The current issue and full text archive of this journal is available on Emerald Insight at:

https://www.emerald.com/insight/2040-7122.htm

Strengthening
Strengthening consumer–brand consumer–
relationships through avatars brand
relationships
Mohamed H. Elsharnouby
Faculty of Business and Economics, Badr University in Cairo, Cairo, Egypt and
Faculty of Commerce, Cairo University, Giza, Egypt 581
Chanaka Jayawardhena Received 2 February 2022
Surrey Business School, University of Surrey, Guildford, UK Revised 30 April 2022
21 July 2022
Hongfei Liu 16 September 2022
Accepted 17 September 2022
Southampton Business School, University of Southampton, Southampton, UK, and
Alaa M. Elbedweihy
Department of Business Administration, Cairo University Faculty of Commerce,
Cairo, Egypt

Abstract
Purpose – Avatars have become increasingly prevalent on brand websites, yet their impact on consumers’
use of these sites remains underexplored. The current study focuses on avatars, which are three-dimensional
animated graphical web interfaces that verbally aid the brand stakeholders (e.g. customers, employees and
suppliers). Avatars provide administrative and technical information through the brand website. Drawing
upon the stimuli–organism–response (S-O-R) paradigm, this research examines the impact of avatars as an
information provision and interacting tool (vs a traditional format) on consumers’ perceptions, attitudes and
behaviors toward a brand. It also investigates the roles of familiarity with avatar use and the language used
by an avatar in shaping consumers’ responses.
Design/methodology/approach – Across two laboratory experiments, the authors examined and
confirmed causal relationships between the use of avatars (vs a traditional format) on a website and
attitudinal and behavioral constructs.
Findings – We show that avatars (vs written information) had a significant effect on controlling
information. The users in our experiments had greater control over the information provided when it was
presented as text on a website compared to the case of avatars “telling” the information. Different languages
and familiarity with avatar use also affected the consumers’ hedonism in terms of website use.
Originality/value – We advance the understanding of avatar use in website design, particularly avatars’ verbal
interaction, in shaping consumers’ cognitive, affective, attitudinal and behavioral responses and add important
empirical evidence to the growing body of research and practices involving avatar use in interactive marketing.
Keywords Avatar elements, Apply intention, WOM, Information recall, Hedonic, Usefulness, Attitude
toward a brand, Familiarity
Paper type Research paper

Introduction
Avatars are digital entities with anthropomorphic appearance that are controlled by a human or
by software (Miao et al., 2022). In practice, avatars offer interactive touch points that may be
verbal (voice) or nonverbal (text and animation) (Liew et al., 2017). Since their introduction as
interactive features, avatars have become increasingly prevalent in the design of websites.
Brands have incorporated avatars into their websites for a variety of reasons, including
customer relationship management and facilitating purchases, increasing the probability to
purchase from the brand (e.g. Choi et al., 2020; Foster et al., 2022). For example, HSBC has two Journal of Research in Interactive
well-known virtual assistants, called “XiaoLingTong” and “Olivia.” Interactive marketing Marketing
Vol. 17 No. 4, 2023
emphasizes the significance of bidirectional communication in which avatars play pp. 581-601
© Emerald Publishing Limited
irreplaceable roles in enriching the customer experience (Foster et al., 2022; Wang, 2021). 2040-7122
DOI 10.1108/JRIM-02-2022-0035
JRIM Previous research mainly focused on understanding consumers’ general perceptions of
17,4 avatars (e.g. Liew et al., 2017). However, as many brands are now using avatars on their
websites, the implications of consumer–avatar interaction remain underexplored.
Pioneering researchers and practitioners highlighted the importance of understanding
customers’ responses elicited by interaction with avatars, thereby improving the
effectiveness of website designs that incorporate these tools. The outcomes of customer–
avatar interactions are important to understand. Unlike static website features, an avatar
582 can communicate in real time, offering real-time information – in that participants’ attitudes
toward the brand in both the avatar and the avatar based on text designs are higher than
those for the written style design – and interacting like a human. Such interaction is distinct
from navigating traditional websites, and the underlying process of how consumers
respond to avatars needs to be examined specifically.
Given the importance of avatars in interactive marketing research and practice, our
research aims to understand how avatars help brands strengthen consumer–brand
relationships. Drawing from the existing research, we identified several important
research gaps that informed and shaped our research objectives. Our research objective is
three-fold. First, despite the growth in the use of avatars, relatively little academic inquiry
to date has explored the brand avatar factors that influence consumers’ perceptions and
behaviors (Crolic et al., 2022). More specifically, as avatars can be incorporated in different
ways into website designs (e.g. with text and with verbal communication), which elements
of an avatar, from a user perspective, are more effective remains an important and
unexplored line of inquiry (e.g. Miao et al., 2022). Therefore, our first research objective
was to examine the effectiveness of avatar use by comparing consumers’ responses when
having information provided as written text on a website, verbally through an avatar and
through an avatar based on text.
Second, avatars are responsible for providing information and assisting brand–consumer
communication, but how avatars contribute to the effectiveness of this interaction poses
a need for urgent inquiry (Hoyet et al., 2019; Lee and Lee, 2006). Therefore, our second
research objective focuses on examining the impact of the use of avatars on consumers’
perceptions of website design. Here, we specify three dimensions of website design (i.e. clarity,
control and convenience) in order to examine the influence of avatars.
Third, previous research highlighted that there is a need to examine human–avatar
interactions because avatars might be perceived as more competent in the utilitarian realm
than in the hedonic (Borau et al., 2021). Previous research also called for an examination of
possible mitigators that may assist consumers’ willingness to follow avatar advice (Lin et al.,
2021) and expand the interaction. Based on this call, our third objective was to understand
whether consumers’ familiarity with using avatars and the language an avatar uses might
mitigate the impact of avatars on those consumers’ responses. In relation to consumers’
responses to marketing stimuli, in addition to purchase intention, word of mouth (WOM) is
one of the most powerful information sources and exerts a strong influence on product
evaluation by consumers (e.g. Sheth, 2021). However, a positive and rewarding consumption
experience is an important requirement (Kohler et al., 2011). Therefore, our inquiry contributes to
understanding the role of avatars in shaping consumer experiences and responses (Farzin and
Fattahi, 2018; Park et al., 2021).
Against this background, the aim of this study was to examine the causal relationships
of interacting and providing information tools (avatars, avatars based on text and
a traditional format [no avatar/information as text]) on (1) the consumer/user perception
including the tool’s elements (clarity, control and convenience) that reflect the perceived
ease of use (e.g. Chen, 2019) and (2) affective (hedonism), (3) cognitive (information provision
and usefulness), (4) attitudinal (attitudes toward brands) and (5) purchase-related and
voluntary behavior (e.g. WOM) outcomes. This research also examines the roles of
consumers’ familiarity with using avatars and the language avatars use in strengthening Strengthening
the relationships between avatars and consumers’ responses. Understanding this influence consumer–
is valuable as consumers become more likely to anticipate having avatars provide
information (e.g. Keeling et al., 2010) and interact with them (Foster et al., 2022).
brand
Our research contributes to the interactive marketing literature. For example, our work relationships
makes contributions to the increasing body of literature that examines how consumers
interact with brands through new media and tools on websites (e.g. Lim and Childs, 2020;
Wang, 2021), showing the significant outcomes of adding an avatar as a new communication 583
tool. Our study also extends previous research that focused on examining chat avatars (e.g.
Lin et al., 2021). We instead used avatars that are highly anthropomorphized (using real
human images and voices) and interacted with users verbally.

