UvA-DARE (Digital Academic Repository)

Try online before you buy: How does shopping with augmented reality affect
brand responses and personal data disclosure

Smink, A.R.; Frowijn, S.; van Reijmersdal, E.A.; van Noort, G.; Neijens, P.C.
DOI
10.1016/j.elerap.2019.100854
Publication date
2019
Document Version
Final published version
Published in
Electronic Commerce Research and Applications
License
Article 25fa Dutch Copyright Act (https://www.openaccess.nl/en/in-the-netherlands/you-share-
we-take-care)
Link to publication

Citation for published version (APA):

Smink, A. R., Frowijn, S., van Reijmersdal, E. A., van Noort, G., & Neijens, P. C. (2019). Try
online before you buy: How does shopping with augmented reality affect brand responses
and personal data disclosure. Electronic Commerce Research and Applications, 35, [100854].
https://doi.org/10.1016/j.elerap.2019.100854

General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)
and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open
content license (like Creative Commons).

Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please
let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material
inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter
to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You
will be contacted as soon as possible.
UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

Download date:١٠ ‫ ربمتبس‬٢٠٢٣

 Electronic Commerce Research and Applications 35 (2019) 100854

Contents lists available at ScienceDirect

Electronic Commerce Research and Applications

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

Try online before you buy: How does shopping with augmented reality affect T
brand responses and personal data disclosure
⁎
Anne R. Smink , Sanne Frowijn, Eva A. van Reijmersdal, Guda van Noort, Peter C. Neijens
Amsterdam School of Communication Research (ASCoR), University of Amsterdam, Amsterdam, the Netherlands

A R T I C LE I N FO A B S T R A C T

Keywords: Augmented Reality (AR) enables consumers to virtually try products on their own face or surroundings in real
Augmented reality time (e.g., make-up, furniture), which could help providing consumers a ‘try before you buy’ experience when
Persuasion shopping online. In an online experiment, we examined the potential positive and negative effects of online
Online shopping product presentation with AR, compared to two non-AR product presentations on a picture of the self or a model.
Informativeness
Results suggest that AR enhances perceived informativeness and enjoyment of the shopping experience, as
Enjoyment
Intrusiveness
opposed to both non-AR product presentations. Consequently, perceived informativeness leads to a cognitive
Personal data disclosure process which enhances purchase intention and willingness to share personal data with the brand, while per-
ceived enjoyment leads to an affective process which enhances attitude towards the brand. At the same time, AR
is perceived as more intrusive, but against expectations, this does not lead to any negative effects.

1. Introduction local presence (Verhagen et al., 2014), stronger self-brand connections

(Baek et al., 2018), and increased utilitarian and hedonic value of the
While online purchases via (mobile) e-commerce continue to grow shopping experience (Hilken et al., 2017). With this study, we aim to
worldwide, retailers face issues such as high return rates, online shopping extend the existing literature in three ways.
card abandonment, and webrooming (browse products online, then shop First, most studies have compared AR experiences involving aug-
product offline; Dacko, 2016; Hilken et al., 2018). These issues may menting a virtual product on the face, to non-AR experiences which
partly be attributed to the lack of direct product experiences when show the product on a model, or solely show the product (Baek et al.,
shopping online, as online product presentations lack the sensory in- 2018; Javornik, 2016; Verhagen et al., 2014; Yim and Park, 2018).
formation one would have when shopping in a physical store (Fang Therefore, it remains unclear which part of these effects can be at-
et al., 2014; Hilken et al., 2017; Overmars and Poels, 2015; Verhagen tributed to AR technology, and which part to merely seeing one’s own
et al., 2014). One way to bridge the gap between offline and online face. In this study, we aim to filter out this possible confound by
shopping is through Augmented Reality (AR) technology (Baek et al., comparing an online product presentation with AR, to a non-AR pro-
2018). AR enables consumers to virtually try products on their own face duct presentation featuring a picture of one’s own face, and a non-AR
or surroundings in real time (e.g., make-up, furniture, sunglasses), and product presentations featuring a picture of a model. We expect that AR
as such deliver a more direct product experience (Hilken et al., 2017; has a strengthening effect on persuasive outcomes (as opposed to both
Verhagen et al., 2014). Because this technology offers a ‘try before you non-AR conditions), due to its interactive and real time features. And,
buy’ experience, AR shows great potential for marketers and retailers to we also expect that showing one’s own face in the non-AR condition (as
improve online conversion rates and reduce return rates (Dacko, 2016; opposed to a model) causes stronger effects because of consumers’
Morgan, 2017). preference and attention for self-relevant cues (self-referencing theory;
Due to its increasing popularity, academics have started comparing Kuiper and Rogers, 1979).
AR experiences to non-AR experiences to assess the value of AR during Second, previous studies have mostly focused on effects of AR on
online shopping. These studies mostly showed that an AR experience brand related affective and behavioral responses (Baek et al., 2018;
was more persuasive compared to a non-AR experience (Baek et al., Javornik, 2016; Verhagen et al., 2014). However, the effectiveness and
2018; Javornik, 2016; Verhagen et al., 2014). This enhancing effect of adoption of AR apps might partly be dependent on consumers’ will-
AR was found via several underlying mechanisms, such as increased ingness to share personal data. In order to use AR apps, retailers ask

