Journal of Marketing Theory and Practice

ISSN: (Print) (Online) Journal homepage: www.tandfonline.com/journals/mmtp20

Determinants of intention to use ChatGPT for

obtaining shopping information

Mohammad Iranmanesh, Madugoda Gunaratnege Senali, Morteza

Ghobakhloo, Behzad Foroughi, Elahe Yadegaridehkordi & Nagaletchimee
Annamalai

To cite this article: Mohammad Iranmanesh, Madugoda Gunaratnege Senali, Morteza

Ghobakhloo, Behzad Foroughi, Elahe Yadegaridehkordi & Nagaletchimee Annamalai (26 Jul
2024): Determinants of intention to use ChatGPT for obtaining shopping information, Journal
of Marketing Theory and Practice, DOI: 10.1080/10696679.2024.2380719

To link to this article: https://doi.org/10.1080/10696679.2024.2380719

Published online: 26 Jul 2024.

Submit your article to this journal

Article views: 37

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=mmtp20
JOURNAL OF MARKETING THEORY AND PRACTICE
https://doi.org/10.1080/10696679.2024.2380719

Determinants of intention to use ChatGPT for obtaining shopping information

Mohammad Iranmanesha, Madugoda Gunaratnege Senalib, Morteza Ghobakhlooc,d, Behzad Foroughie,
Elahe Yadegaridehkordif*, and Nagaletchimee Annamalaig
a
La Trobe Business School, La Trobe University, Melbourne, Victoria, Australia; bSchool of Business and Law, Edith Cowan University,
Joondalup, Washington, Australia; cDivision of Industrial Engineering and Management, Uppsala University, Uppsala, Sweden; dSchool of
Economics and Business, Kaunas University of Technology, Kaunas, Lithuania; eDepartment of International Business Administration, I-Shou
University, Kaohsiung, Taiwan; fSchool of Business, Polytechnic Institute Australia, Sydney, Australia; gSchool of Distance Education, Universiti
Sains Malaysia, George Town, Malaysia

ABSTRACT
Drawing on the “Technology Acceptance Model” (TAM) and “Elaboration Likelihood Model,” this
study explored the determinants of intention to use ChatGPT to obtain shopping information. We
extended TAM by incorporating perceived enjoyment. Data were collected from 412 Malaysians
and analyzed using the “Partial Least Squares” approach. Findings confirmed the significance of
TAM and perceived enjoyment, revealing that information accuracy, completeness, and diagnos­
ticity greatly influence perceived usefulness. These results enhance the understanding of
ChatGPT’s use in shopping information gathering, extend TAM, and highlight how informational
cues impact perceived usefulness, guiding developers and marketers to encourage customer use
of text generators.

Summary Statement of Contribution

This is the first study investigating determinants of intention to use ChatGPT to obtain shopping
information. The study integrated the Technology Acceptance Model (TAM) and the Elaboration
Likelihood Model (ELM) and extended TAM by incorporating perceived enjoyment. The results
provide valuable guidance to developers of text generators and marketing managers on motivat­
ing customers to use text generators to gather information about shopping.

Introduction
paralysis, where customers feel overwhelmed and incap­
In the past few decades, shopping and user purchasing able of making a purchase decision (Manolică et al.,
mechanisms have drastically transformed (Bag et al., 2021). Businesses are now using Artificial Intelligence
2022). In the current business environment, digital tech­ (AI) driven chatbot assistant technologies to assist cus­
nologies play a significant role (Szász et al., 2022). tomers in making purchase decisions in order to over­
Retailers have increasingly integrated technological come this technology-driven dilemma. Chatbots may
innovations to enhance the shopping experience and offer numerous benefits to customers, such as providing
better connect and enhance customer relationships. product suggestions based on customers’ profiles
The digitalization of the marketplace has not been (Landim et al., 2022).
a one-way road. The emergence and diffusion of tech­ Despite being embryonic, AI chatbot technology is
nological innovations have also revolutionized how cus­ unprecedently progressing. This progression mainly man­
tomers choose their products, providing them access to ifests in the unpredictable public launch of ChatGPT by
different sources of information (Jiang et al., 2019). OpenAI. ChatGPT is an AI-powered conversational bot
Although digital transformation enables customers to that operates on the generative pre-trained transformer
obtain more product information, information overload language model. This advanced chatbot differs from con­
may be a double-edged sword and complicate purchase ventional models in its capacity to store and retrieve infor­
decisions (Turri & Watson, 2022). Customers are nowa­ mation from previous interactions, tailor responses to
days bombarded by an overwhelming amount of infor­ specific user needs, and offer data-driven and unbiased
mation when deciding on purchasing a product or recommendations. Despite ChatGPT’s potential to provide
service, leading to a phenomenon known as decision valuable recommendations, customers’ acceptance of this

CONTACT Behzad Foroughi [email protected] Program in International Business Administration, I-Shou University, Kaohsiung 84001, Taiwan
*Present affiliation for Elahe Yadegaridehkordy is College of Information and Communications Technology, School of Engineering and Technology,
CQUniversity, Sydney, Australia
© 2024 Taylor & Francis Group, LLC
2 M. IRANMANESH ET AL.

technological innovation for obtaining shopping informa­ shown the importance of hedonic factors in shopping
tion has not been studied. Given this gap, this study aims to (e.g. Banik & Gao, 2023; Y.-C. Huang et al., 2023), TAM
investigate the influential factors driving the intention to considers only the influences of utilitarian factors (PEU
use ChatGPT in the shopping context. Filling this gap is and PU) (Davis, 1989; Venkatesh et al., 2003).
essential for several reasons. First, it offers insights into Accordingly, we added perceived enjoyment to the
how customers adapt advanced AI technologies to their study’s conceptual framework to better capture the
shopping behavior, which is crucial for businesses that seek hedonic aspects of technology use, which are particu­
to integrate such technologies effectively. Second, it con­ larly salient in the context of AI-driven tools like
tributes to the broader discussion on the adoption of AI in ChatGPT. Furthermore, TAM proposed that scholars
customer markets and provides valuable perspectives for incorporate external factors to explain how practically
marketers and technology developers. The present study PEU and PU can be formed. As ChatGPT is a text-
drew on the “Technology Acceptance Model” (TAM) generator tool, it is crucial to investigate how the
(Davis et al., 1989) to identify influential drivers of inten­ characteristics of provided content form customer per­
tion to use ChatGPT as a purchasing recommendation ceptions. Accordingly, following the “Elaboration
system. TAM was selected as it is one of the most widely Likelihood Model” (ELM), as customers utilize
accepted theories for understanding behaviors toward ChatGPT in their self-interest, the elaboration level is
adopting innovative technologies. It has been used widely high, and they evaluate informational cues cognitively
in marketing and has shown excellent power to explain to determine their value (Petty et al., 1986). This deeper
customer behaviors in adopting new technologies such as level of engagement means factors related to the infor­
chatbots (Rese et al., 2020) and augmented reality (Alam mation are expected to be more critical in shaping
et al., 2021). Furthermore, TAM is flexible and can be customer perceptions than aesthetic and interface-
adapted to various contexts and populations, allowing us related dimensions. As Filieri and McLeay (2014) and
to account for contextual factors that might uniquely affect Filieri (2015) recommended, we considered information
the acceptance of ChatGPT in the purchasing context accuracy, completeness, relevance, understandability,
(Sagnier et al., 2020). value-added information, and diagnosticity as informa­
Given the innovative application of ChatGPT in the tional cues that may contribute to the perceived utility
shopping domain, TAM’s parsimonious structure of ChatGPT in the shopping context. This study aims to:
enables a focused examination of fundamental accep­
tance predictors that are particularly relevant to this ● Determine the factors that influence the intention
emerging technology. This approach allows us to inves­ of customers to use ChatGPT as a tool for obtain­
tigate the direct influences of “perceived ease of use” ing shopping information.
(PEU) and “perceived usefulness” (PU) with greater ● Extend the explanatory power of TAM in the
specificity than broader models might permit. ChatGPT context by incorporating perceived
Furthermore, the flexibility of TAM accommodates the enjoyment as a hedonic factor.
integration of context-specific constructs, such as per­ ● Assess the extent to which informational cues
ceived enjoyment, which addresses the hedonic aspects influence the PU of customers.
of using ChatGPT for shopping.
The focused nature of TAM, with its adaptability to This research advances the field by exploring how
include novel constructs, not only aligns with the ChatGPT’s unique capabilities influence customer shop­
specific research questions addressed by this study but ping decisions. This study goes beyond general adoption
also contributes to the technology acceptance literature models by investigating the factors driving customers to
by providing insights into user acceptance in the context adopt ChatGPT for shopping information. It enhances
of advanced AI-driven recommendation systems. By the TAM framework by uniquely integrating perceived
explaining the role of critical determinants in the enjoyment and highlights how hedonic elements specifi­
novel context of ChatGPT, this study offers targeted cally contribute to accepting text-based generative AI for
contributions that can inform the continuous develop­ shopping-related purposes. Additionally, this study iden­
ment of theoretical models in the face of rapidly tifies how ChatGPT’s information quality – in terms of
evolving technology landscapes. accuracy, completeness, relevance, understandability,
Although PEU and PU, as two main components of value-added, and diagnosticity – impacts customer per­
TAM, have a good power to explain the intention to use ceptions of usefulness. The findings also provide tailored
new technologies, the TAM model should be extended insights for developers and marketers leveraging
by incorporating contextual factors (C.-Y. Lin & Xu, ChatGPT’s distinctive features to boost customer engage­
2021; Rafique et al., 2020). While previous studies have ment and optimize shopping experiences.
 JOURNAL OF MARKETING THEORY AND PRACTICE 3

