse market sectors.
a lack of guidance for companies in conducting marketing and promo­
tional activities (Yuan et al., 2022). The findings of this study provide CRediT authorship contribution statement
insights into the marketing and R&D departments of the companies,
which suggest that the stakeholders have increased the use of AI assis­ Sana Affandi: Writing – review & editing, Writing – original draft,
tants by providing professional abilities and intelligent interactions, Visualization, Data curation, Conceptualization. Muhammad Ishtiaq
adding value to users and enhancing positive intentions among users. Ishaq: Writing – review & editing, Writing – original draft, Supervision,
Managers must focus on the advantages that such AI apps and devices Data curation, Conceptualization. Ali Raza: Writing – review & editing,
offer to improve the product’s performance. Supplementing this, solu­ Writing – original draft, Validation, Formal analysis, Conceptualization.
tions to the problems proposed by AIassistants should be based on users’ Qurat-ul-ain Talpur: Writing – review & editing, Writing – original
preferences and habits, where analysis is supported by the system’s data draft, Methodology, Investigation, Conceptualization. Rehan Ahmad:
processing for personalized and adaptive aspects that increase the use o f Writing – review & editing, Methodology, Formal analysis, Data cura­
AI assistants. Manufacturers consider the design process and other ele­ tion, Conceptualization.
ments to add enjoyment and entertainment to AI. Such AI advantages
considered in this study can provide manufacturers with suggestions and Declaration of competing interest
references when considering user engagement and emotional disclosure.
AI technology and its application in the retail industry are being used The authors confirm that they have no conflict of interest.
extensively, and brands are also paying careful attention to technology
adoption, such as AI assistants, in terms of consumption patterns. Our Data availability
research recommends that the brand managers also involve consumers
in the design process through AI services, which provides a sense of Data will be made available on request.

Appendix 1. Measurement Scales

Variable Measurement Items

AI Assistant Advantage Accuracy

A1. AI assistant produces accurate information.
A2. AI assistant produces adequate service.
(continued on next page)

7
S. Affandi et al. Journal of Retailing and Consumer Services 82 (2025) 104087

(continued )
Variable Measurement Items

A3. AI assistant produces complete service.

A4. AI assistant produces credible service.
Responsiveness
R1. I obtain the information from AI assistant in a timely manner.
R2. AI assistant is very fast in responding to my feedback.
R3. AI assistant processes my input very quickly.
Compatibility
C1. AI assistant fits into my lifestyle.
C2. AI assistant fits well with the way I live.
C3. AI assistant suits me.
C4. AI assistant is compatible with all aspects of my life.
Anthropomorphism
Anth1. Similar to manual service, AI have a mind of their own.
Anth2. Similar to manual service, AI have consciousness.
Anth3. Similar to manual service, AI have their own free will.
Anth4. Similar to manual service, AI will experience emotions.
Affinity
Af1. I feel as though this AI assistant really understands me.
Af2. I feel I have a good relationship with the AI assistant.
Af3. AI makes me enthusiastic about using its services.
User Engagement ENG1. I feel that the engagement I have with AI is very human like.
ENG2. I feel that the conversation I have with AI is very engaging.
ENG3. The engagement I have with AI is very meaningful.
User Emotional Disclosure UED1. I feel more optimistic now that I have talked with the AI.
UED2. Whether user feel better after the conversation with AI.
UED3. I want to disclose how I really felt after using AI assistant.
UED4. I want to disclose my emotions after using AI assistants.
Performance Expectation PE1. I believe that using AI to answer my question
PE2. I believe that using AI to answer my question would save time so I can do other activities in my day to day.
PE3. I believe that using AI to answer my question would bring me greater convenience.
Intention to Reuse INT1. If I had access to AI, I plan to use it all the time.
INT2. I think it will be worth it for me to use AI when it’s available.
INT3. I would recommend using a AI to my friends and family.
INT4. If I have the opportunity to purchase more AI (e.g., Google home, Alexa), I would.
INT5. I rather use a AI to answer my questions than asking another individual for answers.