⁎
Corresponding author at: University of Amsterdam, P.O. Box 15791, 1001 NG Amsterdam, The Netherlands.
E-mail address: [email protected] (A.R. Smink).

https://doi.org/10.1016/j.elerap.2019.100854
Received 15 January 2019; Received in revised form 23 April 2019; Accepted 26 April 2019
Available online 29 April 2019
1567-4223/ © 2019 Elsevier B.V. All rights reserved.
A.R. Smink, et al. Electronic Commerce Research and Applications 35 (2019) 100854

Fig. 1. Conceptual model.

users for permission to use their camera and collect other personal data. time view of themselves or their surroundings through a webcam or
Research showed this is considered a drawback of using AR apps, and smartphone camera, overlaid with virtual products. Third, the virtual
consumers might be reluctant to share personal data with AR apps object is registered in 3D and therefore has a fixed position in the 3D
(Dacko, 2016). However, it has not been studied whether usage of AR space (Azuma, 1997). In this sense, a virtual product can be fixed on the
apps actually affects consumers’ personal data disclosure. Therefore, user’s head through facial recognition (e.g., makeup or sunglasses), or
this study considers willingness to share personal data as a dependent in a fixed position in the user’s surroundings (e.g., furniture), meaning
variable, besides brand attitude and purchase intention as brand re- that users can inspect the product from different angles while moving
sponses. around. Because the virtual objects are registered and fixed in physical
Third, while previous research has mostly shown positive effects of space, it enables interaction between the virtual objects and the real
AR through several underlying mechanisms (Baek et al., 2018; world (Azuma, 1997). These unique characteristics of AR result in on-
Javornik, 2016; Verhagen et al., 2014; Yim and Park, 2018), its po- line product presentations that resemble a more direct product ex-
tential negative effects are often overlooked. Although recent research perience (Verhagen et al., 2014). We expect that due to this, online
has shown that AR is less effective for consumers with specific char- product presentations with AR (compared to the two non-AR condi-
acteristics (i.e., strong privacy concerns; Hilken et al., 2017), it is not tions) will enhance or diminish consumer reactions via three underlying
clear through which process AR could yield negative persuasive effects, mechanisms: perceived informativeness, perceived enjoyment, and
and whether these negative effects are stronger or weaker than the perceived intrusiveness. The conceptual model corresponding to this is
positive persuasive effects (Suh and Prophet, 2018). Therefore, we ex- outlined in Fig. 1.
amine whether differences in online product presentations will either
translate into more positive or negative persuasive effects. 2.2. Perceived informativeness
Following Hilken et al. (2017) we focused on both hedonic (per-
ceived enjoyment), and utilitarian processes (perceived informative- With perceived informativeness, we address the subjective evalua-
ness), as these are important indicators of intended adoption of new tion of how much information the online product presentation offers.
technologies and positive consumer reactions. Moreover, by differ- While online stores can provide factual information about the search
entiating between hedonic and utilitarian processes, we can separate attributes of a product (attributes that can be assessed before using a
affective versus cognitive effects of AR. At the same time, the real time product, such as measurements or price), they are less capable of pro-
and interactive nature of this relatively new and unfamiliar technology viding information about the experience attributes of a product (attri-
could potentially raise the perceived intrusiveness, and therefore ne- butes that can be verified by actual use of the product, such as fit,
gatively affect consumer reactions. We expect, based on equity theory shape, texture, and feel; Overmars and Poels, 2015). Research has
(Adams, 1963) and privacy calculus theory (Culnan and Armstrong, shown that direct product experiences (as opposed to indirect experi-
1999), that consumers make a cost-benefit analysis, which determines ences) provide more vivid mental images of the product, more sensory
whether the positive persuasive effects outweigh the potential negative engagement, and enable consumers to better judge the quality of the
effects and to decide whether they want to disclose personal informa- experience attributes of a product (Hamilton and Thompson, 2007;
tion (Dinev and Hart, 2006; Wottrich et al., 2018). Klein, 2003; Li et al., 2001; Overmars and Poels, 2015; Schlosser,
In sum, this study examines the positive and negative effects of 2003). Therefore, these direct product experiences provide more sub-
online product presentation through AR (compared to a picture of the jective information than indirect product experiences. To illustrate,
self or a model), on brand attitude, purchase intention and willingness previous research showed that more direct product experiences (videos
to share personal data through three underlying processes: perceived or 360 view of the product) led to a higher perceived informedness than
informativeness, enjoyment and intrusiveness. more indirect product experiences (a static picture of the product; Li
and Meshkova, 2013). Thus, from the above, it can be argued that
2. Theoretical framework online product presentations that offer the most direct product ex-
perience, will likely be perceived as the most informative.
2.1. Online product presentation with AR Due to its previously described interactive and real-time features,
AR resembles the most direct product experience (Verhagen et al.,
AR has three unique characteristics that distinguish an online pro- 2014), and is therefore also expected to be perceived as most in-
duct presentation with AR from other forms of online product pre- formative. A consumer survey also showed that the ‘more complete
sentation. First, AR combines the real and virtual world by overlaying information on products’ is considered the biggest benefit of using AR
virtual products onto the consumer or their surroundings, which en- shopping applications (Dacko, 2016). AR provides a multisensory and
ables them to try a product ‘as if’ it is really there (Scholz and Smith, vivid experience, which enables consumers to visually examine specific
2015). Second, the virtual objects in AR are interactive and are shown in experience characteristics of the product (Klein, 2003; Li et al., 2001;
real time (Azuma, 1997), which means that users are exposed to a real Yim et al., 2017). In comparison, a non-AR experience, providing a