Theoretical background and hypotheses development

The S-O-R paradigm (Mehrabian and Russell, 1974) provides an orientation for understanding
users’ cognitive, emotional and behavioral responses in the online environment (e.g.
Rodríguez-Torrico et al., 2021; Sohn et al., 2020). We considered the S-O-R paradigm to be
a suitable theoretical underpinning for the current study. By following this paradigm, this
study frames stimuli as the conditions for providing information (avatars, avatars based on
text and a traditional format of no avatar/information as text). These stimuli determine
a user’s (organism) cognitive and affective reaction processes (the user’s perceptions,
including the elements of the tool, in terms of clarity, control and convenience, hedonism,
usefulness, information recall and attitude toward the brand), which leads to responses that
are desirable as behavioral outcomes (intention to apply and WOM). The current research
study extends the applicability of the S-O-R paradigm into relationship marketing by
investigating the role of using avatars to explain customers’ emotional, cognitive and
conative responses to consumer–brand interaction. We discuss the stimulus, organism and
response in relation to this study below.

Avatars
There are multiple terminologies associated with avatars, such as chatbot, virtual assistant,
virtual agent and conversational agent embodiment (e.g. Aljukhadar and Senecal, 2011; Han,
2021). Many brands choose to humanize their consumer service chatbots by giving them
names and avatars (Crolic et al., 2022). Avatars are virtual characters that present human-like
features, such as a face and body parts (Han, 2019). In this research, we focus on avatars as
“digital entities with anthropomorphic appearance, controlled by a human or software that are
able to interact” (Miao et al., 2022, p. 5). In other words, avatars are three-dimensional (3D)
animated graphical web interfaces that imitate the brand’s representatives by providing
information and responding to users’/visitors’ queries orally and visually (Elsharnouby, 2015).

Avatar elements
Clarity. Clarity is one of the key ease-of-use elements of an avatar. The clarity of an avatar
refers to the extent to which the avatar’s voice, facial cues and the information provided are
clear (Elsharnouby, 2015). An avatar provides information on users’ websites in an oral
method. The logical order of the information provided should also be taken into
consideration. Previous research showed the importance of clarity in similar settings. For
example, the role of clarity in the virtual environment is affected by the avatar–consumer
interaction in terms of content, such as functional and social content (Kohler et al., 2011). In
addition, the type of virtual agent (two-dimensional embodied agents vs 3D embodied
agents) – a similar concept to an avatar – has highly significant effects on the clarity of
JRIM the agent’s voice (McBreen and Jack, 2001). Voices using only an audio format were also
17,4 found to be significantly clearer than images with facial expressions (McBreen and Jack,
2001). Based on previous research and Stæhr’s (2008) study, which found learners’ receptive
vocabulary size to be strongly associated with their reading ability and moderately
associated with their listening ability, we postulate the following:
H1a. The degree of clarity is affected more by an avatar based on text than having only
584 an avatar or a written information design.

Control. In organizations, resources and employees are typically controlled by managers.

Likewise, in classrooms, teachers have control over students in the form of grades and
possess greater resources in the form of knowledge (Rucker et al., 2011). In a virtual reality
context, users can control both the speed and direction of avatars’ motions, very easily and
intuitively (Oshita, 2006). Creating controllable, responsive avatars is a significant problem
and is challenging in virtual settings (Lee and Lee, 2006). This requirement for control
means using a motion capture system to offer an easy solution to interactions simply by
transferring the movements of a performer to an animated avatar in real time (Butt et al.,
2021; Hoyet et al., 2019). However, the current research examines avatars as part of
a brand’s website. Thus, the control issue is related to the degree of difficulty users face
in accomplishing their tasks or receiving complete information from an avatar when having
control problems. Offering more choices in terms of interaction modes (e.g. dealing with
front-desk human staff) may increase brand consumers’ sense of control (Choi et al., 2020).
Previous research that compared reading and listening to information found that listeners
did not seem to be able to take as much advantage of repetition, showing the less
controllable aspects of a sound system (Vidal, 2011). Thus, we postulate:
H1b. The degree of control is affected more by an avatar based on text than having
only an avatar or a written information design.

Convenience. There is increasing evidence that convenience is a central element for customers
when encountering service. Previous studies related to convenience have long been concerned
with consumer expenditure of time and effort (e.g. Gottschalk, 2020). Providing smart
technologies conveys convenient and instant digitalized services to meet brand consumers’
expectations of high technology (Choi et al., 2020). Convenience refers to the level of effort and
time required by the user to use an avatar to obtain the information needed from a website
(Elsharnouby, 2015). The less effort and time required, the more convenient the avatar.
Convenience is related to the appropriateness of avatars as a tool for providing information.
Providing more choices in terms of interaction modes (e.g. dealing with front-desk human staff)
may provide greater convenience for a brand’s consumers (Choi et al., 2020). Thus,
H1c. The degree of convenience is affected more by an avatar based on text than
having only an avatar or a written information design.

Outcomes of avatar use

Outcomes related to avatars and websites. Information recall. One of the main outcomes of
using an avatar is information recall, which refers to the degree to which a user remembers the
information provided by the avatar (Elsharnouby, 2015). Previous research investigated how
technology helps users to recall information from the desired web page (Kiesel et al., 2018).
Recall also varied in relation to the information recall resulting from reading or listening to the
information provided. For example, people receiving written information usually recall
significantly more than people receiving verbal information. In a language-learning setting,
the girls recalled the information better than the boys (Langdon et al., 2002). However, the girls
were equally proficient at the recall task for all three language modes (listening, oral reading Strengthening
and silent reading), with the boys being equally good at the task after listening and after oral consumer–
reading but markedly poorer after silent reading (Johnson, 1982). In addition, a memory of the
user recalls the information after viewing the visualization (Kotlarek et al., 2020). Based on the
brand
above discussion, we postulate that: relationships
H2a. The consumer’s perception of information recall is affected more by an avatar
based on text than having only an avatar or a written information design. 585
Hedonism. An avatar can be perceived as a hedonic tool. Hedonism includes the joy
and excitement of shopping/visiting (e.g. Wakefield and Baker, 1998). In the current
research, hedonism indicates users’ feelings of enjoyment or fun derived from using an
avatar. When the users’ thinking style is primarily experiential (including hedonic
activities) rather than rational (including utilitarian activities), they perceive an
avatar as a hedonic tool (e.g. Han, 2021). The excitement might come from the
newness and animation of having a speaking 3D avatar (advanced technology) rather
than a simple image. Using realistic avatars might have a positive effect on customers’
affective responses, such as enjoyment and entertainment (Butt et al., 2021; Han, 2021;
Miao et al., 2022).
H2b. The consumer’s perception of hedonism is affected more by an avatar based on
text than having only an avatar or a written information design.

Usefulness. A website’s users might frame an avatar as a utilitarian tool (Etemad-Sajadi and
Ghachem, 2015). In this case, the perceived usefulness is more likely to be high. Perceived
usefulness refers to the degree to which a website’s user describes an avatar as being useful in
the sense of performing its tasks (Elsharnouby, 2015). Kohler et al. (2009), for example,
showed the usefulness of avatars in different product types (e.g. banking and
entertainment) through playing various roles such as giving instructions on ATM screens.
Artificial intelligence (AI)-powered avatars are considered useful, as highlighted by positive
results regarding gamers’ attitudes (Butt et al., 2021). The avatars might be perceived as more
competent in the utilitarian (usefulness) realm (Borau et al., 2021; Miao et al., 2022). Thus,
H2c. The consumer’s perception of usefulness is affected more by an avatar based on
text than having only an avatar or a written information design.