References Chung, M., Ko, E., Joung, H., Kim, S.J., 2020. Chatbot e-service and customer satisfaction
regarding luxury brands. J. Bus. Res. 117, 587–595.
Comrey, A.L., Lee, H.B., 1992. A first course in Factor Analysis. Eribaum, Hillsdale: NJ.
Adam, M., Wessel, M., Benlian, A., 2021. AI-based chatbots in customer service and their
DeLone, W.H., McLean, E.R., 2003. The DeLone and McLean model of information
effects on user compliance. Electron. Mark. 31 (2), 427–445.
systems success: a ten-year update. J. Manag. Inf. Syst. 19 (4), 9–30.
Alexander, M.J., Jaakkola, E., Hollebeek, L.D., 2018. Zooming out: actor engagement
Deng, L., Turner, D.E., Gehling, R., Prince, B., 2010. User experience, satisfaction, and
beyond the dyadic. J. Serv. Manag. 29 (3), 333–351.
continual usage intention of IT. Eur. J. Inf. Syst. 19 (1), 60–75.
Arachchi, H.D.M., Samarasinghe, G.D., 2023. Intention to adopt intelligent clothing in
Dewey, A.R., 2023. Metacognitive control in single-vs. dual-process theory. Think. Reas.
the fashion retail industry: extending the Hisam Model with Technology Readiness.
29 (2), 177–212.
Int. J. Hum. Comput. Interact. 1–15.
Dhar, R., Gorlin, M., 2013. A dual-system framework to understand preference
Araujo, T., 2018. Living up to the chatbot hype: the influence of anthropomorphic design
construction processes in choice. J. Consum. Psychol. 23 (4), 528–542.
cues and communicative agency framing on conversational agent and company
Epley, N., Waytz, A., Akalis, S., Cacioppo, J.T., 2008. When we need a human:
perceptions. Comput. Hum. Behav. 85, 183–189.
motivational determinants of anthropomorphism. Soc. Cognit. 26 (2), 143–155.
Ashfaq, M., Yun, J., Yu, S., Loureiro, S.M.C., 2020. I, Chatbot: modeling the determinants
Everitt, B.S., 1975. Multivariate analysis: The need for data, and other problems. Br. J.
of users’ satisfaction and continuance intention of AI-powered service agents.
Psychiatr. 126 (3), 237–240.
Telematics Inf. 54, 101473.
Fornell, C., Larcker, D.F., 1981. Evaluating structural equation models with unobservable
Babin, B.J., Darden, W.R., Babin, L.A., 1998. Negative emotions in marketing research:
variables and measurement error. J. Market. Res. 18 (1), 39–50.
affect or artifact? J. Bus. Res. 42 (3), 271–285.
Gesk, T.S., Leyer, M., 2022. Artificial intelligence in public services: when and why
Benbasat, I., Dimoka, A., Pavlou, P.A., Qiu, L., 2020. The role of demographic similarity
citizens accept its usage. Govern. Inf. Q. 39 (3), 101704.
in people’s decision to interact with online anthropomorphic recommendation
Grimes, G.M., Schuetzler, R.M., Giboney, J.S., 2021. Mental models and expectation
agents: Evidence from a functional magnetic resonance imaging (fMRI) study. Int. J.
violations in conversational AI interactions. Decis. Support Syst. 144, 113515.
Hum. Comput. Stud. 133, 56–70.
Guerreiro, J., Loureiro, S.M.C., 2023. I am attracted to my cool smart assistant! analyzing
Brodie, R.J., Ilic, A., Juric, B., Hollebeek, L., 2013. Consumer engagement in a virtual
attachment-aversion in AI-human relationships. J. Bus. Res. 161, 113863.
brand community: an exploratory analysis. J. Bus. Res. 66 (1), 105–114.
Guha, A., Bressgott, T., Grewal, D., Mahr, D., Wetzels, M., Schweiger, E., 2023. How
Buschmeyer, K., Hatfield, S., Heine, I., Jahn, S., Markus, A.L., 2023. Expectation
artificiality and intelligence affect voice assistant evaluations. J. Acad. Market. Sci.
management in AI implementation projects: a case study. EuroMed J. Bus. 18 (3),
51 (4), 843–866.
441–451.
Guha, A., Grewal, D., Kopalle, P.K., Haenlein, M., Schneider, M.J., Jung, H., et al., 2021.
Chakraborty, D., Polisetty, A., Sowmya, G., Rana, N.P., Khorana, S., 2024. Unlocking the
How artificial intelligence will affect the future of retailing. J. Retailing 97 (1),
potential of AI: enhancing consumer engagement in the beauty and cosmetic product
28–41.
purchases. J. Retailing Consum. Serv. 79, 103842.
Guo, W., Luo, Q., 2023. Investigating the impact of intelligent personal assistants on the
Chen, J.S., Le, T.T.Y., Florence, D., 2021. Usability and responsiveness of artificial
purchase intentions of Generation Z consumers: the moderating role of brand
intelligence chatbot on online customer experience in e-retailing. Int. J. Retail
credibility. J. Retailing Consum. Serv. 73, 103353.
Distrib. Manag. 49 (11), 1512–1531.
Gursoy, D., Chi, O.H., Lu, L., Nunkoo, R., 2019. Consumers acceptance of artificially
Cheng, Y., Jiang, H., 2022. Customer–brand relationship in the era of artificial
intelligent (AI) device use in service delivery. Int. J. Inf. Manag. 49, 157–169.
intelligence: understanding the role of chatbot marketing efforts. J. Prod. Brand
Hair Jr., J.F., Black, W.C., Babin, B.J., Anderson, R.E., Tatham, R.L., 2010. Multivariate
Manag. 31 (2), 252–264.
Data Analysis, seventh ed. Pearson Education, Upper Saddle River, NJ.
Chiang, H.S., 2013. Continuous usage of social networking sites: the effect of innovation
Han, S., Yang, H., 2018. Understanding adoption of intelligent personal assistants: a
and gratification attributes. Online Inf. Rev. 37 (6), 851–871.
parasocial relationship perspective. Ind. Manag. Data Syst. 118 (3), 618–636.
Choi, T.R., Drumwright, M.E., 2021. OK, Google, why do I use you?” Motivations, post-
consumption evaluations, and perceptions of voice AI assistants. Telematics Inf. 62,
101628.