2
A.R. Smink, et al. Electronic Commerce Research and Applications 35 (2019) 100854

picture of the consumers’ own face or a model, lacks these interactive online product presentations, makes users more willing to share their
and real-time features, and as such provides a more indirect experience. personal data.
As we argue that more direct product experiences will be perceived as
Hypothesis 2 ((The Perceived Informativeness Mediation Hypothesis):).
more informative, we expect that online product presentation with AR
Perceived informativeness positively mediates the effect of online product
will be perceived as more informative than both non-AR conditions.
presentation on a) brand attitude, b) purchase intention, and c) willingness
Additionally, both non-AR conditions are compared in order to filter
to share personal data.
out whether the effect is (partly) caused by the presence of one’s own
face rather than AR. We also expected an enhancing effect of seeing
one’s own face (as opposed to a model), through a process called self- 2.3. Perceived enjoyment
referencing (Kuiper and Rogers, 1979). Self-referencing means that,
when consumers are exposed to self-relevant information (in this case, a Besides providing utilitarian value through enhancing the perceived
picture of their own face), they tend to pay more attention and process informativeness, online product presentations can also provide hedonic
the information more easily and deeply (Yan Tam and Ying Ho, 2005). value through increasing perceived enjoyment (Hilken et al., 2017;
We anticipate that this cognitive process will increase the perceived Huang and Liao, 2015; Li and Meshkova, 2013). Research showed that
informativeness of the online product presentation. Moreover, showing enjoyment during online shopping can be induced through increased
the virtual product on the consumers’ face (as opposed to a model) interactivity and vividness of the experience (Li and Meshkova, 2013;
enables them to better judge the fit of the product, and as such provides Yim et al., 2017; Yim and Park, 2018). Interactive and vivid experiences
more subjective information (Baek et al., 2018). (on websites, shopping sites, or through mobile apps) have been shown
In sum, we hypothesize that both online product presentation with to lead to flow, involvement, and immersion, creating a more positive
AR (compared to the non-AR conditions), and online product pre- emotional experience, and as such induce a more fun and enjoyable
sentation on one’s own face (compared to a model) enhance the per- experience (Kim and Forsythe, 2008; Shih, 1998; van Noort et al., 2012;
ceived informativeness. By posing two hypotheses for each process, we Yim et al., 2017). For example, research showed that online product
distinguish between the relative effect of AR (H1a) and the effect of presentations with more interactive and vivid features (i.e., a video of
seeing one’s own face (H1b). the product) led to more enjoyment (operationalized as excitement or
hedonic value) as opposed to static online product presentations
Hypothesis 1a. (The Perceived Informativeness of AR Hypothesis):
without these features (Hilken et al., 2017; Li and Meshkova, 2013; Yim
Online product presentation with AR elicits a higher perceived
and Park, 2018). As discussed, an online product presentation with AR
informativeness than online product presentation without AR (on a picture
provides sensory cues to induce a vivid experience and enables real
of the consumer’s own face or a model).
time interaction with virtual products (Azuma, 1997; Javornik, 2016).
Hypothesis 1b. (The Perceived Informativeness of the Consumer’s Based on this, we expect that online product presentation with AR will
Own Face Hypothesis): Online product presentation on a picture of the be perceived as more enjoyable than online product presentation
consumer’s own face elicits a higher perceived informativeness than online without AR.
product presentation on a picture of a model. Second, research on self-referencing shows that consumers prefer
and attach positive affect to information or advertising that addresses