Literature review Theoretical background

ChatGPT in marketing Various theories have been developed to explain tech­
nology acceptance (Teo et al., 2011). TAM is the most
Chatbots can identify human language using natural
language processing models (Nagarhalli et al., 2020). widely used and validated model for describing accep­
Chatbots are centered on rule-based or self-learning tance of new technologies (Rejali et al., 2023). It was
(AI) methods (Thorat & Jadhav, 2020). ChatGPT is developed by Davis (1989) to explain the determinants
one of the most recent developments in AI-based of the intention to use new technology. Based on TAM,
chatbots (Taecharungroj, 2023), which can produce PU and PEU determine users’ attitudes and intentions
text-based responses (Gordijn & Have, 2023). This to use new technology (Davis, 1989). TAM has demon­
application has been trained on a large dataset and can strated great power to explain technology acceptance in
produce coherent and human-like responses to various various contexts, such as healthcare (Almazroi et al.,
questions and conversations (van Dis et al., 2023). One 2022), education (Alismaiel et al., 2022), finance
of the main strengths of ChatGPT is its ability to execute (Thathsarani & Jianguo, 2022), and marketing
various natural language processing tasks with high (Vrontis et al., 2022).
accuracy and precision (El-Ansari & Beni-Hssane, Although the TAM provides a valuable framework for
2023). Its capabilities include generating content, trans­ describing technology acceptance, contextual factors
lating language, providing shopping tips, and offering must be included to consider the complexity of technol­
personalized product and service recommendations ogy adoption in various contexts (C.-T. Chang et al.,
(Rahaman, 2023). Unlike conventional chatbots, 2017). A number of studies have demonstrated that
ChatGPT can store and retrieve information from pre­ incorporating contextual factors into TAM increases its
vious user interactions, customize the informational predictive power and makes it more applicable to real-
output based on a user’s specific needs, and provide world situations (Doleck et al., 2018; Waris et al., 2022;
personalized responses. Therefore, ChatGPT is Zin et al., 2023). This study extends the TAM by incor­
a valuable tool for various purposes, including customer porating informational factors and perceived enjoyment
service, marketing, education, and entertainment. (Figure 1). This expansion is grounded in the theoretical
Using live chat interfaces to communicate with premises of the Elaboration Likelihood Model (ELM)
customers has become increasingly popular in mar­ (Petty et al., 1986). The ELM is a theory in social psychol­
keting (Adam et al., 2021). Customers use live chat ogy that explains how individuals handle and respond to
services to seek information (e.g. product, price, avail­ persuasive messages (C.-Y. Li, 2013). ELM indicates that
ability) or assistance (e.g. making payments, getting individuals deal with information in two routes; the
a refund) (Adam et al., 2021). With AI, human chat peripheral and central (Filieri & McLeay, 2014). The
services were replaced with conversational software central route, which involves high elaboration, correlates
agents (chatbots) (Pfeuffer et al., 2019). ChatGPT is with users’ cognitive assessment of ChatGPT’s informa­
an AI-powered chatbot that can provide the informa­ tional content (such as accuracy and completeness), while
tion needed to make purchase decisions (Haleem the peripheral route, which involves lower elaboration,
et al., 2023). ChatGPT can provide information aligns with aspects like perceived enjoyment and inter­
about products and their specifications, enabling cus­ face aesthetics. As customers use ChatGPT to gather
tomers to make more informed purchase decisions information relevant to their needs, their elaboration
(Aydın & Karaarslan, 2023). Customers often spend and cognitive evaluation of informational cues are
much time before purchasing, comparing prices, and expected to be high. It means informational cues are
reading reviews to determine which product, brand, expected to influence their attitude and intention to use
and retailer to choose (Maggioni et al., 2020). through the central route, which is a cognitive base (PU).
ChatGPT saves customers time by quickly providing Furthermore, as hedonic factors play a significant role in
relevant shopping information, making purchasing the shopping context, we added perceived enjoyment to
decisions easier (George & George, 2023). address the weakness of TAM in reflecting the influence
Furthermore, ChatGPT can provide personalized pro­ of hedonic factors (Figure 1), thereby providing a more
duct recommendations based on the purchase history holistic depiction of the user experience. The present
and preferences of the customer. ChatGPT assists research aims to present a more comprehensive and
customers by responding to their inquiries (George contextual TAM framework through these augmenta­
& George, 2023). ChatGPT allows customers to make tions. This study argues that informational cues, directed
better decisions and feel more confident about their through the central route of ELM, play a critical role in
purchases (George & George, 2023). shaping technology acceptance (Foroughi, Sitthisirinan,
4 M. IRANMANESH ET AL.

Information
Accuracy

Information Perceived
Completeness Usefulness

Information
Relevance

Perceived Ease of
Information Attitude Intention to Use
Use
Understandability

Information Value-
added

Perceived Control Variables

Information Enjoyment
Diagnosticity Age & Gender

Figure 1. Proposed conceptual framework.

et al., 2024; Harnadi et al., 2024; Won et al., 2023). By the positive association between information accuracy
incorporating a hedonic measure of perceived enjoy­ and PU (Cheung et al., 2019; Priyadarshini et al., 2017;
ment, the study’s framework captures the affective Wu & Zhang, 2014). Accordingly, it is expected that the
dimensions of user interaction. Users’ age and gender accuracy of information provided by ChatGPT plays an
were considered as control variables, and their impacts essential role in shaping customers’ PU. Therefore, we
were controlled. Previous studies have shown that users’ proposed that:
age (Dutta et al., 2023; Muhammad et al., 2024) and
gender (Alnemer, 2022; Niu et al., 2022) may influence H1: Information accuracy positively influences PU.
their intention to use technology. Together, these ele­
ments provide a comprehensive understanding of tech­
nology adoption in environments where users seek to Information completeness
fulfill cognitive needs and derive pleasure.
Information completeness is “the extent to which
information is of sufficient depth and breadth” (Filieri
Hypotheses development & McLeay, 2014, p. 48). Information completeness mea­
sures the extent to which the information provided to
Information accuracy
customers is comprehensive enough to assist them in
Information accuracy refers to the degree to which making an informed decision (Wisker et al., 2019).
information given is perceived to be accurate, correct, Ashfaq et al. (2020) and D.-H. Huang and Chueh
and reliable (Doll & Torkzadeh, 1998). Precise, consis­ (2021) discussed that the completeness of chatbot-
tent, and up-to-date information always leads to faster generated information is vital in making timely deci­
decision-making (Sadrzadehrafiei et al., 2013). In shop­ sions. Individuals expect to receive detailed information
ping, individuals always seek correct, reliable, and up-to on their inquiries when interacting with a chatbot
-date information to make effective purchase decisions (D. M. Nguyen et al., 2021). For example, before making
(C.-H. Chen et al., 2015). With the introduction of AI, a purchase decision, customers always seek complete
companies have employed chatbots in their systems to product information, such as price, functionality, and
provide instant responses to customer queries (C. T. Lee quality (Prasetyo et al., 2021). If a chatbot provides
et al., 2022). AI-based Chatbots can provide accurate unnecessary or incomplete information, individuals
and personalized information by gathering information may perceive it as unhelpful and useless (Hari et al.,
from users’ past inquiries, shopping habits, and prefer­ 2022). Customers are more likely to find a helpful chat­
ences and analyzing online reviews and product features bot when it provides complete information (Brandtzaeg
(M.-H. Huang & Rust, 2021). The functional design of & Følstad, 2017). Similarly, it is expected that if
AI chatbots may positively influence PU (Rietz et al., ChatGPT provides complete information about pro­
2019). The studies in different contexts have confirmed ducts/services/brands, including features, benefits,
 JOURNAL OF MARKETING THEORY AND PRACTICE 5

pricing, and other essential information, customers are Information value-added

more likely to find it helpful. Accordingly, we proposed
Information value-added refers to the degree to which
that:
information is beneficial and offers advantages to users
(R. Y. Wang & Strong, 1996). Online reviews are
H2: Information completeness positively influences
expected to be more valuable than information shared
PU.
by companies, as marketers usually communicate posi­
tive aspects of products, whereas online reviewers com­
Information relevance municate both positive and negative aspects (Chenini &
Touaiti, 2018). It provides customers with comprehen­
Information relevance is determined by the needs of
sive information that may assist them in making better
individuals in specific circumstances and refers to the
decisions (Filieri & McLeay, 2014). Chatbots can pro­
degree to which information is appropriate and useful
vide value-added information to customers by analyzing
for completing a particular task (Filieri & McLeay,
their shopping histories, online reviews, company-
2014). For instance, when purchasing a cell phone,
generated information, and other sources of informa­
customers expect to receive information about price,
tion (Español et al., 2021; S. M. Lee & Lee, 2020).
battery life, features, camera, and network compatibility.
ChatGPT has been trained on a massive amount of
In this case, information relevance measures the extent
information through different topics and can offer
to which provided information covers expected infor­
expert insights and knowledge on various topics.
mation about cell phones. Previous studies have found
Accordingly, ChatGPT may provide useful information
information relevance as a significant driver of PU
and help customers to detect unknown aspects of pro­
(Baker et al., 2010; Y.-C. Liu et al., 2014; Zou, 2019).
ducts or brands by analyzing and processing large
When the information provided by a chatbot is relevant
volumes of data such as customer reviews, social
to a customer’s expectations and needs, they are more
media posts, and other online content. Accordingly,
likely to find the chatbot helpful in making purchase
we proposed that:
decisions (Kull et al., 2021). Similarly, customers are
more likely to find the information provided by
H5: Information value added positively influences PU.
chatGPT useful if it is relevant to their purchase deci­
sions. Accordingly, we proposed that:
Information diagnosticity
H3: Information relevance positively influences PU.
Information diagnosticity refers to “a consumer’s
improved understanding or knowledge about
Information understandability
a product and how it works as a consequence of adopt­
Information understandability refers to the degree to ing information from a specific website” (Filieri, 2015,
which information is easy to understand, interpretable, p. 1262). When the presented information can be used
and readable (Filieri & McLeay, 2014). Individuals will to evaluate products or brands, users are more likely to
find valuable information for decision-making if pre­ find the content useful (Henni et al., 2022; L.-T. Huang,
sented logically and straightforwardly rather than using 2016). Individuals perceive information with low diag­
technical language. Customers expect simple and easy- nosticity as less valuable in making decisions (Kim et al.,
to-understand conversations with chatbots (Dahiya, 2020; Y. Liu & Hu, 2021). AI-powered chatbots can offer
2017; Mehrolia et al., 2023). As a result of the clear, relevant and useful information to customers, which can
simple and step-by-step guidance provided by the chat­ help them evaluate products (Brandtzaeg & Følstad,
bot, customers can understand and follow instructions, 2017). Chatbots, which provide high diagnosticity infor­
which makes it useful to them (Kushwaha et al., 2021). mation, can enhance the overall customer experience by
Customers will likely find ChatGPT information helpful enabling customers to find what they are looking for
for selecting products, services, or brands if it is easy to and make better decisions (Zhang et al., 2023). Natural
understand and interpret. Making informed decisions language processing enables ChatGPT to understand
and having greater confidence in them can be achieved customers’ questions and offer customized recommen­
through simple conversational dialogue. Accordingly, dations. When a customer is looking for a particular
we proposed that: product, ChatGPT can provide highly relevant informa­
tion, such as its features, benefits, and comparison with
H4: Information understandability positively influ­ similar products. As a result of this information, custo­
ences PU. mers may be able to make a more informed purchase
6 M. IRANMANESH ET AL.