Outcomes related to brand. Attitude toward a brand. A conversational human voice

appeared to be a key factor in enhancing positive attitudes toward brands that used social
media platforms to communicate with their audience (e.g. Van Noort and Willemsen, 2012).
It is expected that an avatar enhances attitudes toward a brand since it speaks in a real-life
voice. Previous research showed that face-to-face communication produces a greater
positive attitude change toward a brand than does audio or video (Williams, 1977). The
presence of an avatar has positive impacts, including on attitude toward a product (e.g. Jin
and Bolebruch, 2009; Keeling et al., 2010). Thus,
H3a. The consumer’s attitude toward a brand is affected more by an avatar based on
text than having only an avatar or a written information design.

Intention to apply. Attitudes play a vital role in determining consumers’ behaviors

(Kroesen et al., 2017). Attitude is considered a type of behavioral intention since it refers
to the degree to which an avatar’s presence can encourage potential consumers to apply for
and join a brand’s services. A number of previous studies showed that an avatar-enabled
user interface had an impact on purchase intention in online shopping (e.g. Holzwarth et al.,
JRIM 2006). In addition, the presence of a spokesperson cartoon advertisement can increase the
17,4 purchase intention for a brand (Heiser et al., 2008). Other results also confirmed a positive
indirect effect of a chatbot on consumers’ purchase intentions through perceived enjoyment
of chatbot commerce (Han, 2021). Thus, we hypothesize:
H3b. The consumer’s intention to apply is affected more by an avatar based on text
than having only an avatar or a written information design.
586
Word of mouth. WOM is becoming increasingly predominant (e.g. Sheth, 2021).
Consumers engage in communicating their experiences with a brand with other
consumers in terms of their desire for social interaction and enhancing their self-worth,
which leads to electronic WOM (eWOM) behavior (e.g. Kohler et al., 2011). WOM can be
positive or negative, although negative word of mouth (NWOM) has stronger effects than
Positive Word of Mouth in terms of reach and impact (e.g. Roy et al., 2021). Hence, brands
call for appropriate strategies to control NWOM and its potential damage (Van Noort and
Willemsen, 2012). In the current research, WOM refers to the degree to which adding an
avatar motivates the users of a website to talk positively or negatively about the brand.
After an avatar is added as a new feature on a brand website, users potentially spread
positive words about this feature. Previous studies showed that information influence had
a positive and significant effect on eWOM (e.g. Farzin and Fattahi, 2018). Thus,
H3c. WOM is affected more by an avatar based on text than having only an avatar or
a written information design.

Own language and familiarity

From a linguistic perspective, listening to a second language, in comparison to reading it,
poses a significant problem, in that some listeners are more likely to grasp only words and
others might try to construct a sensible context to organize what little they are able to
perceive and decode (Lund, 1991). In addition, a reader, compared to a listener, can pause
over new words and look around in the text for contextual cues that a listener might well
miss. The listener of a second language who attends to a single word will miss the following
parts of the message (Lund, 1991). Language barriers can prevent effective communication
clarity between a brand and its consumers (Reeves et al., 2005). Previous studies have
shown that language can influence consumers’ perceptions and attitudes (e.g. Kronrod et al.,
2012; Luna et al., 2003).
A user’s familiarity with an avatar is a key issue and can play an important role in
improving its use as a tool on a website. It is expected that people who are unfamiliar with
avatar technology are more likely to use the traditional format of a website, whether texts,
images and/or videos. Thus, familiarity with an avatar helps the website audience use it
more easily, thereby improving information recall (e.g. Hauge et al., 2013). Consumers’
familiarity with a product/tool was found to be positively correlated with preferences for
and choice of a product/tool with greater functionality (Thompson et al., 2009). This means
that familiarity with an avatar can increase the perceived usefulness that consumers have
of it. In addition, familiarity as a result of frequent encounters with talking avatars may
potentially moderate the effects of avatars on online shopping perceptions and behaviors
(Liew et al., 2017). Based on the above discussion related to language and familiarity, it is
predicted that:
H4. When the user is familiar with avatar technology using their own language, an
avatar based on text will have a greater effect on the study constructs (clarity,
control, convenience, information recall, hedonism, usefulness, attitude toward
brand and WOM), as observed in similar previous research.
Study one Strengthening
Design and participants consumer–
This study aimed to compare the impact of three designs for providing information (i.e.
a written information style as the control condition and, as the two experimental conditions,
brand
an avatar and an avatar based on text) on visitors to a website (i.e. students) in relation to the relationships
design elements (i.e. clarity, control and convenience), information recall, hedonism,
usefulness, attitude toward the brand and apply intention. A within-subject experimental
design (repeated measures) was selected to conduct this study. Each participant was exposed 587
to the three designs (i.e. written information, an avatar and an avatar based on text).
A fictitious brand was used to exclude other effects that might appear, such as brand
loyalty, thereby leading to more valid results (Jin and Bolebruch, 2009). Based on Cohen’s
(1992) and Dattalo’s (2008) studies of power and effect size, considering the required sample
sizes for power level 0.8 (recommended), the effect size taken in this research was 0.5. This
effect size suggested an ideal sample size of 28 participants. Thirty-nine university students
participated in this within-group experimental study. Responses from two participants were
omitted as their questionnaires were not valid.

Materials and procedures

A website for a fictitious university brand was developed. Each web page included some
information about the brand – in particular, study topics (programs) and frequently asked
questions. Each web page was designed with one specific condition for providing
information (an avatar, an avatar based on text or written information) as the stimulus.
The information was uploaded to the avatar on the developed website. Instruction files were
prepared and printed to give to participants during the experiment sessions. In addition,
two specific computer laboratories were booked to conduct all the experiment sessions.
The online questionnaire was transformed and used in this study. When a participant
attended the laboratory, he/she received a general instruction sheet and a specific file that
included the main instructions for the experiments. Figure 1(a) shows a web page of the
avatar condition, Figure 1(b) shows a web page of the avatar based on text condition and
Figure 1(c) shows a web page of the written information condition.
Randomization was used to prevent the order effect (Field and Hole, 2010). On average,
every six participants were allocated to each condition. After receiving, reading and
following the instructions, the participants used a link to a questionnaire and answered
questions based on the design to which they were exposed. After that, participants took 2
minutes to rest before going to the next condition. Each participant completed the three
conditions. Since the influence of an avatar increases in cases of difficult purchase
processes, complicated products and limited buyer’s knowledge of the product (e.g.
Holzwarth et al., 2006), applying this research to university brands was valuable because
the process of comparing universities to choose the most appropriate institution is not easy.
Therefore, an avatar would be likely to play an important role in students’ decision-making
process.