8
S. Affandi et al. Journal of Retailing and Consumer Services 82 (2025) 104087

Hari, H., Iyer, R., Sampat, B., 2022. Customer brand engagement through chatbots on Pelau, C., Dabija, D.C., Ene, I., 2021. What makes an AI device humanlike? The role of
bank websites–examining the antecedents and consequences. Int. J. Hum. Comput. interaction quality, empathy and perceived psychological anthropomorphic
Interact. 38 (13), 1212–1227. characteristics in the acceptance of artificial intelligence in the service industry.
Hasan, R., Shams, R., Rahman, M., 2021. Consumer trust and perceived risk for voice- Comput. Hum. Behav. 122, 106855.
controlled artificial intelligence: the case of Siri. J. Bus. Res. 131, 591–597. Pennebaker, J.W., Chung, C.K., 2007. Expressive writing, emotional upheavals, and
Ho, A., Hancock, J., Miner, A.S., 2018. Psychological, relational, and emotional effects of health. Foundations of health psychology 263–284.
self-disclosure after conversations with a chatbot. J. Commun. 68 (4), 712–733. Pihlström, M., Brush, G.J., 2008. Comparing the perceived value of information and
Hollebeek, L.D., Srivastava, R.K., Chen, T., 2019. SD logic–informed customer entertainment mobile services. Psychol. Market. 25 (8), 732–755.
engagement: integrative framework, revised fundamental propositions, and Pillai, R., Sivathanu, B., Dwivedi, Y.K., 2020. Shopping intention at AI-powered
application to CRM. J. Acad. Market. Sci. 47, 161–185. automated retail stores (AIPARS). J. Retailing Consum. Serv. 57, 102207.
Huang, M.H., Rust, R.T., 2021. Engaged to a robot? The role of AI in service. J. Serv. Res. Pitardi, V., Marriott, H.R., 2021b. Alexa, she’s not human but… Unveiling the drivers of
24 (1), 30–41. consumers’ trust in voice-based artificial intelligence. Psychol. Market. 38 (4),
Huh, J., Kim, H.Y., Lee, G., 2023. "Oh, happy day!" Examining the role of AI-powered 626–642.
voice assistants as a positive technology in the formation of brand loyalty. J. Res. Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y., Podsakoff, N.P., 2003. Common method
Indian Med. 17 (5), 794–812. biases in behavioral research: a critical review of the literature and recommended
Hult, G.T.M., Hair, Proksch, D., Sarstedt, M., Pinkwart, A., Ringle, C.M., 2018. remedies. J. Appl. Psychol. 88 (5), 879.
Addressing endogeneity in international marketing applications of partial least Podsakoff, P.M., Organ, D.W., 1986. Self-reports in organizational research: problems
squares structural equation modeling. J. Int. Market. 26 (3), 1–21. and prospects. J. Manag. 12 (4), 531–544.
Jan, I.U., Ji, S., Kim, C., 2023. What (de) motivates customers to use AI-powered Prentice, C., Nguyen, M., 2020. Engaging and retaining customers with AI and employee
conversational agents for shopping? The extended behavioral reasoning perspective. service. J. Retailing Consum. Serv. 56, 102186.
J. Retailing Consum. Serv. 75, 103440. Prizant-Passal, S., Shechner, T., Aderka, I.M., 2016. Social anxiety and internet use–A
Jenneboer, L., Herrando, C., Constantinides, E., 2022. The impact of chatbots on meta-analysis: what do we know? What are we missing? Comput. Hum. Behav. 62,
customer loyalty: a systematic literature review. Journal of theoretical and applied 221–229.
electronic commerce research 17 (1), 212–229. Rahman, M.S., Bag, S., Hossain, M.A., Fattah, F.A.M.A., Gani, M.O., Rana, N.P., 2023.
Kamoonpuri, S.Z., Sengar, A., 2023. Hi, May AI help you? An analysis of the barriers The new wave of AI-powered luxury brands online shopping experience: the role of
impeding the implementation and use of artificial intelligence-enabled virtual digital multisensory cues and customers’’ engagement. J. Retailing Consum. Serv.
assistants in retail. J. Retailing Consum. Serv. 72, 103258. 72, 103273.
Kaur, D., Uslu, S., Rittichier, K.J., Durresi, A., 2022. Trustworthy artificial intelligence: a Ramadan, Z.B., 2021. ""Alexafying" shoppers: the examination of Amazon’s captive