decision, increase their confidence in the purchase, and with chatbots increases their likelihood of considering
therefore perceive ChatGPT as a useful tool. them useful (Y.-C. Lee et al., 2021; Richad et al., 2019).
Accordingly, we proposed that: When customers find using ChatGPT to obtain product
and brand information easy, they are more likely to find
H6: Information diagnosticity positively influences it useful and entertaining, resulting in a positive attitude
PU. toward it. Additionally, the simplicity and user-
friendliness of ChatGPT directly contribute to increas­
ing users’ willingness to employ this technology for their
Perceived usefulness
shopping needs. Accordingly, we proposed that:
PU refers to an individual’s belief that using certain
technologies could improve his/her performance H8: PEU positively influences a) PU, b) perceived
(Davis, 1989). PU is one of the important predictors of enjoyment, c) attitude, and d) intention to use
attitudes toward using innovative technology (Cabero- ChatGPT to obtain shopping information.
Almenara et al., 2019; Y.-M. Huang et al., 2012).
Customers’ attitudes toward applications are more posi­
Perceived enjoyment
tive when they perceive them as beneficial when shop­
ping online (Kasilingam, 2020). Al Amin (2022), Perceived enjoyment refers to “the extent to which using
T. T. H. Nguyen et al. (2019), and Vahdat et al. (2021) a given technology is perceived to be personally enjoyable
validated the positive association between PU and atti­ in its own right aside from any instrumental benefits to
tudes. In addition, customers who perceive a particular be derived from using the technology” (Rouibah et al.,
technology as useful are more likely to use it in the 2016, p. 37). Research indicates a strong correlation
future (Al Amin, 2022; Al-Hattami, 2023; An et al., between perceived enjoyment and positive attitudes
2023). AI-powered chatbots can provide fast and effi­ toward technology (Gumulya, 2020; J. Lin, 2022). This
cient customer service, resulting in positive attitudes (Q. relationship becomes particularly salient in the context of
Chen et al., 2023). Furthermore, when customers find AI-driven technologies, such as ChatGPT. ChatGPT can
chatbots to help answer their questions, they are more provide human-like interactions and personalized
likely to have a positive attitude toward them and use responses, making it easier to access information and
them (Zarouali et al., 2018). Accordingly, it is expectable enhance the overall user experience while providing
that if customers find that the information provided by a sense of enjoyment (Foroughi et al., 2023). This positive
ChatGPT helps them make shopping decisions, they are user experience is crucial in forming favorable attitudes
likely to have a positive attitude toward ChatGPT and toward ChatGPT. The conversational and interactive
subsequently use it. Accordingly, we proposed that: nature of ChatGPT, which transcends mere information
retrieval, plays a pivotal role in creating an engaging and
H7: PU positively influences a) attitudes and b) beha­ enjoyable experience. This aligns with findings from
vioral intention to use ChatGPT to obtain shopping Przegalinska et al. (2019), who note the value of custo­
information. mized interactions and emotional support in forming
users’ positive attitudes toward chatbot technologies.
Furthermore, we posit that this enjoyment experi­
Perceived ease of use
enced during interaction with ChatGPT directly influ­
PEU refers to the belief that using a system will be free of ences users’ intention to use the technology. This
mental effort (Siagian et al., 2022). A simple and easy-to statement is supported by studies from M. Li and
-use system is likely to result in positive emotions in Wang (2023) and Marjerison et al. (2022), which high­
customers, such as enjoyment (Cabero-Almenara et al., light the impact of hedonic motivations, such as per­
2019). The ability to interact effectively and efficiently ceived enjoyment, on users’ intention to use AI-based
with less effort leads to greater enjoyment and positive chatbots. Therefore, the enjoyment derived from using
attitudes (Ertz et al., 2022). Previous studies have ChatGPT is not only a determinant of positive attitudes
confirmed the positive effect of PEU on PU (Eneizan but also a significant predictor of users’ behavioral
et al., 2022; Fauzi et al., 2021; Jameel et al., 2021), intentions to use ChatGPT for shopping information.
perceived enjoyment (Wirtz & Göttel, 2016; Accordingly, we proposed that:
Zimmermann et al., 2021), attitude (Ho et al., 2015;
W. M. Huang et al., 2016; Khurshid et al., 2014), and H9: Perceived enjoyment positively influences a) atti­
intention to use (de Andrés-Sánchez & Gené-Albesa, tude and b) intention to use ChatGPT to obtain infor­
2023; X. Wang et al., 2023). The ease of interaction mation about products/services/brands.
 JOURNAL OF MARKETING THEORY AND PRACTICE 7

Attitude the measurement items. For example, information accu­

racy had a Cronbach’s alpha of 0.910, information com­
Attitude can be perceived as the extent to which an
pleteness had 0.833, and information relevance had 0.938.
individual is favorably or unfavorably oriented toward
Similarly, other constructs like information understand­
an object (M. Wang et al., 2021). The behavioral inten­
ability, information value-added, and information diag­
tion to use a system is determined by the attitude toward
nosticity also showed high reliability with Cronbach’s
using it (Patwary et al., 2022; Weerathunga et al., 2021).
alpha values of 0.939, 0.773, and 0.925, respectively.
Favorable attitudes are expected to result in system
usage (Suki et al., 2012). In shopping, a positive attitude
toward a product may lead to future use of the product Sample and data collection
or service (Gumulya, 2020). A favorable attitude toward
Malaysians over the age of 18 comprise the population for
ChatGPT is expected to lead to its use. Accordingly, we
this study. The sample was limited to respondents aged 18
proposed that:
and over to ensure they were in a position to make
purchasing decisions. Due to the limitations of conduct­
H10: Attitude positively influences intention to use
ing an online survey, we used a convenience-based sam­
ChatGPT to obtain shopping information.
pling approach while striving for as diverse a respondent
pool as possible. The survey link was shared on various
Research design social media platforms, specifically targeting online com­
munities and forums visited by Malaysian users inter­
Measurement of constructs
ested in technology and e-commerce. To facilitate
The measurement items of the study were adapted from respondents’ understanding of ChatGPT and its applica­
previous studies to ensure the constructs are reliable and tion in shopping, we included an instructional video at
valid. All items were measured using a Likert scale from the beginning of the survey. This video demonstrated the
1, “Strongly Disagree,” to 5, “Strongly Agree.” The items practical use cases of ChatGPT in comparing products
measuring information accuracy, completeness, rele­ and making purchase recommendations. We also asked
vance, understandability, and value-added were adapted respondents to log into ChatGPT using the link provided
from Filieri and McLeay (2014). Information diagnosti­ and use it to make a purchase decision. We received 453
city’s items were borrowed from Filieri (2015). To mea­ responses, of which 41 were excluded due to the low
sure perceived ease of use and usefulness, this study degree of variation in the replies. This decision was
adapted items from Tandon et al. (2020). Items for made to address potential response set bias, specifically
perceived enjoyment and attitude toward using acquiescence bias and extreme response style
ChatGPT were borrowed from Kasilingam (2020). To (Baumgartner & Steenkamp, 2001; Paulhus, 1991),
measure behavioral intention to use, items were adapted which could compromise the data’s validity (Weijters
from Nikou and Economides (2017). To further validate et al., 2010). To ensure the integrity and reliability of
the questionnaire, three academicians were engaged to our study, we conducted sensitivity analyses to assess
review and pretest the constructs. They independently the impact of this exclusion, finding no significant demo­
evaluated each item for clarity, relevance, and appropri­ graphic or response differences between the excluded and
ateness. Based on their feedback, we revised the ques­ included data sets. We tested the study model using 412
tionnaire. The inter-rater reliability was calculated using usable data. Table 1 presents the profile of the respon­
Cohen’s Kappa coefficient to measure the agreement dents. In this study, data collection was conducted using
among the academicians. The obtained Kappa score a singular source, which necessitates an evaluation for
was 0.82, indicating a strong level of agreement. This common method bias (CMB). To address this, this study
substantial agreement confirms the reliability of the employed two methods to check for CMB. First, we used
questionnaire beyond chance. the unmeasured marker variable technique (Lindell &
The measurement instrument was further refined Whitney, 2001). An unmeasured marker variable was
through a pilot test involving 67 customers in Malaysia. created using SPSS and analyzed with SmartPLS. The
The pilot sample comprised 50.7% females and 49.3% findings revealed that the inclusion of the unmeasured
males, with a diverse age range and representation from marker variable in the research model resulted in an R2
different academic levels. The demographic details change in the dependent variable (intention to use) of
ensured a comprehensive understanding of the sample’s 4.1%, which is below the 10% threshold. Second, we
composition. The constructs’ reliability was demon­ implemented the full collinearity assessment approach
strated by Cronbach’s alpha scores, all exceeding the as recommended by Kock and Lynn (2012). The variance
threshold of 0.7, confirming the internal consistency of inflation factors (VIF) for all variables ranged from
8 M. IRANMANESH ET AL.

Table 1. Profile of respondents.

Demographic Factors Categories Frequency Percentage (%)
Gender Female 217 52.7
Male 195 47.3
Age 18–25 131 31.8
26–35 103 25.0
36–45 94 22.8
46–55 56 13.6
Above 55 28 6.8
Race Malay 216 52.4
Chinese 95 23.1
Indian 62 15.0
Others 39 9.5
Academic Level Diploma 43 10.4
Bachelor 219 53.2
Master 107 26.0
PhD 24 5.8
Others 19 4.6

a minimum of 1.224 to a maximum of 2.578, which are assess the reliability and convergent validity of the mea­
below the threshold of 3.3, indicating that CMB did not surement model (Jozef et al., 2019; Zailani et al., 2020).
significantly influence our findings. These results indicate Analysis of Table 2 reveals that the loadings of all items
that common method bias is unlikely to be a significant surpassed the recommended threshold of 0.7, and both
issue in this dataset. the AVE and CR values for each of the eleven constructs
were deemed acceptable, exceeding the threshold values
of 0.5 and 0.7, respectively.
Data analysis The constructs’ discriminant validity was assessed by
Within the context of this investigation, we have imple­ examining the Fornell-Larcker criterion and the
mented the “Partial Least Squares” (PLS) technique to Heterotrait-Monotrait (HTMT) ratio. In line with the
scrutinize the postulated relationships. PLS was chosen Fornell-Larcker criterion, evidence for discriminant
due to its efficacy in illuminating potential variable validity was observed as the square roots of the AVE
associations, which aligns well with the explanatory exceeded the corresponding inter-construct correlations,
nature of our study (Foroughi, Iranmanesh, et al., as shown in the construct correlation matrix (Table 3).
2024; Hair et al., 2019; Senali et al., 2024). Moreover, Complementing this, as illustrated in Table 4, the find­
the present research used an online tool (Web Power) to ings showed that all HTMT values fell below the prede­
test multivariate normality. The findings specified that fined threshold of 0.85 (Annamalai et al., 2020; Foon
the gathered data are not multivariate normal with et al., 2020), further confirming that the constructs met
Mardia’s multivariate skewness (b = 9.106, p < 0.01) the requirements for discriminant validity.
and kurtosis (b = 69.712, p < 0.01), confirming the
logic for choosing PLS-SEM. PLS is suitable for non-
normal data distributions and, therefore, deemed Structural model
appropriate for our analysis (Nikbin et al., 2014; The proposed model’s ability to explain the variance in
Subramaniam et al., 2020). The present research utilized the endogenous constructs was evaluated in this
a two-step analytical approach to assess the measure­ research. The results show that the R2 values for PU,
ment and structural models (Beh et al., 2021; perceived enjoyment, attitude, and intention to use were
W. F. Chang et al., 2019). The initial step involved 0.662, 0.295, 0.612, and 0.681, respectively, indicating
assessing the measurement’s reliability and validity, a reasonable level of explained variance. As revealed by
while the second step entailed testing the relationships the results of the blindfolding procedure, the Q2 esti­
using a non-parametric bootstrapping technique. mates for PU (0.457), perceived enjoyment (0.208), ATT
(0.498), and intention to use (0.531) surpassed zero,
Results signifying the presence of predictive relevance in the
research (Ponnusamy et al., 2020; Zainuddin et al.,
Measurement model 2021). Based on the outcomes, all proposed associations
This research employed multiple techniques, including were verified, except for the impacts of information
“indicator loadings,” “composite reliability” (CR), and relevance, information understandability, and informa­
“average variance extracted” (AVE) calculations, to tion value-added on PU (Table 5).
 JOURNAL OF MARKETING THEORY AND PRACTICE 9

Table 2. Measurement model evaluation.