Measures and pretest

The survey contained eight items relating to clarity (adapted from Freling et al., 2011), five
items to convenience (adapted from Keh and Pang, 2010; Seiders et al., 2005), three items to
usefulness (adapted from Sheinin et al., 2011), five items to hedonism (adapted from Voss
et al., 2003), four items to information recall, six items to attitude toward the brand, six items
to control and four items to apply intention, which were developed from the qualitative
results (Elsharnouby, 2015). The questionnaire was tested on a small sample of 10 students.
JRIM
17,4

588

Figure 1.
One web page from
each scenario
(continued )
 Strengthening
consumer–
brand
relationships

589

Figure 1.
Notes: (a) Avatar presence – study one, (b) avatar based on text – study one and (c) no avatar – study one

Data analysis and results

Instrument validation. Table 1 shows means and Cronbach’s alphas for all the constructs.
The Cronbach’s alphas were above 0.7 for all constructs. The constructs were also assessed
for convergent and discriminant validity through Confirmatory Factor Analysis using
AMOS. We checked the unidimensionality of each construct and the measurement model
(see Table 1 for the remaining items), in accordance with Janssens et al. (2008), for goodness
of fit (χ2(271) = 326.441, p = 0.012; goodness-of-fit index [GFI] = 0.83; comparative fit index
[CFI] = 0.977; Tucker–Lewis index [TLI] = 973; root mean square error of approximation
[RMSEA] = 0.043; standardized root mean residual [SRMR] = 0.0629). Table 1 shows that
the estimates of all the factor loadings are greater than 0.5, all the t-values are greater than
1.96 (loadings ranged from 6.453 to 14.687), all composite reliability (CR) values are above
0.7 and all average variance extracted (AVE) estimates are above 0.5. Thus, all four criteria
provided support to the convergent validity of the constructs (Fornell and Larcker, 1981;
Janssens et al., 2008). Table 2 shows that all the square roots of AVE estimates on the
diagonal are greater than the construct correlations with another factor, providing evidence
of the discriminant validity of all the constructs.
Repeated measures analysis of variance (ANOVA). Clarity, control and convenience. As
shown in Table 3, there are no significant differences between the three designs in respect of
clarity, the convenience of the information provided or usefulness. These results suggest
that although the information was provided using three different designs, clarity and
convenience did not differ as the participants understood most of this information; thus, H1a
and H1c are not supported. Repeated measures analysis of variance (ANOVA) revealed
significant differences between an avatar design (M = 4.523, F(2,72) = 3.862, p < 0.05) and
a written style design (M = 5.261, F(2,72) = 3.862, p < 0.05), in that participants had greater
control over the information provided in the latter type of design compared to the former. In the
control condition (written style), the participants obtained all the information they needed
JRIM STD Cronbach’s
17,4 Constructs Mean loadings t-Value alpha AVE CR

Clarity 5.49 0.864 0.634 0.870

The (design) is
Unclear concerning the language/clear 0.821 12.024
concerning the language
590 Unclear/clear 0.943
Vague/well-defined 0.815 11.849
Indistinct/distinct 0.554 6.453
Control 4.98 0.870 0.671 0.859
The (design) allows me to control
The provision of information 0.864
The search process 0.793 9.689
Getting the difficult words to get their meaning, 0.798 9.789
for example from a dictionary
Convenience 5.27 0.860 0.714 0.908
It is easy to deal with the (design) 0.871 13.668
The time required to receive the required 0.921 15.554
information is appropriate
I am able to get to the information from the 0.909
(design) quickly
Getting information from the (design) requires 0.651 8.064
little effort
Hedonism 4.22 0.865 0.802 0.942
I feel about the (design) is
Not fun/fun 0.880
Dull/exciting 0.924 14.687
Not delightful/delightful 0.916 14.408
Not thrilling/thrilling 0.861 12.607
Usefulness 5.47 0.861 0.700 0.823
The (design) provides relevant information 0.767
The (design) does a good job of presenting the 0.901 9.273
information
Information recall 4.74 0.855 0.873 0.932
I can remember
Most information provided by the (design) 0.893
The information provided by the (design) easily 0.974 15.256
Attitude toward brand 5.12 0.875 0.811 0.944
After I navigate the website, I feel that the brand is
Technologically developed 0.963 10.465
Adapting new ways 0.961 10.435
Having paperless work 0.727
Modern 0.929 10.066
Apply intention 4.20 0.878 0.582 0.805
I would join the brand 0.852
Table 1. I would join the brand as getting information 0.651 6.769
Means, loadings, through the (design) is easy
t-values, Cronbach’s The chance of joining the brand is high 0.772 8.014
alphas, AVE and CR
(study one) Note: STD, standardized

compared to the experimental condition (only listening to the information via an avatar).
However, the two experimental conditions (avatar and avatar based on text) did not differ
significantly (p > 0.05), as shown in Table 3 and Figure 2. Thus, H1b is partially supported.
 Attitude
Strengthening
Information toward Apply consumer–
Clarity Control Convenience Hedonism Usefulness recall brand intention brand
Clarity 0.796 relationships
Control 0.565 0.819
Convenience 0.633 0.729 0.845
Hedonism 0.443 0.434 0.519 0.896 591
Usefulness 0.720 0.688 0.809 0.504 0.837
Information 0.618 0.677 0.715 0.571 0.705 0.934
recall
Attitude 0.388 0.165 0.347 0.690 0.420 0.508 0.900
toward Table 2.
brand Correlations and
Apply 0.391 0.407 0.344 0.485 0.371 0.447 0.610 0.763 square roots of the
intention AVE (study one)

Significance between conditions

Avatar and
Avatar, Avatar based Written avatar based on Avatar and Avatar based on text
Construct M on text, M style, M text written style and written style

Clarity 5.432 5.716 5.331 0.209 0.689 0.424

Control 4.523 5.144 5.261 0.074* 0.047* 1.000
Convenience 5.460 5.466 4.892 0.565 0.110 0.783
Information 4.757 5.135 4.324 0.196 0.128 0.016*
recall
Hedonism 4.662 4.710 3.297 1.000 0.000* 0.000*
Usefulness 5.527 5.554 5.324 1.000 0.254 0.312
Attitude 5.595 5.750 4.007 1.000 0.000* 0.000*
toward
brand Table 3.
Apply 4.225 4.441 3.928 0.380 0.268 0.030* Results of repeated
intention measures ANOVA
Note: * p < 0.05 (study one)

Information recall, hedonism and usefulness. For the information recall construct,
although the avatar condition did not differ from the avatar based on text and the control
conditions (written style), there was a significant difference between the avatar based on
text and the written style, in that participants recalled the information provided by an
avatar based on text; thus, H2a is partially supported. Concerning the hedonism construct,
the results show significant differences between the two experimental conditions – avatar
design (M = 4.662; F(2,72) = 13.492, p < 0.01) and avatar based on text design (M = 4.710; F
(2,72) = 13.492, p < 0.01) – and the control condition – written style design (M = 3.297; F
(2,72) = 13.492, p < 0.01) – in that participants perceived both an avatar and an avatar based
on text design as more hedonic than the written style design. Thus, H2b is supported. For
the usefulness construct, the three designs were not significantly different, and thus H2c is
not supported.
Attitude toward brand and intention to apply. The results in Figure 2 and Table 3 show
significant differences between the two experimental conditions – avatar design (M = 5.595, F
JRIM
17,4

592
Figure 2.
Influence of conditions
(study one)

(2,72) = 21.788, p < 0.01) and avatar based on text design (M = 5.750, F(2,72) = 21.788, p < 0.01) –
and the control condition – written style design (M = 4.007, F(2,72) = 21.788, p < 0.01) – in that
participants’ attitudes toward the brand in both the avatar and avatar based on text designs are
higher than those for the written style design. Thus, H3a is supported. Finally, a significant
difference was only found between an avatar based on text and the control conditions (written
style), in that the intention to deal with a brand was higher in the former (avatar based on text).
Thus, H3b is partially supported.