The product information obtained through online product pre-
the self (Ahn et al., 2017; Holland et al., 2009; Maslowska et al., 2016;
sentations can consequently lead to positive brand responses. Providing
Yan Tam and Ying Ho, 2005). Therefore, we expect that besides AR,
enriching product information can help consumers to reduce un-
there will also be an enhancing effect on perceived enjoyment through
certainty about product choices, make more informed decisions when
showing the product on the self (as opposed to a model). Thus, we
shopping online, and, as such, can lead to a higher purchase intention
argue that both the presence of one’s own face (compared to a model),
(Chen and Yen, 2004; Hilken et al., 2017; Overmars and Poels, 2015).
and AR (versus non-AR), will have an ameliorating effect on perceived
For example, research showed that greater informedness through in-
enjoyment.
teractive product presentations positively enhanced purchase intention
(Li and Meshkova, 2013). Additionally, because product information Hypothesis 3a. (The Perceived Enjoyment of AR Hypothesis): Online
obtained through AR is considered more reliable and reduces the risk product presentation with AR elicits a higher perceived enjoyment than
associated with buying the product, we expect that consumers will online product presentation without AR (on a picture of the consumer’s own
experience more positive affect, which will be transferred to their at- face or a model).
titude towards the brand. Similarly, we expect that perceived in-
Hypothesis 3b. (The Perceived Enjoyment of the Consumer’s Own
formativeness will positively mediate the effect of online product pre-
Face Hypothesis): Online product presentation on a picture of the
sentation on brand attitude and purchase intention.
consumer’s own face elicits a higher perceived enjoyment than online
Lastly, we expect that the enhanced informativeness will have a
product presentation on a picture of a model.
positive effect on the willingness to disclose personal data with the
brand. Research on AR showed that when users feel they have control Enjoyment while shopping online can lead to positive affect
over the access to personal information in AR applications, this has a (Mathwick et al., 2001), which, in turn, can be carried over to the brand
positive effect on user satisfaction (Poushneh, 2018). In has not been through affect transfer (Baek et al., 2018). Research indeed showed that
studied however, whether users are actually more willing to disclose increasing enjoyment while shopping online consequently enhances
their personal information with AR applications (as opposed to non-AR brand attitude and purchase intention (Javornik, 2016; Li and
applications). Research has shown that product information gathered Meshkova, 2013; Pantano et al., 2017; Yim et al., 2017). At the same
through a more direct experience on online shopping sites (induced by time, enjoyment while shopping online is characterized by feeling im-
image interactivity) is considered as more reliable and trustworthy mersed and involved, which makes consumers less critical towards the
(Overmars and Poels, 2015), which can make users less reluctant to content and more likely to be persuaded (Huang and Liao, 2017;
disclose personal information (Schoenbachler and Gordon, 2002). Javornik, 2016; Yim et al., 2017). This can also translate into a higher
Moreover, research showed that websites which were considered as willingness to disclose personal data. Within this study, we expect that
more informative, induced trust in the user, reduced users’ privacy enjoyment will reduce the natural resistance to share personal in-
concerns, and consequently made them more willing to share personal formation. Consequently, we expect that perceived enjoyment will
data with the website (Kaushik et al., 2018; Pavlou et al., 2007). In positively mediate the effect on willingness to share personal data.
relation to AR, we expect that the enhanced informativeness through
Hypothesis 4 ((The Perceived Enjoyment Mediation Hypothesis):).