Constructs Items Loadings
Information Accuracy The product/service/brand information I obtain from ChatGPT is correct. 0.878
The product/service/brand information I obtain from ChatGPT is accurate. 0.869
The product/service/brand information I obtain from ChatGPT is reliable. 0.857
Information The product/service/brand information I obtain from ChatGPT is of sufficient depth. 0.863
Completeness The product/service/brand information I obtain from ChatGPT is of sufficient breadth. 0.903
Information The product/service/brand information I get through ChatGPT is relevant as it matches my needs. 0.896
Relevance The product/service/brand information I get through ChatGPT is appropriate for satisfying my needs. 0.865
Information The product/service/brand information I obtain from ChatGPT is easy to understand. 0.828
Understandability The product/service/brand information I obtain from ChatGPT is easy to interpret. 0.869
The product/service/brand information I obtain from ChatGPT is easy to read. 0.844
Information Value- The information I obtain from ChatGPT enables me to understand both the positive and the negative aspects of a specific 0.884
Added product/service/brand.
The information I obtain from ChatGPT enables me to detect unknown aspects of a specific product/service/brand. 0.901
Information The information provided by ChatGPT is helpful for me to evaluate products/services/brands. 0.879
Diagnosticity The information provided by ChatGPT is helpful in familiarizing me with products/services/brands. 0.916
The information provided by ChatGPT is helpful for me to understand the performance of the products/services/brands. 0.890
Perceived Usefulness It is easy to learn how to use ChatGPT. 0.844
It is easy to use ChatGPT to find product/service/brand information. 0.824
It is easy to use the information provided by ChatGPT to select a product/service/brand. 0.845
It is easy for me to become skillful at using ChatGPT. 0.843
Overall, I find ChatGPT easy to use. 0.843
Perceived Ease of Use The information provided by ChatGPT will make my shopping faster. 0.798
The information provided by ChatGPT will facilitate the comparison of different products/services/brands. 0.791
The information provided by ChatGPT will make my shopping better. 0.763
The information provided by ChatGPT will help me make better product/service/brand decisions. 0.737
I find The information provided by ChatGPT will be useful in my shopping. 0.781
Perceived Enjoyment I find using ChatGPT for obtaining information about products/services/brands enjoyable. 0.866
The actual process of using ChatGPT for obtaining information about products/services/brands is pleasant. 0.830
I will have fun while using ChatGPT for obtaining information about products/services/brands. 0.849
Attitude Using chatbots for obtaining information about products/services/brands is a good idea. 0.906
I like using chatbots for obtaining information about products/services/brands. 0.903
Using chatbots for obtaining information about products/services/brands would be pleasant. 0.909
Intention to Use I indent to use ChatGPT to obtain information about products/services/brands in the future. 0.925
ChatGPT I plan to use ChatGPT to obtain information about products/services/brands in the future. 0.896
I predict I would use ChatGPT to obtain information about products/services/brands in the future. 0.905
Note: CR: Composite Reliability; AVE: Average Variance Extracted.

Table 3. Fornell-larcker criterion.

Mean SD CR AVE ATT IA IC ID IR IU IVA PE PEU PU UI
ATT 3.502 0.873 0.932 0.821 0.906
IA 3.463 0.888 0.902 0.753 0.619 0.868
IC 3.451 0.896 0.877 0.781 0.583 0.615 0.884
ID 3.572 0.848 0.924 0.801 0.552 0.557 0.447 0.895
IR 3.589 0.902 0.873 0.775 0.409 0.417 0.345 0.535 0.880
IU 3.564 0.930 0.884 0.718 0.499 0.487 0.437 0.381 0.358 0.847
IVA 3.534 0.880 0.887 0.797 0.193 0.213 0.215 0.310 0.370 0.258 0.892
PE 3.457 0.968 0.885 0.719 0.661 0.509 0.451 0.450 0.304 0.382 0.101 0.848
PEU 3.434 0.891 0.882 0.599 0.685 0.677 0.638 0.593 0.472 0.604 0.194 0.543 0.774
PU 3.499 0.881 0.923 0.705 0.684 0.693 0.630 0.644 0.410 0.475 0.202 0.602 0.628 0.840
UI 3.536 0.855 0.934 0.826 0.761 0.643 0.564 0.577 0.441 0.459 0.224 0.659 0.685 0.716 0.909
Note: Diagonal terms (in bold) are square roots of AVE. ATT: attitude; IA: information accuracy; IC: information completeness; ID: information diagnosticity; IR:
information relevance; IU: information understandability; IVA: information value-added; PE: perceived enjoyment; PEU: perceived ease of use; PU: perceive
usefulness; UI: behavioral intention.

Discussion aligns with the peripheral route of ELM, which sug­

ChatGPT can revolutionize shopping decisions by gests that affective components, like enjoyment, can
facilitating access to required information about pro­ greatly impact attitudes and decisions. Information
ducts and brands. Drawing on TAM and ELM, this accuracy, completeness, and diagnosticity are three
study investigated the determinants of using informational cues that significantly influence PU,
ChatGPT to obtain shopping information. The find­ which reflects the central route of ELM, where users
ings showed that all relationships proposed in TAM engage in more elaborate cognitive processing of the
were significant. Furthermore, perceived enjoyment as content provided by ChatGPT. The influences of
a significant driver of attitude depends on PEU. This another three informational cues, namely information
10 M. IRANMANESH ET AL.

Table 4. Heterotrait-monotrait (HTMT).

ATT IA IC ID IR IU IVA PE PEU PU UI
ATT
IA 0.717
IC 0.726 0.793
ID 0.625 0.650 0.558
IR 0.515 0.539 0.473 0.676
IU 0.590 0.595 0.571 0.454 0.473
IVA 0.237 0.270 0.290 0.384 0.511 0.334
PE 0.778 0.619 0.585 0.534 0.401 0.473 0.131
PEU 0.791 0.809 0.819 0.693 0.611 0.741 0.249 0.659
PU 0.765 0.800 0.780 0.726 0.512 0.560 0.247 0.708 0.839
UI 0.825 0.743 0.699 0.651 0.552 0.541 0.274 0.776 0.789 0.800

Table 5. Path coefficients and hypotheses testing.

Hypotheses Relationships Path Coefficients T values P values Decision
H1 IA -> PU 0.240 4.579 0.000** Supported
H2 IC -> PU 0.180 3.678 0.000** Supported
H3 IR -> PU −0.037 1.051 0.294 Not Supported
H4 IU -> PU 0.011 0.270 0.787 Not Supported
H5 IVA -> PU −0.020 0.569 0.569 Not Supported
H6 ID -> PU 0.271 5.313 0.000** Supported
H7a PU -> ATT 0.247 3.322 0.001* Supported
H7b PU -> UI 0.236 3.430 0.001** Supported
H8a PEU -> PU 0.306 4.736 0.000** Supported
H8b PEU -> PE 0.543 9.778 0.000** Supported
H8c PEU -> ATT 0.332 5.141 0.000** Supported
H8d PEU -> UI 0.161 2.819 0.005* Supported
H9a PE -> ATT 0.337 6.185 0.000** Supported
H9b PE -> UI 0.189 3.167 0.001** Supported
H10 ATT-> UI 0.364 4.854 0.000** Supported
Control Variables
– Age -> UI 0.043 1.279 0.202 –
– Gender -> UI 0.031 0.771 0.441 –
Note: *p < 0.01; **p < 0.001.

relevance, information understandability, and infor­ The results indicated a positive association between
mation value-added on PU, were not supported. information completeness and PU. This finding aligns
The positive association between information accu­ with previous studies by D.-H. Huang and Chueh
racy and PU was confirmed. This finding is consistent (2021) and D. M. Nguyen et al. (2021). Besides accuracy,
with previous studies by Cheung et al. (2019) and the depth of responses is important in making well-
Priyadarshini et al. (2017). This finding implies that informed purchasing decisions. Customers need various
customers are likely to find the information provided information about products and brands to make
by ChatGPT useful if it is accurate and up-to-date. As informed decisions, such as price, quality, features, and
products are updated rapidly, customers may make availability. ChatGPT should be trained using various
poor purchasing decisions relying on outdated infor­ marketing data sources such as customer reviews, social
mation. Furthermore, ChatGPT may also result in media, industry reports, official brand websites, and
professional blogs.
poor decisions by providing inaccurate information.
The results show a positive relationship between
The accuracy of information depends on the complete­
information diagnosticity and PU, which is consistent
ness and quality of training information (Du et al.,
with Henni et al. (2022) and L.-T. Huang (2016). The
2023). As such, ChatGPT should improve the accuracy
main purpose of using ChatGPT to obtain information
of responses by regularly refining and updating its about products is to become familiar with products and
training data, enhancing the quality and quantity of their performance and evaluate the products.
training data, and improving its algorithm. It is impor­ Accordingly, information diagnosticity contributes to
tant to note that the current ChatGPT version is ChatGPT’s effectiveness in delivering useful shopping
trained with information until September 2021. information. According to the results, information diag­
ChatGPT should be trained in real-time and frequently nosticity is the most significant driver of PU. Providing
to reflect rapid changes in products and services. quality information and revealing the sources of
 JOURNAL OF MARKETING THEORY AND PRACTICE 11

information may enhance the perception of information affects attitude and intention to use ChatGPT to obtain
diagnosticity (Filieri, 2015). Accordingly, ChatGPT shopping information. It is worthwhile to highlight that
should disclose the sources of information in order to the impact of perceived enjoyment on attitude is higher
enhance information diagnosticity and limit their train­ than utilitarian factors, namely PU and PEU. This finding
ing data to reliable and credible sources. is expectable as shopping is a hedonic behavior, and
The influences of information relevance, information hedonic factors play an important role in shaping it. If
understandability, and information value-added on PU customers enjoy using ChatGPT to obtain information
were not supported. The insignificant effects of infor­ about products and brands, this may enhance their shop­
mation relevance and value-added information contra­ ping experience and result in favorable attitudes toward
dict the findings of Filieri and McLeay (2014) those who ChatGPT and intentions to use it.
find these two attributes significant drivers of adopting
information. Furthermore, Mehrolia et al. (2023) infor­
mation understandability is an important factor in the Theoretical and practical implications
chatbot context. The insignificant influences of these
Theoretical contributions
factors can be attributed to the advanced capabilities of
ChatGPT, which inherently ensure a high level of per­ The present study contributes to the literature on AI-
formance across these informational cues. This leads to powered chatbots in several ways. To the best of our
a potential ceiling effect, wherein the excellence in knowledge, this is the first attempt to explore the deter­
ChatGPT’s performance sets a new standard that users minants of ChatGPT adoption as a tool for obtaining
have come to expect as a given rather than shopping information. We extended the explanatory
a distinguishing feature that enhances PU. This phe­ power of the TAM model by incorporating perceived
nomenon aligns with the concept of a “hygiene factor” enjoyment. While TAM emphasizes PU and PEU as key
in Herzberg’s Two-Factor Theory, where the presence predictors of technology acceptance, our study intro­
of certain factors is not necessarily motivational, but duces a significant extension. We found that perceived
their absence can cause dissatisfaction (Herzberg et al., enjoyment emerges as a crucial predictor of attitudes
2011). Thus, in the context of advanced AI-driven tech­ toward using ChatGPT for obtaining shopping informa­
nologies like ChatGPT, traditional attributes such as tion. This addition to the traditional TAM model under­
information relevance, understandability, and value- scores the importance of hedonic factors in shaping user
added may not exhibit a strong differential impact on attitudes, particularly in contexts where enjoyment
PU because users perceive them as baseline features. plays a pivotal role. Our extended TAM model
However, these factors could become critical when explained 61.2% of the variation in attitude, which is
their performance is lacking and potentially lead to higher than previous studies in the marketing context
a negative influence on PU. This is in accordance with that neglected hedonic factors (e.g. Song et al., 2021).
the principle that users tend to take the presence of Drawing on ELM, the study assessed the influences of
specific favorable attributes for granted over time, informational cues on PU. The findings showed that
which diminishes their influence on decision-making although information accuracy, information complete­
(Foroughi et al., 2019). This reflects the interplay of ness, and information diagnosticity significantly influ­
ELM’s routes in technology acceptance, where the cen­ ence PU, information relevance, information
tral route factors might become less distinguishable in understandability, and information value-added have
highly efficient AI systems. no significant effect on PU. We can conclude from
The results confirmed that TAM is a powerful model comparing significant and insignificant informational
for explaining attitudes and intentions toward using cues that the information characteristics for which pre­
ChatGPT for shopping information. Following TAM, vious reports have questioned ChatGPT’s ability to meet
we found PU, PEU, and attitude to be significant drivers them (information accuracy and information complete­
of intention to use. Furthermore, PEU and PU signifi­ ness) or that customers do not have adequate experience
cantly influence attitude (Foroughi, Yadegaridehkordi, to evaluate them (information diagnosticity) signifi­
et al., 2024). Studies on Chatbots also confirmed the cantly influence PU. In contrast, information character­
power of TAM in explaining attitudes and behaviors istics which ChatGPT has illustrated its ability to fulfill
toward adopting Chatbots (H.-L. Chen et al., 2020; (information relevance, information understandability,
Richad et al., 2019). We added perceived enjoyment to and information value-added) do not affect PU. The
the TAM model to consider the influence of hedonic findings of this study contribute to the ELM literature
factors and extend the explanatory power of TAM. The by showing that the model is not only relevant in tradi­
findings confirmed that perceived enjoyment significantly tional persuasive settings but is also applicable in the
12 M. IRANMANESH ET AL.