Discussion
Study one demonstrates partial support of H1. The users of the website did not find any
significant differences between the three conditions concerning clarity (H1a) and convenience
(H1c). However, they were more likely to find significant differences concerning their control
over information (H1b) between interacting with an avatar and using a written style. In the case
of the presence of an avatar, the user does not control the information provided, which is
consistent with previous linguistic studies (e.g. Vidal, 2011) that found that listeners did not
seem to be able to take as much advantage of the repetition, showing the less controllable
aspects of a sound system. Concerning information recall (H2a), the results of study one also
demonstrate that the users found significant differences between an avatar based on text and
having written information. The results also demonstrate that the users did not find any
significant differences between the three conditions concerning usefulness (H2c). These
results are contrary to those in Borau et al. (2021), which anticipated that avatars might be
perceived as more competent in the utilitarian realm than in the hedonic. Our results are also not
in line with those of Butt et al. (2021), who showed the usefulness of AI-powered avatars.
However, users are more likely to find significant differences concerning the degree of hedonism
(H2b) between the two conditions (avatar and avatar based on text) and a written style design.
In cases in which an avatar is present, users perceived the website as more hedonic, which
increased their perceived enjoyment, which is consistent with previous research (e.g. Han, 2021)
that found that users usually perceive an avatar as a hedonic tool. The results of study one also
demonstrate that adding an avatar significantly increases join/apply likelihood (H3a). However,
the current study shows that the presence of avatars significantly enhanced attitude toward the
brand. These results are in line with Borau et al. (2021) and Han (2021).

Study two
Design and participants
Study two aimed to compare the impacts of two conditions for providing information
(avatar and avatar based on text), user’s familiarity with an avatar (familiar vs
unfamiliar) and the language used (English vs user’s native language) on design Strengthening
elements (clarity, control and convenience), information recall, hedonism, usefulness, consumer–
attitude toward the brand and WOM. A 2 × 2 × 2 factorial design was used to pursue
this objective. This study manipulated two main factors (two independent variables:
brand
avatar design and avatar language), but familiarity with avatar technology was not relationships
manipulated. Therefore, the strategy used in study two was one of quasi-
experimental research since we employed some of the rigor and control that exist
in experiments, but it contains a flaw that prevents the research from obtaining an 593
absolute cause-and-effect answer, such as familiarity (Gravetter and Forzano, 2018).
Four student groups participated in this study, and each participant was exposed to
only one condition. As we did in study one, we carried out the power analysis to
determine the sample size. One hundred thirty-five university students participated
in the experiment (between groups) in study two. The four groups contained the
following numbers of participants: 33 (avatar in English design), 35 (avatar based on
text in English design), 34 (avatar in own language design) and 33 (avatar based on
text in own language design).

Materials, procedures and measures

Similar to study one, a website was developed with specific information provided by
different types of avatar (only avatar or avatar based on text) as stimuli in a laboratory
experiment. In addition, the information on the web page was provided in English as the
default language, with an option to change the language to Arabic or Chinese for Arabic
and Chinese students (see Figure 3 for sample stimuli). The experimental procedure was the
same as for study one. We adopted the same measurements in study one and also captured
familiarity with avatar technology (Steenkamp et al., 2003) and WOM (Brüggen et al., 2011).

Data analysis and results

Instrument validation. Table 4 shows the means and Cronbach’s alphas for all the
constructs in study two. The Cronbach’s alphas were above 0.7 for all the constructs. We
checked the unidimensionality of each construct and the measurement model (see Table 4),
as per study one, for goodness of fit (χ2(288) = 382.055; GFI = 0.834; CFI = 0.955; TLI =
0.945; RMSEA = 0.049; SRMR = 0.0552). As shown in Table 4, all the estimates of the factor
loadings are above 0.5, all the t-values are greater than 1.96 (loadings ranged from 3.065 to
17.069), all CR values are above the recommended level of 0.7 and AVE estimates are above
the recommended threshold of 0.5. Therefore, all four criteria provided support for the
convergent validity of the constructs. Same as study one, the discriminant validity of all the
constructs were conducted.
Factorial design ANOVA. Three-way independent ANOVA was conducted as the
study included 2 avatar levels (avatar vs avatar based on text) × 2 language levels
(English vs Own language) × 2 familiarity levels (low vs high). The results revealed
the significant main effect of language and familiarity only on the hedonism
construct. In other words, the analysis did not reveal any significant differences
between the three factors (the avatar, language and familiarity conditions) related to
outcome constructs other than hedonism. The main effect of language indicated that
participants perceived an avatar as more hedonic when it was talking in their
language (M = 5.819) than when it was doing so in English (M = 4.212; F(1,127) =
5.449, p = 0.021), as would be expected. Similarly, a main effect of familiarity
indicates that participants perceived the avatar as more hedonic (enjoyable) when
they were highly familiar with avatar technology (M = 4.902) compared to if they
were unfamiliar with the technology (M = 4.128; F(1,127) = 8.868, p = 0.003). Finally,
JRIM
17,4

594

Figure 3.
One web page from
each scenario (study
two)

there was no interaction between any two of the three factors (avatar, language and
familiarity) or three-way interaction among these factors, as p > 0.05. Thus, H4 is
partially supported.

Discussion
Study two demonstrates partial support for H4. The results do not confirm an interaction
between avatar levels (avatar vs avatar based on text), language levels (English vs Own
language) and familiarity levels (low vs high). The results reveal the significant main effect of
language and familiarity only on the hedonism construct. A main effect of language indicates
that participants perceived the avatar as more hedonic when the avatar was talking in their own
language than when it was talking in English. These results confirm the argument of the
possibility of language being a cue for the hedonic character of a promoted product (Kronrod
et al., 2012). Similarly, a main effect of familiarity indicates that participants perceived an avatar
as more hedonic (enjoyable) when they were highly familiar with avatar technology compared
to if they were unfamiliar with this technology. These results are in line with previous research
showing that higher familiarity levels with technology (e.g. robots) increases the influence of
these technologies on consumers’ perceptions (e.g. Belanche et al., 2019).
 STD Cronbach’s
Strengthening
Constructs Mean loading t-value alpha AVE CR consumer–
Clarity 5.649 0.817 0.550 0.859
brand
The (design) is relationships
Unclear concerning the language/clear concerning 0.703
the language
Not obvious/obvious 0.674 7.125 595
Not apparent/apparent 0.782 8.149
Unclear/clear 0.828 8.541
Unclear concerning the organization of 0.710 7.471
information/clear concerning the organization of
information
Control 4.759 0.803 0.530 0.817
The (design) allows me to control
The provision of information 0.644
The access to required information from the entire 0.646 6.280
body of information
The speed of getting the required information 0.777 7.224
The search process 0.828 7.502
Convenience 5.296 0.799 0.604 0.818
It would not take much time to get the information 0.625
from the (design)
The time required to receive the required 0.863 7.581
information is appropriate
I am able to get to the information from the (design) 0.823 7.409
quickly
Hedonism 4.563 0.801 0.831 0.936
I feel about the (design) is
Dull/exciting 0.925 17.069
Not delightful/delightful 0.906
Not enjoyable/enjoyable 0.903 16.199
Usefulness 5.474 0.799 0.638 0.779
The (design) provides relevant information 0.763
The (design) does a good job of presenting the 0.833 8.763
information
Information recall 5.109 0.803 0.548 0.781
I can remember
Information provided by the (design) through 0.611 6.967
taking notes
The most information provided by the (design) 0.744 8.619
The information provided by the (design) easily 0.847
Attitude toward brand 5.590 0.801 0.745 0.898
After I navigate the website, I feel that the brand is
Technologically developed 0.898
Looking forward to change 0.837 12.795
Modern 0.854 13.246
Word of mouth 4.926 0.804 0.830 0.907
Say positive things about the brand to other people 0.896
Speak positively of the brand to people close to me 0.926 13.311
Familiarity 3.637 0.864 0.589 0.729
By browsing the (design)
It is very unfamiliar to me/it is very familiar to me 0.932 3.065 Table 4.
I am not at all knowledgeable about it/I am very 0.556 Means, loadings,
knowledgeable about it t-values, Cronbach’s
alphas, AVE and CR
Note: STD, standardized (study two)
JRIM Conclusions
17,4 General discussion
Avatars are used by a growing number of brands that have become an accepted tool to
many consumers. Avatars as a marketing communications tool represent the “brand voice”
on the website and offer a crucial way to influence consumers’ buying perceptions and
behaviors. In this research, we examined one specific type of avatar (avatars that can
provide information verbally) and the outcomes of using this. Table 5 summarizes the
596 results of the two studies we conducted for our research.