review. ACM Comput. Surv. 55 (2), 1–38. relationship strategy. J. Retailing Consum. Serv. 62, 102610.
Kim, S., Baek, T.H., 2018. Examining the antecedents and consequences of mobile app Rhim, J., Kwak, M., Gong, Y., Gweon, G., 2022. Application of humanization to survey
engagement. Telematics Inf. 35 (1), 148–158. chatbots: change in chatbot perception, interaction experience, and survey data
Klein, K., Martinez, L.F., 2023. The impact of anthropomorphism on customer quality. Comput. Hum. Behav. 126, 107034.
satisfaction in chatbot commerce: an experimental study in the food sector. Electron. Rosenthal-Von Der Pütten, Krämer, N.C., 2014. How design characteristics of robots
Commer. Res. 23 (4), 2789–2825. determine evaluation and uncanny valley related responses. Comput. Hum. Behav.
Kumar, A., Bala, P.K., Chakraborty, S., Behera, R.K., 2024. Exploring antecedents 36, 422–439.
impacting user satisfaction with voice assistant app: a text mining-based analysis on Saiful Bahry, Masrom, M., Masrek, M.N., 2021. Measuring validity and reliability of
Alexa services. J. Retailing Consum. Serv. 76, 103586. website credibility factors in influencing user engagement questionnaire. Int. J. Web
Kunduru, A.R., 2023. Artificial intelligence usage in cloud application performance Inf. Syst. 17 (1), 18–28.
improvement. Central Asian Journal of Mathematical Theory and Computer Sciences Saleem, T., Ishaq, M.I., Raza, A., Junaid, M., 2024. Exploring the effect of telepresence
4 (8), 42–47. and escapism on consumer post-purchase intention in an immersive virtual reality
Lee, J., Lee, D., Lee, J.G., 2022b. Influence of rapport and social presence with an AI environment. J. Retailing Consum. Serv. 81, 104014.
psychotherapy chatbot on users’ self-disclosure. Int. J. Hum. Comput. Interact. 1–12. Sheehan, B., Jin, H.S., Gottlieb, U., 2020. Customer service chatbots: anthropomorphism
Letheren, K., Russell-Bennett, R., Mulcahy, R.F., McAndrew, R., 2019. Rules of and adoption. J. Bus. Res. 115, 14–24.
(household) engagement: technology as manager, assistant and intern. Eur. J. Shimizu, Y., Osaki, S., Hashimoto, T., Karasawa, K., 2021. How do people view various
Market. 53 (9), 1934–1961. kinds of smart city services? Focus on the acquisition of personal information.
Liang, Y., Lee, S.H., Workman, J.E., 2020. Implementation of artificial intelligence in Sustainability 13 (19), 11062.
fashion: are consumers ready? Cloth. Text. Res. J. 38 (1), 3–18. Sierra, J.J., Hyman, M.R., 2011. Outlet mall shoppers’ intentions to purchase apparel: a
Liu, W., Jiang, M., Li, W., Mou, J., 2024. How does the anthropomorphism of AI chatbots dual-process perspective. J. Retailing Consum. Serv. 18 (4), 341–347.
facilitate users’ reuse intention in online health consultation services? The Silva, D.G., Coutinho, C., Costa, C.J., 2023. Factors influencing free and open-source
moderating role of disease severity. Technol. Forecast. Soc. Change 203, 123407. software adoption in developing countries—an empirical study. J. Open Innovat.:
Lu, L., Cai, R., Gursoy, D., 2019. Developing and validating a service robot integration Technol. Market Complex. 9 (1), 100002.
willingness scale. Int. J. Hospit. Manag. 80, 36–51. Skjuve, M., Følstad, A., Fostervold, K.I., Brandtzaeg, P.B., 2021. My chatbot companion-a
Lucas, G.M., Gratch, J., King, A., Morency, L.P., 2014. It’s only a computer: virtual study of human-chatbot relationships. Int. J. Hum. Comput. Stud. 149, 102601.
humans increase willingness to disclose. Comput. Hum. Behav. 37, 94–100. So, K.K.F., King, C., Sparks, B.A., Wang, Y., 2016. The role of customer engagement in
Luo, X., Tong, S., Fang, Z., Qu, Z., 2019. Frontiers: machines vs. humans: the impact of building consumer loyalty to tourism brands. J. Trav. Res. 55 (1), 64–78.
artificial intelligence chatbot disclosure on customer purchases. Market. Sci. 38 (6), Song, J.Y., Pycha, A., Culleton, T., 2022. Interactions between voice-activated AI