realm of technology acceptance, particularly for sophis­ Furthermore, developers can consider “Frequently
ticated AI platforms. This indicates a broader utility of Asked Questions” and customer behavior information
ELM and suggests that its principles can be effectively provided by social media analytical tools to provide
applied to understand user behavior in contexts invol­ more appropriate responses to general inquiries.
ving complex information processing and decision- Information about customer purchase history and pre­
making with AI systems. ferences can also be used to provide customized
responses. Finally, with respect to the importance of
information diagnosticity, text generators should be
Practical contributions
trained with various sources of data and disclose the
This study offers practical implications for develo­ sources of information. Official brand websites, social
pers of ChatGPT and text generator tools and for media, professional blogs, and customer reviews and
marketing managers who wish to incorporate these ratings on e-commerce websites are some sources of
tools into their marketing activities. Our findings information that can provide useful information about
revealed the importance of perceived enjoyment, products and how they work. Disclosing the source of
PEU, and PU in shaping attitudes and behaviors information may assure customers that information is
toward using ChatGPT for obtaining shopping infor­ generated from reliable sources.
mation. PEU derives PU and perceived enjoyment.
As such, developers should simplify the process of
Limitations and future research
using text generators. Marketing managers can pro­
vide training and guidelines on using text generators There are some limitations to the study which provide
to obtain information about products and services. directions for further research. Firstly, the study’s
Text generators can be made easier for customers methodology relied on cross-sectional data which
through training and guidelines. Customers who inherently limits the ability to draw causal inferences.
find ChatGPT easy to use are more likely to find it As such, Longitudinal studies are needed to track how
valuable and entertaining, resulting in a positive atti­ user perceptions and behaviors evolve over time, pro­
tude. Furthermore, developers may incorporate viding a clearer picture of the dynamic interaction
attractive graphics, icons, and images to increase between users and ChatGPT. Secondly, the study did
the platforms’ visual appeal and make the user not account for the potential variability in the levels of
experience more pleasant and enjoyable. information relevance and understandability. While
The findings showed that besides ease of use, infor­ high levels of these factors did not show a significant
mation accuracy, information completeness, and infor­ effect on PU, varying levels may yield different
mation diagnosticity play an essential role in shaping impacts. This suggests that the effect of these informa­
customers’ PU. The significant influence of information tional attributes on PU may be more complex and
accuracy implies that developers of ChatGPT and text warrants further investigation, particularly at different
generators should train the system frequently using levels of information quality. Thirdly, we assessed the
reliable data sources. Products and services are evolving influence of utilitarian and hedonic factors on the
rapidly in the current competitive market. Accordingly, intention to use. Future studies can use “Adaptive
a text generator trained with outdated data cannot pro­ Neuro-Fuzzy Inference Systems” (ANFIS) to demon­
vide accurate and reliable shopping recommendations. strate non-linear relationships between these factors
For instance, a customer may look for a mobile phone and the impacts of information relevance and infor­
with the best camera. ChatGPT, trained with informa­ mation understandability at different levels. Moreover,
tion until September 2021, may not provide accurate the direct and moderating effects of personal factors
recommendations and may mislead the customer. As such as technology anxiety and personal innovative­
such, it is necessary to improve the machine-learning ness were not assessed. Future studies should consider
algorithms of text generators to enable them to be these personal factors to understand how they might
trained in real-time. With respect to the importance of influence user acceptance of AI technologies.
information completeness, customer search capabilities Moreover, this study focuses on specific variables –
and text generation systems’ ability to identify custo­ information accuracy, completeness, relevance, under­
mers’ needs and wants to play a significant role in standability, value-added information, and diagnosti­
providing detailed information to customers. city – to understand the acceptance of ChatGPT for
Customers should receive training on how to ask ques­ shopping. While these variables provide critical
tions effectively so that they can obtain detailed insights into the cognitive aspects influencing user
information. acceptance, the present research did not include
 JOURNAL OF MARKETING THEORY AND PRACTICE 13

other potentially influential factors such as user inter­ education. Heliyon, 9(1), e13141. https://doi.org/10.1016/j.
face design (Jin & Eastin, 2023), personalization (M. Li heliyon.2023.e13141
& Wang, 2023), and social influence (Ma & Huo, Alismaiel, O. A., Cifuentes-Faura, J., & Al-Rahmi, W. M.
(2022). Social media technologies used for education: An
2023). These elements could play a significant role in
empirical study on TAM model during the COVID-19
shaping the user experience and acceptance of AI- pandemic. Frontiers in Education, 7. https://doi.org/10.
driven shopping assistants. Future research should 3389/feduc.2022.882831
explore these additional dimensions to offer a more Almazroi, A. A., Mohammed, F., Al-Kumaim, N. H., &
comprehensive understanding of the factors that drive Hoque, M. R. (2022). An empirical study of factors influencing
user acceptance of advanced AI technologies in shop­ e-health services adoption among public in Saudi Arabia.
ping contexts. By examining personal factors such as Health Informatics Journal, 28(2), 14604582221102316.
https://doi.org/10.1177/14604582221102316
technology anxiety and personal innovativeness, Alnemer, H. A. (2022). Determinants of digital banking adop­
researchers can uncover how these variables influence tion in the Kingdom of Saudi Arabia: A technology accep­
user acceptance, thus enabling marketers to tailor tance model approach. Digital Business, 2(2), 100037.
strategies to different user segments. Additionally, https://doi.org/10.1016/j.digbus.2022.100037
investigating the role of user interface design, perso­ An, S., Eck, T., & Yim, H. (2023). Understanding consumers’
nalization, and social influence on the acceptance of acceptance intention to use mobile food delivery applica­
tions through an extended technology acceptance model.
AI-driven shopping assistants can provide valuable
Sustainability, 15(1), 832. https://doi.org/10.3390/
insights into improving user experiences and lever­ su15010832
aging social proof and influencer marketing strategies. Annamalai, N., Foroughi, B., Iranmanesh, M., & Buathong, S.
Furthermore, future studies should also assess the (2020). Needs and Facebook addiction: How important are
impact of varying levels of information quality on psychological well-being and performance-approach goals?
user acceptance, with the aim of optimizing the infor­ Current Psychology, 39(6), 1942–1953. https://doi.org/10.
mation provided by AI systems to enhance perceived 1007/s12144-019-00516-2
Ashfaq, M., Yun, J., Yu, S., & Loureiro, S. M. C. (2020). I,
usefulness and ease of use. Lastly, we acknowledge
chatbot: Modeling the determinants of users’ satisfaction
that our use of a convenience-based sampling and continuance intention of AI-powered service agents.
approach, while effective for gathering a large and Telematics and Informatics, 54, 101473. https://doi.org/10.
diverse sample, may not provide the targeted specifi­ 1016/j.tele.2020.101473
city of a purposive sample. Future studies should Aydın, Ö., & Karaarslan, E. (2023). Is ChatGPT leading gen­
consider using purposive sampling methods to better erative AI? What is beyond expectations? SSRN Electronic
Journal. https://doi.org/10.2139/ssrn.4341500
target respondents with specific characteristics rele­
Bag, S., Srivastava, G., Bashir, M. M. A., Kumari, S.,
vant to the research, thereby potentially yielding Giannakis, M., & Chowdhury, A. H. (2022). Journey of
more precise and relevant insights. customers in this digital era: Understanding the role of
artificial intelligence technologies in user engagement and
conversion. Benchmarking: An International Journal, 29(7),
Disclosure statement 2074–2098. https://doi.org/10.1108/BIJ-07-2021-0415
Baker, E. W., Al-Gahtani, S. S., & Hubona, G. S. (2010).
No potential conflict of interest was reported by the author(s). Cultural impacts on acceptance and adoption of informa­
tion technology in a developing country. Journal of Global
Information Management (JGIM), 18(3), 35–58. https://doi.
References org/10.4018/jgim.2010070102
Adam, M., Wessel, M., & Benlian, A. (2021). AI-based chat­ Banik, S., & Gao, Y. (2023). Exploring the hedonic factors
bots in customer service and their effects on user affecting customer experiences in phygital retailing. Journal
compliance. Electronic Markets, 31(2), 427–445. https:// of Retailing & Consumer Services, 70, 103147. https://doi.
doi.org/10.1007/s12525-020-00414-7 org/10.1016/j.jretconser.2022.103147
Alam, S. S., Susmit, S., Lin, C.-Y., Masukujjaman, M., & Baumgartner, H., & Steenkamp, J.-B. E. M. (2001). Response
Ho, Y.-H. (2021). Factors affecting augmented reality adop­ styles in marketing research: A cross-national investigation.
tion in the retail industry. Journal of Open Innovation: Journal of Marketing Research, 38(2), 143–156. https://doi.
Technology, Market, and Complexity, 7(2), 142. https:// org/10.1509/jmkr.38.2.143.18840
doi.org/10.3390/joitmc7020142 Beh, P. K., Ganesan, Y., Iranmanesh, M., & Foroughi, B.
Al Amin, M. (2022). The influence of psychological, situa­ (2021). Using smartwatches for fitness and health monitor­
tional and the interactive technological feedback-related ing: The UTAUT2 combined with threat appraisal as
variables on customers’ technology adoption to use online moderators. Behaviour & Information Technology, 40(3),
shopping applications. Journal of Global Marketing, 35(5), 282–299. https://doi.org/10.1080/0144929X.2019.1685597
384–407. https://doi.org/10.1080/08911762.2022.2051157 Brandtzaeg, P. B., & Følstad, A. (2017). Why people use
Al-Hattami, H. M. (2023). Understanding perceptions of aca­ chatbots. Internet Science: 4th International Conference,
demics toward technology acceptance in accounting INSCI 2017 (pp. 377–392), Thessaloniki, Greece.
14 M. IRANMANESH ET AL.