Theoretical contributions
Our research contributes to the literature on interactive marketing. First, our work contributes to
the increasing body of literature that examines how consumers interact with brands through
new media and tools on websites (e.g. Lim and Childs, 2020; Wang, 2021), showing the significant
outcomes of adding an avatar as a communication tool. To date, much of the work on avatars has
focused on the virtual environment (e.g. Kohler et al., 2009), and few studies have examined
avatars as consumer assistants on a brand website (e.g. Choi et al., 2020; Holzwarth et al., 2006).
Specifically, our research confirms the relationships proposed in the emerging theory of avatar
marketing (Miao et al., 2022). Our study also extends previous research that focused on
examining chat avatars (e.g. Lin et al., 2021). We instead used avatars that were highly
anthropomorphized (using real human images and voices) and interacted with users verbally.
Second, highlighting the important roles of avatars on brands’ websites, this study advances
previous research by adding to the knowledge of the quality of consumer–brand relationships in
online service contexts (Foster et al., 2022; Lin et al., 2021) since the brand avatar serves as an
interactive relationship partner for consumers (Foster et al., 2022). We thus extend the literature
on consumer–avatar interactions and the part they play in enhancing information recall,
enjoyment, attitude toward a brand and intention to deal with the brand. Our results support
the presence of causal relationships between an avatar and perceived control, hedonism and
attitude toward a brand after conducting two experimental studies, showing the high internal
validity of the results. Another important contribution of our work is the findings on the impacts
of familiarity with using avatars and the availability of language options in the model.

Practical implications
Practitioners need to consider placing more emphasis on the social interaction of a brand
avatar as a key factor contributing to positive attitudes toward the brand, as this could

Study two
Interaction effect
Construct Study one (3 factors or 2 factors) Main effect

Clarity Not supported Not supported Not supported

Control Partially supported Not supported Not supported
Convenience Not supported Not supported Not supported
Hedonism Supported Not supported Supported (language
and familiarity)
Usefulness Not supported Not supported Not supported
Information recall Partially supported Not supported Not supported
Table 5. Attitude toward brand Supported Not supported Not supported
Summary of research Apply intention Partially supported
results Word of mouth Not supported Not supported
eventually be used as a substitute for traditional interactions between consumers and Strengthening
employees (Foster et al., 2022). Our findings provide key insights for practitioners into the consumer–
benefits of adding avatars as a marketing communication tool. The results identify the main
dimensions of avatars that can be compared to other tools, such as text and videos. For
brand
instance, the main aspects of an avatar should be highlighted in relation to its ease of use, relationships
such as clarity of voice, use of easy language when delivering information and the addition of
some features (e.g. play, pause, rewind and forward buttons) to control the information flow.
Such aspects could also guide managers in identifying the main deficiencies of avatars and 597
finding solutions. For example, one of the main problems with avatars is that some users
cannot extract the required information quickly. In this case, practitioners should work on
determining the operations that help in obtaining the required information more rapidly.
Furthermore, the results offer practitioners more insights into the roles that avatars can play
on a website, such as guidance, instruction and responding to inquiries. Creating an avatar on
a brand website can lead to valuable outcomes that practitioners and employees should
consider when developing their marketing plans. The outcomes identified in this research
show that the addition of an avatar creates value for the brand website and consumers in
terms of hedonism. In other words, avatars enhance the customer experience on the brand
website (Rogers et al., 2022).

Limitations and future research

Our research has limitations that could be regarded as future research opportunities.
Avatars are gradually being used in more varied industries, such as hotels, banking,
smartphones and restaurants (e.g. Choi et al., 2020), and future research might also
examine the impact of product or service type on consumers’ responses to avatars as
a communication tool. Although adopting experimental research in the current study
increases the internal validity of the results, using a relatively small sample size harms
the generalizability. Thus, future research might also replicate examining these
relationships with a larger sample. We assumed that users usually prefer receiving
information in their mother tongue as their first language would be easier to understand.
However, forcing users to receive information in a specific language might have limited the
generalizability of the results because of the language backfire effect (Holmqvist et al.,
2019). Further research could be conducted by making a comparison between forcing
participants to use a specific language and providing more than one language from
which the participants can choose.