937–947. assistants and human speakers and their implications for second-language
Maduku, D.K., Mpinganjira, M., Rana, N.P., Thusi, P., Ledikwe, A., Mkhize, N.H.B., 2023. acquisition. Front. Commun. 7, 995475.
Assessing customer passion, commitment, and word-of-mouth intentions in digital Stanovich, K.E., West, R.F., 2000. Advancing the rationality debate. Behav. Brain Sci. 23
assistant usage: the moderating role of technology anxiety. J. Retailing Consum. (5), 701–717.
Serv. 71, 103208. Storbacka, K., Brodie, R.J., Böhmann, T., Maglio, P.P., Nenonen, S., 2016. Actor
Malloch, Y.Z., Zhang, J., 2019. Seeing others receive support online: Effects of self- engagement as a microfoundation for value co-creation. J. Bus. Res. 69 (8),
disclosure and similarity on perceived similarity and health behavior intention. 3008–3017.
J. Health Commun. 24 (3), 217–225. Sung, E.C., Bae, S., Han, D.I.D., Kwon, O., 2021. Consumer engagement via interactive
Mehmood, K., Kautish, P., Shah, T.R., 2024. Embracing digital companions: unveiling artificial intelligence and mixed reality. Int. J. Inf. Manag. 60, 102382.
customer engagement with anthropomorphic AI service robots in cross-cultural Tam, K.P., Lee, S.L., Chao, M.M., 2013. Saving Mr. Nature: anthropomorphism enhances
context. J. Retailing Consum. Serv. 79, 103825. connectedness to and protectiveness toward nature. J. Exp. Soc. Psychol. 49 (3),
Moore, S., Bulmer, S., Elms, J., 2022. The social significance of AI in retail on customer 514–521.
experience and shopping practices. J. Retailing Consum. Serv. 64, 102755. Thorbjørnsen, H., Supphellen, M., Nysveen, H., Egil, P., 2002. Building brand
Moriuchi, E., 2021. An empirical study on anthropomorphism and engagement with relationships online: A comparison of two interactive applications. J. Interact.
disembodied AIs and consumers’ reuse behavior. Psychol. Market. 38 (1), 21–42. Market. 16 (3), 17–34.
Munnukka, J., Talvitie-Lamberg, K., Maity, D., 2022. Anthropomorphism and social Tinsley, H.E., Tinsley, D.J., 1987. Uses of factor analysis in counseling psychology
presence in Human–Virtual service assistant interactions: the role of dialog length research. J. Counsel. Psychol. 34 (4), 414.
and attitudes. Comput. Hum. Behav. 135, 107343. Transparency Market Research, 2016. Growing focus on strengthening customer
Nass, C., Moon, Y., 2000. Machines and mindlessness: social responses to computers. relations spurs adoption of intelligent virtual assistant technology. : www.
J. Soc. Issues 56 (1), 81–103. prnewswire.com/news-releases/growing-focus-on-strengthening-customer-relati
Nazir, S., Khadim, S., Asadullah, M.A., Syed, N., 2023. Exploring the influence of onsspurs-adoption-ofintelligent-virtual-assistant-technology-says-tmr-589042721.
artificial intelligence technology on consumer repurchase intention: the mediation html. (Accessed 6 March 2018). says TMR available at.
and moderation approach. Technol. Soc. 72, 102190. Tsai, W.H.S., Liu, Y., Chuan, C.H., 2021. How chatbots’ social presence communication
Okazaki, S., Navarro-Bailón, M.Á., Molina-Castillo, F.J., 2012. Privacy concerns in quick enhances consumer engagement: the mediating role of parasocial interaction and
response code mobile promotion: the role of social anxiety and situational dialogue. J. Res. Indian Med. 15 (3), 460–482.
involvement. Int. J. Electron. Commer. 16 (4), 91–120. Venkatesh, V., Davis, F., Morris, M.G., 2007. Dead or alive? The development, trajectory
Pantano, E., Pizzi, G., 2020. Forecasting artificial intelligence on online customer and future of technology adoption research. The Development, Trajectory and
assistance: Evidence from chatbot patents analysis. J. Retailing Consum. Serv. 55, Future of Technology Adoption Research (April 27, 2007). Venkatesh, V., Davis, FD,
102096. and Morris, MG "Dead or Alive 267–286.