Cabero-Almenara, J., Fernández-Batanero, J. M., & Barroso- Du, H., Teng, S., Chen, H., Ma, J., Wang, X., Gou, C., Li, B.,
Osuna, J. (2019). Adoption of augmented reality technol­ Ma, S., Miao, Q., Na, X., Ye, P., Zhang, H., Luo, G., &
ogy by university students. Heliyon, 5(5), e01597. https:// Wang, F.-Y. (2023). Chat with chatgpt on intelligent vehi­
doi.org/10.1016/j.heliyon.2019.e01597 cles: An IEEE tiv perspective. IEEE Transactions on
Chang, C.-T., Hajiyev, J., & Su, C.-R. (2017). Examining the Intelligent Vehicles, 8(3), 2020–2026. https://doi.org/10.
students’ behavioral intention to use e-learning in 1109/TIV.2023.3253281
Azerbaijan? The general extended technology acceptance Dutta, A., Kovid, R. K., Drave, V. A., & Bhatia, M. S. (2023).
model for e-learning approach. Computers & Education, Internet of things adoption: Unpacking the role of per­
111, 128–143. https://doi.org/10.1016/j.compedu.2017.04.010 ceived brand credibility. Global Knowledge, Memory &
Chang, W. F., Amran, A., Iranmanesh, M., & Foroughi, B. Communication, 1–22. https://doi.org/10.1108/GKMC-05-
(2019). Drivers of sustainability reporting quality: Financial 2023-0160
institution perspective. International Journal of Ethics and El-Ansari, A., & Beni-Hssane, A. (2023). Sentiment analysis
Systems, 35(4), 632–650. https://doi.org/10.1108/IJOES-01- for personalized chatbots in E-Commerce applications.
2019-0006 Wireless Personal Communications, 129(3), 1623–1644.
Chen, C.-H., Nguyen, B., Klaus, P. P., & Wu, M.-S. (2015). https://doi.org/10.1007/s11277-023-10199-5
Exploring electronic word-of-mouth (eWOM) in the con­ Eneizan, B., Obaid, T., Abumandil, M. S. S., Mahmoud, A. Y.,
sumer purchase decision-making process: The case of Abu-Naser, S. S., Arif, K., & Abulehia, A. F. S. (2022).
online holidays–evidence from United Kingdom (UK) con­ Acceptance of Mobile banking in the era of COVID-19.
sumers. Journal of Travel & Tourism Marketing, 32(8), International Conference on Information Systems and
953–970. https://doi.org/10.1080/10548408.2014.956165 Intelligent Applications: ICISIA 2022 (pp. 29–42).
Chen, H.-L., Vicki Widarso, G., & Sutrisno, H. (2020). Ertz, M., Jo, M.-S., Kong, Y., & Sarigöllü, E. (2022). Predicting
A chatbot for learning Chinese: Learning achievement m-shopping in the two largest m-commerce markets: The
and technology acceptance. Journal of Educational United States and China. International Journal of Market
Computing Research, 58(6), 1161–1189. https://doi.org/10. Research, 64(2), 249–268. https://doi.org/10.1177/
1177/0735633120929622 14707853211023036
Chen, Q., Lu, Y., Gong, Y., & Xiong, J. (2023). Can AI chatbots Español, S. J., Fronda, R., Jr., Villarica, M. R., & Intal, G. L.
help retain customers? Impact of AI service quality on (2021). SHOP BUDDY enhanced automated Chatbot for
customer loyalty. Internet Research, 33(6), 2205–2243. any E-commerce website with data analytics functionalities.
https://doi.org/10.1108/INTR-09-2021-0686 International Conference on Industrial Engineering and
Chenini, A., & Touaiti, M. (2018). Building destination loyalty Operations Management (pp. 1–13).
using tourist satisfaction and destination image: A holistic Fauzi, A., Wandira, R., Sepri, D., & Hafid, A. (2021).
conceptual framework. Journal of Tourism, Heritage & Exploring students’ acceptance of Google classroom during
Services Marketing, 4(2), 37–43. the COVID-19 pandemic by using the technology accep­
Cheung, M. L., Chau, K. Y., Lam, M. H. S., Tse, G., Ho, K. Y., tance model in west sumatera universities. Electronic
Flint, S. W., Broom, D. R., Tso, E. K. H., & Lee, K. Y. (2019). Journal of E-Learning, 19(4), 233–240. https://doi.org/10.
Examining consumers’ adoption of wearable healthcare 34190/ejel.19.4.2348
technology: The role of health attributes. International Filieri, R. (2015). What makes online reviews helpful? A
Journal of Environmental Research and Public Health, 16 diagnosticity-adoption framework to explain informational
(13), 2257. https://doi.org/10.3390/ijerph16132257 and normative influences in e-WOM. Journal of Business
Dahiya, M. (2017). A tool of conversation: Chatbot. Research, 68(6), 1261–1270. https://doi.org/10.1016/j.
International Journal of Computer Sciences & Engineering, jbusres.2014.11.006
5(5), 158–161. Filieri, R., & McLeay, F. (2014). E-WOM and accommodation.
Davis, F. D. (1989). Perceived usefulness, perceived ease of Journal of Travel Research, 53(1), 44–57. https://doi.org/10.
use, and user acceptance of information technology. MIS 1177/0047287513481274
Quarterly, 13(3), 319–340. https://doi.org/10.2307/249008 Foon, P. Y., Ganesan, Y., Iranmanesh, M., & Foroughi, B. (2020).
Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User Understanding the behavioural intention to dispose of unused
acceptance of computer technology: A comparison of two medicines: An extension of the theory of planned behaviour.
theoretical models. Management Science, 35(8), 982–1003. Environmental Science and Pollution Research, 27(22),
https://doi.org/10.1287/mnsc.35.8.982 28030–28041. https://doi.org/10.1007/s11356-020-09125-0
de Andrés-Sánchez, J., & Gené-Albesa, J. (2023). Assessing Foroughi, B., Iranmanesh, M., & Hyun, S. S. (2019).
attitude and behavioral intention toward chatbots in an Understanding the determinants of mobile banking con­
insurance setting: A mixed method approach. tinuance usage intention. Journal of Enterprise Information
International Journal of Human–Computer Interaction, Management, 32(6), 1015–1033. https://doi.org/10.1108/
1–16. https://doi.org/10.1080/10447318.2023.2227833 JEIM-10-2018-0237
Doleck, T., Bazelais, P., & Lemay, D. J. (2018). Is a general Foroughi, B., Iranmanesh, M., Nilashi, M., Ghobakhloo, M.,
extended technology acceptance model for e-learning Asadi, S., & Khoshkam, M. (2024). Determinants of fol­
generalizable? Knowledge Management and E-Learning, 10 lowers’ purchase intentions toward brands endorsed by
(2), 133–147. social media influencers: Findings from PLS and fsQCA.
Doll, W. J., & Torkzadeh, G. (1998). Developing Journal of Consumer Behaviour, 23(2), 888–914. https://doi.
a multidimensional measure of system-use in an organiza­ org/10.1002/cb.2252
tional context. Information & Management, 33(4), Foroughi, B., Senali, M. G., Iranmanesh, M., Khanfar, A.,
171–185. https://doi.org/10.1016/S0378-7206(98)00028-7 Ghobakhloo, M., Annamalai, N., & Naghmeh-
 JOURNAL OF MARKETING THEORY AND PRACTICE 15

Abbaspour, B. (2023). Determinants of intention to use Huang, L.-T. (2016). Exploring utilitarian and hedonic ante­
ChatGPT for educational purposes: Findings from cedents for adopting information from a recommendation
PLS-SEM and fsQCA. International Journal of Human– agent and unplanned purchase behaviour. New Review of
Computer Interaction, 1–20. https://doi.org/10.1080/ Hypermedia and Multimedia, 22(1–2), 139–165. https://
10447318.2023.2226495 doi.org/10.1080/13614568.2015.1052098
Foroughi, B., Sitthisirinan, S., Iranmanesh, M., Nikbin, D., & Huang, M.-H., & Rust, R. T. (2021). A strategic framework for
Ghobakhloo, M. (2024). Determinants of travel apps con­ artificial intelligence in marketing. Journal of the Academy
tinuance usage intention: Extension of technology continu­ of Marketing Science, 49(1), 30–50. https://doi.org/10.1007/
ance theory. Current Issues in Tourism, 27(4), 619–635. s11747-020-00749-9
https://doi.org/10.1080/13683500.2023.2169109 Huang, W. M., Chen, C. H., Lin, H. R., & Sun, C. F. (2016).
Foroughi, B., Yadegaridehkordi, E., Iranmanesh, M., Decision support system for material management in
Sukcharoen, T., Ghobakhlo, M., & Nilashi, M. (2024). intensive care units: An empirical study in Taiwan.
Determinants of continuance intention to use food delivery International Journal of Healthcare Management, 9(2),
apps: Findings from PLS and fsQCA. International Journal 102–109. https://doi.org/10.1080/20479700.2015.1101912
of Contemporary Hospitality Management, 36(4), Huang, Y.-C., Li, L.-N., Lee, H.-Y., Browning, M. H. E. M., &
1235–1261. https://doi.org/10.1108/IJCHM-10-2022-1209 Yu, C.-P. (2023). Surfing in virtual reality: An application of
George, A. S., & George, A. S. H. (2023). A review of ChatGPT extended technology acceptance model with flow theory.
AI’s impact on several business sectors. Partners Universal Computers in Human Behavior Reports, 9, 100252. https://
International Innovation Journal, 1(1), 9–23. doi.org/10.1016/j.chbr.2022.100252
Gordijn, B., & Have, H. (2023). ChatGPT: Evolution or Huang, Y.-M., Huang, Y.-M., Huang, S.-H., & Lin, Y.-T.
revolution? Medicine, Health Care and Philosophy, 26(1), (2012). A ubiquitous English vocabulary learning system:
1–2. https://doi.org/10.1007/s11019-023-10136-0 Evidence of active/passive attitudes vs. usefulness/ease-of-
Gumulya, D. (2020). The role of perceived enjoyment in use. Computers & Education, 58(1), 273–282. https://doi.
people’s attitude to accept toy and equipment renting for org/10.1016/j.compedu.2011.08.008
children: A comparative study between people who have Jameel, A. S., Khald Hamzah, A., Raad Al-Shaikhli, T., Ihsan
been using the service versus those who have never used the Alanssari, A., & Ibrahim, M. K. (2021). System character­
rental service. Management Science Letters, 10(9), istics and behavioural intention to use E-Learning.
2119–2130. https://doi.org/10.5267/j.msl.2020.1.017 Learning, 12(10), 7724–7733. https://doi.org/10.17762/tur
Hair, J. F., Risher, J. J., Sarstedt, M., & Ringle, C. M. (2019). comat.v12i10.5738
When to use and how to report the results of PLS-SEM. Jiang, C., Rashid, R. M., & Wang, J. (2019). Investigating the
European Business Review, 31(1), 2–24. https://doi.org/10. role of social presence dimensions and information support
1108/EBR-11-2018-0203 on consumers’ trust and shopping intentions. Journal of
Haleem, A., Javaid, M., & Singh, R. P. (2023). An era of Retailing & Consumer Services, 51, 263–270. https://doi.
ChatGPT as a significant futuristic support tool: A study org/10.1016/j.jretconser.2019.06.007
on features, abilities, and challenges. BenchCouncil Jin, E., & Eastin, M. (2023). Gender bias in virtual doctor
Transactions on Benchmarks, Standards and Evaluations, interactions: Gender matching effects of chatbots and
2(4), 100089. https://doi.org/10.1016/j.tbench.2023.100089 users on communication satisfactions and future intentions
Hari, H., Iyer, R., & Sampat, B. (2022). Customer brand to use the chatbot. International Journal of Human–
engagement through chatbots on bank websites–examining Computer Interaction, 1–13. https://doi.org/10.1080/
the antecedents and consequences. International Journal of 10447318.2023.2279402
Human–Computer Interaction, 38(13), 1212–1227. https:// Jozef, E., Kumar, K. M., Iranmanesh, M., & Foroughi, B.
doi.org/10.1080/10447318.2021.1988487 (2019). The effect of green shipping practices on multi­
Harnadi, B., Widiantoro, A. D., & Prasetya, F. X. H. (2024). national companies’ loyalty in Malaysia. The International
Investigating the behavioral differences in the acceptance of Journal of Logistics Management, 30(4), 974–993. https://
MOOCs and E-learning technology. Computers in Human doi.org/10.1108/IJLM-01-2019-0005
Behavior Reports, 14, 100403. https://doi.org/10.1016/j. Kasilingam, D. L. (2020). Understanding the attitude and
chbr.2024.100403 intention to use smartphone chatbots for shopping.
Henni, S., Franz, P., Staudt, P., Peukert, C., & Weinhardt, C. Technology in Society, 62, 101280. https://doi.org/10.1016/
(2022). Evaluation of an interactive visualization tool to j.techsoc.2020.101280
increase energy literacy in the building sector. Energy & Khurshid, A., Rizwan, M., & Tasneem, E. (2014). Factors
Buildings, 266, 112116. https://doi.org/10.1016/j.enbuild. contributing towards adoption of E-banking in Pakistan.
2022.112116 International Journal of Accounting and Financial
Herzberg, F., Mausner, B., & Snyderman, B. B. (2011). The Reporting, 1(1), 437–455. https://doi.org/10.5296/ijafr.
motivation to work. Transaction Publishers. v4i2.6584
Ho, C. K. Y., Ke, W., & Liu, H. (2015). Choice decision of Kim, M., Han, J., & Jun, M. (2020). Do same-level review
e-learning system: Implications from construal level theory. ratings have the same level of review helpfulness? The role
Information & Management, 52(2), 160–169. https://doi. of information diagnosticity in online reviews. Information
org/10.1016/j.im.2014.07.003 Technology & Tourism, 22(4), 563–591. https://doi.org/10.
Huang, D.-H., & Chueh, H.-E. (2021). Chatbot usage inten­ 1007/s40558-020-00191-1
tion analysis: Veterinary consultation. Journal of Kock, N., & Lynn, G. (2012). Lateral collinearity and mislead­
Innovation & Knowledge, 6(3), 135–144. https://doi.org/ ing results in variance-based SEM: An illustration and
10.1016/j.jik.2020.09.002 recommendations. Journal of the Association for
16 M. IRANMANESH ET AL.