ORCID iD
Mohamed H. Elsharnouby http://orcid.org/0000-0001-9447-5717

References
Aljukhadar, M. and Senecal, S. (2011), “Usage and success factors of commercial recommendation
agents: a consumer qualitative study of MyProductAdvisor.com”, Journal of Research in
Interactive Marketing, Vol. 5 Nos 2/3, pp. 130-152.
Belanche, D., Casaló, L. and Flavián, C. (2019), “Artificial intelligence in FinTech: understanding
robo-advisors adoption among customers”, Industrial Management & Data Systems, Vol. 119
No. 7, pp. 1411-1430.
Borau, S., Otterbring, T., Laporte, S. and Fosso Wamba, S. (2021), “The most human bot: female
gendering increases humanness perceptions of bots and acceptance of AI”, Psychology and
Marketing, Vol. 38 No. 7, pp. 1052-1068.
JRIM Brüggen, E., Foubert, B. and Gremler, D. (2011), “Extreme makeover: short- and long-term effects of
a remodeled servicescape”, Journal of Marketing, Vol. 75 No. 5, pp. 71-87.
17,4
Butt, A., Ahmad, H., Goraya, M., Akram, M. and Shafique, M. (2021), “Let’s play: me and my AI-
powered avatar as one team”, Psychology and Marketing, Vol. 38 No. 6, pp. 1014-1025.
Chen, C. (2019), “Factors affecting the decision to use autonomous shuttle services: evidence from a
scooter-dominant urban context”, Transportation Research Part F: Traffic Psychology and
Behaviour, Vol. 67 No. 2, pp. 195-204.
598
Choi, Y., Mehraliyev, F. and Kim, S. (2020), “Role of virtual avatars in digitalized hotel service”,
International Journal of Contemporary Hospitality Management, Vol. 32 No. 3, pp. 977-997.
Cohen, J. (1992), “A power primer”, Psychological Bulletin, Vol. 112 No. 1, p. 155.
Crolic, C., Thomaz, F., Hadi, R. and Stephen, A. (2022), “Blame the bot: anthropomorphism and anger in
customer-chatbot interactions”, Journal of Marketing, Vol. 86 No. 1, pp. 132-148.
Dattalo, P. (2008), Determining Sample Size: Balancing Power, Precision, and Practicality, Oxford
University Press, New York, NY.
Elsharnouby, M. (2015), “Conceptualization and development of the avatar taxonomy: antecedents,
attitudinal and behavioral consequences”, Doctoral dissertation, University of Hull, Kingston
upon Hull, UK.
Etemad-Sajadi, R. and Ghachem, L. (2015), “The impact of hedonic and utilitarian value of online
avatars on e-service quality”, Computers in Human Behavior, Vol. 52, pp. 81-86.
Farzin, M. and Fattahi, M. (2018), “eWOM through social networking sites and impact on purchase
intention and brand image in Iran”, Journal of Advances in Management Research, Vol. 15 No. 2,
pp. 161-183.
Field, A. and Hole, G. (2010), How to Design and Report Experiments, Sage, London, UK.
Fornell, C. and Larcker, D.F. (1981), “Structural equation models with unobservable variables and
measurement error: algebra and statistics”, Journal of Marketing Research, Vol. 18 No. 3,
pp. 382-388.
Foster, J., McLelland, M. and Wallace, L. (2022), “Brand avatars: impact of social interaction on
consumer-brand relationships”, Journal of Research in Interactive Marketing, Vol. 16 No. 2,
pp. 237-258.
Freling, T., Crosno, J. and Henard, D. (2011), “Brand personality appeal: conceptualization and
empirical validation”, Journal of the Academy of Marketing Science, Vol. 39 No. 3, pp. 392-406.
Gottschalk, P. (2020), The Convenience of White-Collar Crime in Business, Springer Nature, New York,
USA.
Gravetter, F. and Forzano, L. (2018), Research Methods for the Behavioral Sciences, Cengage Learning,
Belmont, USA.
Han, M. (2019), “Instant messaging chat bot: your new best friend?”, in Barbosa, S.F., Santos, C.A. and
Simoes, D. (Eds), Smart Marketing with the Internet of Things, IGI Global, pp. 164-184.
Han, M. (2021), “The impact of anthropomorphism on consumers’ purchase decision in chatbot
commerce”, Journal of Internet Commerce, Vol. 20 No. 1, pp. 46-65.
Hauge, A.P. and Jorgensen, C.B. (2013), “Voice and avatar face recognition with focus on familiarity and
recall accuracy for use in a contact book designed for illiterates”, Master thesis, Aalborg
University, Denmark.
Heiser, R., Sierra, J. and Torres, I. (2008), “Creativity via cartoon spokespeople in print ads: capitalizing
on the distinctiveness effect”, Journal of Advertising, Vol. 37 No. 4, pp. 75-84.
Holmqvist, J., Van Vaerenbergh, Y., Lunardo, R. and Dahlén, M. (2019), “The language backfire effect:
how frontline employees decrease customer satisfaction through language use”, Journal of
Retailing, Vol. 95 No. 2, pp. 115-129.
Holzwarth, M., Janiszewski, C. and Neumann, M. (2006), “The influence of avatars on online consumer Strengthening
shopping behavior”, Journal of Marketing, Vol. 70 No. 4, pp. 19-36.
consumer–
Hoyet, L., Spies, C., Plantard, P., Sorel, A., Kulpa, R. and Multon, F. (2019), “Influence of motion speed on brand
the perception of latency in avatar control”, 2019 IEEE International Conference on Artificial
Intelligence and Virtual Reality (AIVR), 9 December 2019, IEEE, San Diego, CA, USA, pp. 286-2863. relationships
Janssens, W., Wijnen, K., De Pelsmacker, P. and Van Kenhove, P. (2008), Marketing Research with
SPSS, Pearson, Harlow, UK.
599
Jin, S. and Bolebruch, J. (2009), “Avatar-based advertising in second life: the role of presence and
attractiveness of virtual spokespersons”, Journal of Interactive Advertising, Vol. 10 No. 1,
pp. 51-60.
Johnson, S. (1982), “Listening and reading: the recall of 7- to 9-year-olds”, British Journal of Educational
Psychology, Vol. 52 No. 1, pp. 24-32.
Keeling, K., McGoldrick, P. and Beatty, S. (2010), “Avatars as salespeople: communication style, trust,
and intentions”, Journal of Business Research, Vol. 63, pp. 793-800.
Keh, H. and Pang, J. (2010), “Customer reactions to service separation”, Journal of Marketing, Vol. 74
No. 2, pp. 55-70.
Kiesel, J., de Vries, A., Hagen, M., Stein, B. and Potthast, M. (2018), “WASP: web archiving and search
personalized”, available at: http://ceur-ws.org/Vol-2167/paper6.pdf (accessed 23 October 2022).
Kohler, T., Fueller, J., Matzler, K. and Stieger, D. (2011), “Co-creation in virtual worlds: the design of the
user experience”, MIS Quarterly, Vol. 35 No. 3, pp. 773-788.
Kohler, T., Matzler, K. and Füller, J. (2009), “Avatar-based innovation: using virtual worlds for
real-world innovation”, Technovation, Vol. 29 Nos 6-7, pp. 395-407.
Kotlarek, J., Kwon, O., Ma, K., Eades, P., Kerren, A., Klein, K. and Schreiber, F. (2020), “A study of
mental maps in immersive network visualization”, 2020 IEEE Pacific Visualization Symposium
(PacificVis), IEEE, Tianjin, China, pp. 1-10.
Kroesen, M., Handy, S. and Chorus, C. (2017), “Do attitudes cause behavior or vice versa? An
alternative conceptualization of the attitude-behavior relationship in travel behavior
modeling”, Transportation Research Part A: Policy and Practice, Vol. 101, pp. 190-202.
Kronrod, A., Grinstein, A. and Wathieu, L. (2012), “Enjoy! Hedonic consumption and compliance with
assertive messages”, Journal of Consumer Research, Vol. 39 No. 1, pp. 51-61.
Langdon, I., Hardin, R. and Learmonth, I. (2002), “Informed consent for total hip arthroplasty: does
a written information sheet improve recall by patients?”, Annals of the Royal College of Surgeons
of England, Vol. 84 No. 6, p. 404.
Lee, J. and Lee, K. (2006), “Precomputing avatar behavior from human motion data”, Graphical Models,
Vol. 68 No. 2, pp. 158-174.
Liew, T., Tan, S. and Ismail, H. (2017), “Exploring the effects of a non-interactive talking avatar on
social presence, credibility, trust, and patronage intention in an e-commerce website”, Human-
centric Computing and Information Sciences, Vol. 7 No. 1, pp. 1-21.
Lim, H. and Childs, M. (2020), “Visual storytelling on Instagram: branded photo narrative and the role
of telepresence”, Journal of Research in Interactive Marketing, Vol. 14 No. 1, pp. 33-50.
Lin, Y., Doong, H. and Eisingerich, A. (2021), “Avatar design of virtual salespeople: mitigation of
recommendation conflicts”, Journal of Service Research, Vol. 24 No. 1, pp. 141-159.
Luna, D., Peracchio, L.A. and De Juan, M.D. (2003), “The impact of language and congruity on
persuasion in multicultural e-marketing”, Journal of Consumer Psychology, Vol. 13 Nos 1-2,
pp. 41-51.
Lund, R. (1991), “A comparison of second language listening and reading comprehension”, The
Modern Language Journal, Vol. 75 No. 2, pp. 196-204.
JRIM McBreen, H. and Jack, M. (2001), “Evaluating humanoid synthetic agents in e-retail applications”,
IEEE Transactions on Systems, Man, and Cybernetics – Part A: Systems and Humans, Vol. 31
17,4 No. 5, pp. 394-405.
Mehrabian, A. and Russell, J. (1974), An Approach to Environmental Psychology, The MIT Press,
Cambridge, MA, USA.
Miao, F., Kozlenkova, I., Wang, H., Xie, T. and Palmatier, R. (2022), “An emerging theory of avatar
marketing”, Journal of Marketing, Vol. 86 No. 1, pp. 67-90.
600
Oshita, M. (2006), “Motion-capture-based avatar control framework in third-person view virtual
environments”, Proceedings of the 2006 ACM SIGCHI International Conference on Advances
in Computer Entertainment Technology, Hollywood, USA, pp. 2-es
Park, J., Hyun, H. and Thavisay, T. (2021), “A study of antecedents and outcomes of social media
WOM towards luxury brand purchase intention”, Journal of Retailing and Consumer Services,
Vol. 58 No. 1, p. 102272.
Reeves, T., Ford, E., Duncan, W. and Ginter, P. (2005), “Communication clarity in strategic
management data sources”, Strategic Organization, Vol. 3 No. 3, pp. 243-278.
Rodríguez-Torrico, P., Cabezudo, R., San-Martín, S. and Apadula, L. (2021), “Let it flow: the role of
seamlessness and the optimal experience on consumer word of mouth in omnichannel
marketing”, Journal of Research in Interactive Marketing, Forthcoming.
Rogers, S., Broadbent, R., Brown, J., Fraser, A. and Speelman, C. (2022), “Realistic motion avatars
are the future for social interaction in virtual reality”, Frontiers in Virtual Reality, Vol. 2,
pp. 1-12.
Roy, G., Datta, B., Mukherjee, S., Basu, R. and Shrivastava, A. (2021), “Effect of eWOM valence on
purchase intention: the moderating role of product”, International Journal of Technology
Marketing, Vol. 15 Nos 2-3, pp. 158-180.
Rucker, D., Dubois, D. and Galinsky, A. (2011), “Generous paupers and stingy princes: power drives
consumer spending on self versus others”, Journal of Consumer Research, Vol. 37 No. 6,
pp. 1015-1029.
Seiders, K., Voss, G., Grewal, D. and Godfrey, A. (2005), “Do satisfied customers buy more? Examining
moderating influences in a retailing context”, Journal of Marketing, Vol. 69 No. 4, pp. 26-43.
Sheinin, D., Varki, S. and Ashley, C. (2011), “The differential effect of ad novelty and message
usefulness on brand judgments”, Journal of Advertising, Vol. 40 No. 3, pp. 5-18.
Sheth, J. (2021), “Post-pandemic marketing: when the peripheral becomes the core”, Journal of Research
in Interactive Marketing, Vol. 16 No. 1, pp. 37-44.
Sohn, S., Seegebarth, B., Kissling, M. and Sippel, T. (2020), “Social cues and the online purchase
intentions of organic wine”, Foods, Vol. 9 No. 5, p. 643.
Stæhr, L. (2008), “Vocabulary size and the skills of listening, reading and writing”, Language Learning
Journal, Vol. 36 No. 2, pp. 139-152.
Steenkamp, J., Batra, R. and Alden, D. (2003), “How perceived brand globalness creates brand value”,
Journal of International Business Studies, Vol. 34 No. 1, pp. 53-65.
Thompson, D., Hamilton, R. and Petrova, P. (2009), “When mental simulation hinders behavior: the
effects of process-oriented thinking on decision difficulty and performance”, Journal of
Consumer Research, Vol. 36 No. 4, pp. 562-574.
Van Noort, G. and Willemsen, L. (2012), “Online damage control: the effects of proactive versus reactive
webcare interventions in consumer-generated and brand-generated platforms”, Journal of
Interactive Marketing, Vol. 26 No. 3, pp. 131-140.
Vidal, K. (2011), “A comparison of the effects of reading and listening on incidental vocabulary
acquisition”, Language Learning, Vol. 61 No. 1, pp. 219-258.
Voss, K., Spangenberg, E. and Grohmann, B. (2003), “Measuring the hedonic and utilitarian Strengthening
dimensions of consumer attitude”, Journal of Marketing Research, Vol. 40 No. 3,
pp. 310-320.
consumer–
Wakefield, K. and Baker, J. (1998), “Excitement at the mall: determinants and effects on shopping
brand
response”, Journal of Retailing, Vol. 74 No. 4, pp. 515-539. relationships
Wang, C. (2021), “New frontiers and future directions in interactive marketing: inaugural Editorial”,
Journal of Research in Interactive Marketing, Vol. 15 No. 1, pp. 1-9.
601
Williams, E. (1977), “Experimental comparisons of face-to-face and mediated communication: a
review”, Psychological Bulletin, Vol. 84 No. 5, p. 963.