9
S. Affandi et al. Journal of Retailing and Consumer Services 82 (2025) 104087

Verhagen, J., Wagenmakers, E.J., 2014. Bayesian tests to quantify the result of a environment: an empirical study of B2B E-commerce platform. Ind. Market. Manag.
replication attempt. J. Exp. Psychol. Gen. 143 (4), 1457. 92, 101–110.
Vijayasarathy, L.R., Jones, J.M., 2000. Print and Internet catalog shopping: assessing Yuan, C., Zhang, C., Wang, S., 2022. Social anxiety as a moderator in consumer
attitudes and intentions. Internet Res. 10 (3), 191–202. willingness to accept AI assistants based on utilitarian and hedonic values.
Xie, C., Wang, Y., Cheng, Y., 2024. Does artificial intelligence satisfy you? A meta- J. Retailing Consum. Serv. 65, 102878.
analysis of user gratification and user satisfaction with AI-powered chatbots. Int. J. Zhang, R., McNeese, N.J., Freeman, G., Musick, G., 2021. An ideal human" expectations
Hum. Comput. Interact. 40 (3), 613–623. of AI teammates in human-AI teaming. Proceedings of the ACM on Human-Computer
Xu, K., Lombard, M., 2017. Persuasive computing: feeling peer pressure from multiple Interaction 4 (CSCW3), 1–25.
computer agents. Comput. Hum. Behav. 74, 152–162. Zolkepli, I.A., Kamarulzaman, Y., 2015. Social media adoption: the role of media needs
Yuan, C., Moon, H., Wang, S., Yu, X., Kim, K.H., 2021. Study on the influencing of B2B and innovation characteristics. Comput. Hum. Behav. 43, 189–209.
parasocial relationship on repeat purchase intention in the online purchasing

10