Information Systems, 13(7), 546–580. https://doi.org/10. Ma, X., & Huo, Y. (2023). Are users willing to embrace
17705/1jais.00302 ChatGPT? Exploring the factors on the acceptance of chat­
Kull, A. J., Romero, M., & Monahan, L. (2021). How may bots from the perspective of AIDUA framework.
I help you? Driving brand engagement through the warmth Technology in Society, 75, 102362. https://doi.org/10.1016/
of an initial chatbot message. Journal of Business Research, j.techsoc.2023.102362
135, 840–850. https://doi.org/10.1016/j.jbusres.2021.03.005 Maggioni, I., Sands, S. J., Ferraro, C. R., Pallant, J. I.,
Kushwaha, A. K., Kumar, P., & Kar, A. K. (2021). What Pallant, J. L., Shedd, L., & Tojib, D. (2020). Consumer
impacts customer experience for B2B enterprises on using cross-channel behaviour: Is it always planned?
AI-enabled chatbots? Insights from big data analytics. International Journal of Retail & Distribution
Industrial Marketing Management, 98, 207–221. https:// Management, 48(12), 1357–1375. https://doi.org/10.1108/
doi.org/10.1016/j.indmarman.2021.08.011 IJRDM-03-2020-0103
Landim, A., Pereira, A. M., Vieira, T., de Costa, E., Manolică, A., Guță, A.-S., Roman, T., & Dragăn, L. M. (2021).
Moura, J. A. B., Wanick, V., & Bazaki, E. (2022). Chatbot Is consumer overchoice a reason for decision paralysis?
design approaches for fashion E-commerce: An interdisci­ Sustainability, 13(11), 5920. https://doi.org/10.3390/
plinary review. International Journal of Fashion Design, su13115920
Technology and Education, 15(2), 200–210. https://doi. Marjerison, R. K., Zhang, Y., & Zheng, H. (2022). AI in
org/10.1080/17543266.2021.1990417 E-Commerce: Application of the use and gratification
Lee, C. T., Pan, L.-Y., & Hsieh, S. H. (2022). Artificial intelli­ model to the acceptance of chatbots. Sustainability, 14
gent chatbots as brand promoters: A two-stage structural (21), 14270. https://doi.org/10.3390/su142114270
equation modeling-artificial neural network approach. Mehrolia, S., Alagarsamy, S., Moorthy, V., & Jeevananda, S.
Internet Research, 32(4), 1329–1356. https://doi.org/10. (2023). Will users continue using banking chatbots? The
1108/INTR-01-2021-0030 moderating role of perceived risk. FIIB Business Review,
Lee, S. M., & Lee, D. (2020). “Untact”: A new customer service 1–19. https://doi.org/10.1177/23197145231169900
strategy in the digital age. Service Business, 14(1), 1–22. Muhammad, G., Siddiqui, M. S., Rasheed, R., Shabbir, H., &
https://doi.org/10.1007/s11628-019-00408-2 Sher, R. F. (2024). Role of external factors in adoption of
Lee, Y.-C., Yamashita, N., & Huang, Y. (2021). Exploring the HR analytics: Does statistical background, gender and age
effects of incorporating human experts to deliver journal­ matters? Journal of Business Analytics, 7(1), 1–14. https://
ing guidance through a chatbot. Proceedings of the ACM on doi.org/10.1080/2573234X.2023.2231966
Human-Computer Interaction, 5(CSCW1), 1–27. https:// Nagarhalli, T. P., Vaze, V., & Rana, N. K. (2020). A review of
doi.org/10.1145/3449196 current trends in the development of chatbot systems. 2020
Li, C.-Y. (2013). Persuasive messages on information system 6th International Conference on Advanced Computing and
acceptance: A theoretical extension of elaboration likeli­ Communication Systems (ICACCS) (pp. 706–710).
hood model and social influence theory. Computers in Nguyen, D. M., Chiu, Y.-T. H., & Le, H. D. (2021).
Human Behavior, 29(1), 264–275. https://doi.org/10.1016/ Determinants of continuance intention towards banks’
j.chb.2012.09.003 chatbot services in Vietnam: A necessity for sustainable
Li, M., & Wang, R. (2023). Chatbots in e-commerce: The development. Sustainability, 13(14), 7625. https://doi.org/
effect of chatbot language style on customers’ continuance 10.3390/su13147625
usage intention and attitude toward brand. Journal of Nguyen, T. T. H., Nguyen, N., Nguyen, T. B. L., Phan, T. T. H.,
Retailing & Consumer Services, 71, 103209. https://doi. Bui, L. P., & Moon, H. C. (2019). Investigating consumer
org/10.1016/j.jretconser.2022.103209 attitude and intention towards online food purchasing in
Lin, C.-Y., & Xu, N. (2021). Extended TAM model to explore an emerging economy: An extended tam approach. Foods, 8
the factors that affect intention to use AI robotic architects (11), 576. https://doi.org/10.3390/foods8110576
for architectural design. Technology Analysis & Strategic Nikbin, D., Hyun, S. S., Albooyeh, A., & Foroughi, B. (2014).
Management, 34(3), 349–362. https://doi.org/10.1080/ Effects of perceived justice for coaches on athletes’ satisfac­
09537325.2021.1900808 tion, commitment, effort, and team unity. International
Lin, J. (2022). The effects of gamification instruction on the Journal of Sport Psychology, 45(2), 100–120.
roles of perceived ease of learning, enjoyment, and useful Nikou, S. A., & Economides, A. A. (2017). Mobile-based
knowledge toward learning attitude. Turkish Online assessment: Integrating acceptance and motivational fac­
Journal of Educational Technology-TOJET, 21(2), 81–91. tors into a combined model of self-determination theory
Lindell, M. K., & Whitney, D. J. (2001). Accounting for com­ and technology acceptance. Computers in Human Behavior,
mon method variance in cross-sectional research designs. 68, 83–95. https://doi.org/10.1016/j.chb.2016.11.020
Journal of Applied Psychology, 86(1), 114–121. https://doi. Niu, Z., Chen, C., Gao, Y., Wang, Y., Chen, Y., & Zhao, K.
org/10.1037/0021-9010.86.1.114 (2022). Peer effects, attention allocation and farmers’ adop­
Liu, Y.-C., Huang, Y.-A., & Wang, M.-Y. (2014). Exploring tion of cleaner production technology: Taking green con­
users’intention to use health information on a bulletin trol techniques as an example. Journal of Cleaner
board system. International Journal of Cyber Society and Production, 339, 130700. https://doi.org/10.1016/j.jclepro.
Education, 7(1), 89–96. https://doi.org/10.7903/ijcse.1166 2022.130700
Liu, Y., & Hu, H. (2021). Online review helpfulness: The Patwary, A. K., Rasoolimanesh, S. M., Rabiul, M. K.,
moderating effects of review comprehensiveness. Aziz, R. C., & Hanafiah, M. H. (2022). Linking environ­
International Journal of Contemporary Hospitality mental knowledge, environmental responsibility, altruism,
Management, 33(2), 534–556. https://doi.org/10.1108/ and intention toward green hotels through ecocentric and
IJCHM-08-2020-0856 anthropocentric attitudes. International Journal of
 JOURNAL OF MARKETING THEORY AND PRACTICE 17