About the authors

Mohamed H. Elsharnouby is an Associate Professor of Marketing in the Department of Business
Administration at Badr University in Cairo and is currently on leave from the Cairo University, where
he is a tenured Faculty Member. He received his Ph.D. from the Hull University Business School, UK.
Current research interests include new technology in marketing, advertising and relationship
marketing. His work appeared in a number of journals and proceedings, including Journal of
Business Research, Journal of Research in Interactive Marketing, Academy of Marketing Science,
American Marketing Association Conference, European Marketing Academy Conference and
Academy of Marketing Conference.
Chanaka Jayawardhena is the Head of the Department of Marketing and Retailing and Professor of
Marketing at the University of Surrey. He specialises in word of mouth marketing, relationship
marketing and consumption behaviors. Chanaka has particular interest in the virtual marketplaces,
with particular attention on how consumers evaluate the services they receive/consume over these
remote channels. He examines how organizations use technology (Internet and mobile channels) in
attracting, developing and maintaining customer relationships. Chanaka has had over 100 articles
published in specialist journals and proceedings, including Industrial Marketing Management, Journal
of Business Research, Tourism Management and European Journal of Marketing, among others.
Alaa M. Elbedweihy is an Assistant Professor of Marketing in the Business Administration
Department at the Faculty of Commerce, Cairo University. She received her Ph.D. (Marketing) from
the Hull University Business School, UK. Alaa’s research interests include brand relationships,
relationship marketing, online consumer behavior and digital marketing and has published papers
in Journal of Business Research and Marketing Review Journal. She has also presented various
research-based papers at several international conferences including Academy of Marketing Science,
American Marketing Association Conference, European Marketing Academy Conference and Academy
of Marketing Conference.
Hongfei Liu is an Associate Professor in Marketing at the Southampton Business School,
University of Southampton, UK. Hongfei’s research interests primarily lie in digital marketing and
social media. His research focuses on the impact of digitalization and technologization on individuals,
businesses and societies. His work has been published in highly regarded journals, including Journal
of Business Research, Journal of Business Ethics, European Journal of Marketing, Industrial Marketing
Management, Tourism Management and Journal of Travel Research, among others. Hongfei sits in the
editorial board of international journals and chairs tracks in international conferences. Hongfei Liu is
the corresponding author and can be contacted at: hongfei.liu@soton.ac.uk

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com