Contemporary Hospitality Management, 34(12), Rietz, T., Benke, I., & Maedche, A. (2019). The impact of
4653–4673. https://doi.org/10.1108/IJCHM-01-2022-0039 anthropomorphic and functional chatbot design features
Paulhus, D. L. (1991). Measurement and control of response in enterprise collaboration systems on user acceptance.
bias. In J. P. Robinson, P. R. Shaver, & L. S. Wrightsman Bioorganic & Medicinal Chemistry Letters, 29(6), 786–790.
(Eds.), Measures of personality and social psychological atti­ https://doi.org/10.1016/j.bmcl.2019.01.028
tudes (pp. 17–59). Academic Press. https://doi.org/10.1016/ Rouibah, K., Lowry, P. B., & Hwang, Y. (2016). The effects of
B978-0-12-590241-0.50006-X perceived enjoyment and perceived risks on trust forma­
Petty, R. E., Cacioppo, J. T., Petty, R. E., & Cacioppo, J. T. tion and intentions to use online payment systems: New
(1986). The elaboration likelihood model of persuasion. perspectives from an Arab country. Electronic Commerce
Springer. Research and Applications, 19, 33–43. https://doi.org/10.
Pfeuffer, N., Adam, M., Toutaoui, J., Hinz, O., & Benlian, A. 1016/j.elerap.2016.07.001
(2019). Mr. And Mrs. Conversational agent-gender stereo­ Sadrzadehrafiei, S., Chofreh, A. G., Hosseini, N. K., &
typing in judge-advisor systems and the role of egocentric Sulaiman, R. (2013). The benefits of enterprise resource
bias. planning (ERP) system implementation in dry food packa­
Ponnusamy, S., Iranmanesh, M., Foroughi, B., & Hyun, S. S. ging industry. Procedia Technology, 11, 220–226. https://
(2020). Drivers and outcomes of Instagram addiction: doi.org/10.1016/j.protcy.2013.12.184
Psychological well-being as moderator. Computers in Sagnier, C., Loup-Escande, E., Lourdeaux, D., Thouvenin, I.,
Human Behavior, 107, 107. https://doi.org/10.1016/j.chb. & Valléry, G. (2020). User acceptance of virtual reality: An
2020.106294 extended technology acceptance model. International
Prasetyo, Y. T., Tanto, H., Mariyanto, M., Hanjaya, C., Journal of Human–Computer Interaction, 36(11),
Young, M. N., Persada, S. F., Miraja, B. A., & 993–1007. https://doi.org/10.1080/10447318.2019.1708612
Redi, A. A. N. P. (2021). Factors affecting customer satis­ Senali, M. G., Iranmanesh, M., Ghobakhloo, M., Foroughi, B.,
faction and loyalty in online food delivery service during Asadi, S., & Rejeb, A. (2024). Determinants of trust and
the COVID-19 pandemic: Its relation with open purchase intention in social commerce: Perceived price
innovation. Journal of Open Innovation: Technology, fairness and trust disposition as moderators. Electronic
Market, and Complexity, 7(1), 76. https://doi.org/10.3390/ Commerce Research and Applications, 64, 101370. https://
doi.org/10.1016/j.elerap.2024.101370
joitmc7010076
Siagian, H., Tarigan, Z., Basana, S., & Basuki, R. (2022). The
Priyadarshini, C., Sreejesh, S., & Anusree, M. R. (2017). Effect
effect of perceived security, perceived ease of use, and
of information quality of employment website on attitude
perceived usefulness on consumer behavioral intention
toward the website: A moderated mediation study.
through trust in digital payment platform. International
International Journal of Manpower, 38(5), 729–745.
Journal of Data & Network Science, 6(3), 861–874. https://
https://doi.org/10.1108/IJM-12-2015-0235
doi.org/10.5267/j.ijdns.2022.2.010
Przegalinska, A., Ciechanowski, L., Stroz, A., Gloor, P., &
Song, H., Ruan, W. J., & Jeon, Y. J. J. (2021). An integrated
Mazurek, G. (2019). In bot we trust: A new methodology
approach to the purchase decision making process of
of chatbot performance measures. Business Horizons, 62(6),
food-delivery apps: Focusing on the TAM and AIDA
785–797. https://doi.org/10.1016/j.bushor.2019.08.005 models. International Journal of Hospitality Management,
Rafique, H., Almagrabi, A. O., Shamim, A., Anwar, F., & 95, 102943. https://doi.org/10.1016/j.ijhm.2021.102943
Bashir, A. K. (2020). Investigating the acceptance of mobile Subramaniam, P. L., Iranmanesh, M., Kumar, K. M., &
library applications with an extended technology accep­ Foroughi, B. (2020). The impact of multinational corpora­
tance model (TAM). Computers & Education, 145, tions’ socially responsible supplier development practices
103732. https://doi.org/10.1016/j.compedu.2019.103732 on their corporate reputation and financial performance.
Rahaman, M. (2023). Can ChatGPT Be your friend? Emergence International Journal of Physical Distribution & Logistics
of entrepreneurial research. Emergence of Entrepreneurial Management, 50(1), 3–25. https://doi.org/10.1108/
Research. February 18, 2023. IJPDLM-01-2019-0002
Rejali, S., Aghabayk, K., Esmaeli, S., & Shiwakoti, N. (2023). Suki, N. M., Ramayah, T., & Ly, K. K. (2012). Empirical
Comparison of technology acceptance model, theory of investigation on factors influencing the behavioral inten­
planned behavior, and unified theory of acceptance and tion to use facebook. Universal Access in the Information
use of technology to assess a priori acceptance of fully Society, 11(2), 223–231. https://doi.org/10.1007/s10209-
automated vehicles. Transportation Research Part A: 011-0248-6
Policy and Practice, 168, 103565. https://doi.org/10.1016/j. Szász, L., Bálint, C., Csíki, O., Nagy, B. Z., Rácz, B.-G.,
tra.2022.103565 Csala, D., & Harris, L. C. (2022). The impact of
Rese, A., Ganster, L., & Baier, D. (2020). Chatbots in retailers’ COVID-19 on the evolution of online retail: The pandemic
customer communication: How to measure their accep­ as a window of opportunity. Journal of Retailing &
tance? Journal of Retailing & Consumer Services, 56, Consumer Services, 69, 103089. https://doi.org/10.1016/j.
102176. jretconser.2022.103089
Richad, R., Vivensius, V., Sfenrianto, S., & Kaburuan, E. R. Taecharungroj, V. (2023). “What can ChatGPT do?” analyz­
(2019). Analysis of factors influencing millennial’s technol­ ing early reactions to the innovative AI chatbot on Twitter.
ogy acceptance of chatbot in the banking industry in Big Data and Cognitive Computing, 7(1), 1–10. https://doi.
Indonesia. International Journal of Civil Engineering & org/10.3390/bdcc7010035
Technology, 10(3), 1270–1281. https://doi.org/10.34218/ Tandon, A., Dhir, A., Kaur, P., Kushwah, S., & Salo, J. (2020).
IJM.10.3.2019.011 Behavioral reasoning perspectives on organic food
18 M. IRANMANESH ET AL.

purchase. Appetite, 154, 104786. https://doi.org/10.1016/j. e-learning among university students: The role of precipi­
appet.2020.104786 tating events. Education Sciences, 11(8), 436. https://doi.
Teo, T., Faruk Ursavaş, Ö., & Bahçekapili, E. (2011). org/10.3390/educsci11080436
Efficiency of the technology acceptance model to explain Weijters, B., Geuens, M., & Schillewaert, N. (2010). The sta­
pre‐service teachers’ intention to use technology: A Turkish bility of individual response styles. Psychological Methods,
study. Campus-Wide Information Systems, 28(2), 93–101. 15(1), 96–110. https://doi.org/10.1037/a0018721
https://doi.org/10.1108/10650741111117798 Wirtz, B. W., & Göttel, V. (2016). Technology acceptance in
Thathsarani, U. S., & Jianguo, W. (2022). Do digital finance social media: Review, synthesis and directions for future
and the technology acceptance Model strengthen financial empirical research. Journal of Electronic Commerce
inclusion and SME performance? Information, 13(8), 390. Research, 17(2), 97.
https://doi.org/10.3390/info13080390 Wisker, Z. L., Kadirov, D., & Bone, C. (2019). Modelling P2P
Thorat, S. A., & Jadhav, V. (2020). A review on implementa­ Airbnb online host advertising effectiveness: The role of
tion issues of rule-based chatbot systems. Proceedings of the emotional appeal information completeness creativity and
International Conference on Innovative Computing & social responsibility. International Journal of Culture,
Communications (ICICC). Tourism and Hospitality Research, 13(4), 505–523. https://
Turri, A. M., & Watson, A. (2022). Product assortment, choice doi.org/10.1108/IJCTHR-03-2019-0045
overload, and filtering technology across retail contexts. Won, D., Chiu, W., & Byun, H. (2023). Factors influencing
The International Review of Retail, Distribution & consumer use of a sport-branded app: The technology accep­
Consumer Research, 33(3), 219–239. https://doi.org/10. tance model integrating app quality and perceived enjoyment.
1080/09593969.2022.2056904 Asia Pacific Journal of Marketing & Logistics, 35(5),
Vahdat, A., Alizadeh, A., Quach, S., & Hamelin, N. (2021). 1112–1133. https://doi.org/10.1108/APJML-09-2021-0709
Would you like to shop via mobile app technology? The Wu, B., & Zhang, C. (2014). Empirical study on continuance
technology acceptance model, social factors and purchase intentions towards E-Learning 2.0 systems. Behaviour &
intention. Australasian Marketing Journal, 29(2), 187–197. Information Technology, 33(10), 1027–1038. https://doi.
https://doi.org/10.1016/j.ausmj.2020.01.002 org/10.1080/0144929X.2014.934291
van Dis, E. A. M., Bollen, J., Zuidema, W., van Rooij, R., & Zailani, S., Iranmanesh, M., Foroughi, B., Kim, K., &
Bockting, C. L. (2023). ChatGPT: Five priorities for
Hyun, S. S. (2020). Effects of supply chain practices, inte­
research. Nature, 614(7947), 224–226. https://doi.org/10.
gration and closed-loop supply chain activities on
1038/d41586-023-00288-7
cost-containment of biodiesel. Review of Managerial
Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D.
Science, 14(6), 1299–1319. https://doi.org/10.1007/s11846-
(2003). User acceptance of information technology:
019-00332-9
Toward a unified view. MIS Quarterly, 27(3), 425–478.
Zainuddin, Z., Iranmanesh, M., Tseng, M., Foroughi, B., &
https://doi.org/10.2307/30036540
Tengku Hamzah, T. A. A. (2021). Clean development
Vrontis, D., Massoud, M., Dennaoui, H., & Nemar, S. E.
mechanism implementation: External and organizational
(2022). The impact of e-service on hotels’ booking:
Adjusted TAM framework for customers’ intentions to factors drives expected business benefits. Business Strategy
book hotels online. Global Business & Economics Review, and the Environment, 30(8), 3444–3453. https://doi.org/10.
26(3), 285–313. https://doi.org/10.1504/GBER.2022.122385 1002/bse.2812
Wang, M., Jin, Z., Fan, S., Ju, X., & Xiao, X. (2021). Chinese Zarouali, B., Van den Broeck, E., Walrave, M., & Poels, K.
residents’ preferences and consuming intentions for hotels (2018). Predicting consumer responses to a chatbot on
after COVID-19 pandemic: A theory of planned behaviour facebook. Cyberpsychology, Behavior and Social
approach. Anatolia, 32(1), 132–135. https://doi.org/10. Networking, 21(8), 491–497. https://doi.org/10.1089/cyber.
1080/13032917.2020.1795894 2017.0518
Wang, R. Y., & Strong, D. M. (1996). Beyond accuracy: What Zhang, J., Zhu, Y., Wu, J., & Yu-Buck, G. F. (2023). A natural
data quality means to data consumers. Journal of apology is sincere: Understanding chatbots’ performance in
Management Information Systems, 12(4), 5–33. https:// symbolic recovery. International Journal of Hospitality
doi.org/10.1080/07421222.1996.11518099 Management, 108, 103387. https://doi.org/10.1016/j.ijhm.
Wang, X., Luo, R., Liu, Y., Chen, P., Tao, Y., & He, Y. (2023). 2022.103387
Revealing the complexity of users’ intention to adopt Zimmermann, S. K., Wagner, H.-T., Rössler, P., Gewald, H., &
healthcare chatbots: A mixed-method analysis of antece­ Krcmar, H. (2021). The Role of Utilitarian Vs. Hedonic
dent condition configurations. Information Processing & Factors for the Adoption of AI-Based Smart Speakers.
Management, 60(5), 103444. https://doi.org/10.1016/j.ipm. Proceedings of the 27th Americas Conference on
2023.103444 Information Systems, Montreal, Canada.
Waris, I., Ali, R., Nayyar, A., Baz, M., Liu, R., & Hameed, I. Zin, K. S. L. T., Kim, S., Kim, H.-S., & Feyissa, I. F. (2023).
(2022). An empirical evaluation of customers’ adoption of A study on technology acceptance of digital healthcare
drone food delivery services: An extended technology among older Korean adults using extended tam (extended
acceptance model. Sustainability, 14(5), 2922. https://doi. technology acceptance Model). Administrative Sciences, 13
org/10.3390/su14052922 (2), 42. https://doi.org/10.3390/admsci13020042
Weerathunga, P. R., Samarathunga, W., Rathnayake, H. N., Zou, W. (2019). An integrated theoretical model of E-Wom
Agampodi, S. B., Nurunnabi, M., & Madhunimasha, M. adoption—a case in hotel industry. Oklahoma State
(2021). The COVID-19 pandemic and the acceptance of University.