3
A.R. Smink, et al. Electronic Commerce Research and Applications 35 (2019) 100854

Perceived enjoyment positively mediates the effect of online product decision to use a technology by comparing what they receive (input or
presentation on (a) brand attitude, (b) purchase intention, and (c) benefits) to what they sacrifice (output or costs). When the benefits of
willingness to share personal data. the technology outweigh the costs, user satisfaction will be positive. At
the same time, user satisfaction will be negative when the costs out-
weigh the benefits (Adams, 1963). For example, a study from Poushneh
2.4. Perceived intrusiveness
(2018) showed that the quality of the AR experience (the benefits) and
the perceived user control over personal information (potential cost)
In order to use AR applications, consumers need to provide personal
affected user satisfaction. When both the quality of the AR experience
information to enable the technology, such as camera access, an e-mail
and the control over personal information was low (as opposed to high),
address, or location. Consumers can be reluctant to share this in-
it led to a lower (higher) user satisfaction. Research based on privacy
formation, because they do not know how their information will be
calculus theory also showed that consumers make a cost-benefit analysis
used (Poushneh, 2018). When consumers do not feel in control over the
to decide whether or not they want to disclose personal information
information they are sharing, it can lead to an uncomfortable feeling
(Culnan and Armstrong, 1999; Dinev and Hart, 2006; Wottrich et al.,
and raise feelings of intrusiveness (Baek and Morimoto, 2012;
2018). For example, a study on branded apps showed that when the
Poushneh, 2018; van Doorn and Hoekstra, 2013; White et al., 2008).
relative benefits (app value) where stronger than the costs (intrusive-
Especially because AR is a relatively new technology, consumers are
ness and privacy concerns), consumers were more willing to disclosure
unaware of the benefits of using AR for online shopping and could
personal information (Wottrich et al., 2018). In this study, the potential
perceive AR apps asking for camera access as intrusive (Dacko, 2016;
benefits of using AR for online shopping (perceived informativeness and
Hopp and Gangadharbatla, 2016; Tokunaga, 2013).
enjoyment), are compared to the potential cost (perceived intrusive-
First of all, we expect that perceived intrusiveness will be highest for
ness). This cost-benefit analysis would then either yield more positive
online product presentation with AR, because consumers need to give
or more negative effects on brand responses and willingness to share
access to their live camera to use the application. This will lead to a
data depending on which mediating effect is stronger. To examine this,
higher perceived intrusiveness than using online product presentation
we pose a research question:
without AR (showing a picture of the self or a model). Additionally, we
Research Question 1: Is the (hypothesized) positive mediating ef-
expect that online product presentation using a picture of the con-
fect through perceived informativeness and perceived enjoyment on
sumer’s own face, will also raise feelings of intrusiveness, as consumers
brand responses and willingness to share data stronger than the (hy-
still need to provide more personal information (a picture), which is not
pothesized) negative mediating effect through perceived intrusiveness?
needed for the model condition. Therefore, we hypothesize:
Hypothesis 5a. (The Perceived Intrusiveness of AR Hypothesis):
Online product presentation with AR elicits a higher perceived 3. Method
intrusiveness than online product presentation without AR (on a picture of
the consumer’s own face or a model). 3.1. Design and participants

Hypothesis 5b. (The Perceived Intrusiveness of the Consumer’s Own

An online experiment was conducted, using a one-way factorial
Face Hypothesis): Online product presentation on a picture of the
design with three conditions differing in online product presentation:
consumer’s own face elicits a higher perceived intrusiveness than online
Augmented Reality (AR), a picture of the self (non-AR self) or a picture
product presentation on a picture of a model.
of a model (non-AR model). Participants were recruited using a snow-
When consumers perceive information as too intrusive, it can lead ball sampling procedure and were contacted through e-mail and social
to reactance and avoidance (Baek and Morimoto, 2012; Li et al., media over a period of three weeks in May 2018 in The Netherlands.
2002b). According to reactance theory, consumers try to resist per- Because the online application concerned a make-up app, only
suasion and respond negatively when they experience a lack of control women could participate in the experiment. We selected participants
or freedom (Brehm, 1966). This negative effect can spill over to brand aged between 18 and 30 years old, as this age group is in general more
attitude and purchase intention. For example, research showed that open to innovative technologies, and form the main target group for
intrusiveness induced by personalized or location-based advertising can online shopping (Blake et al., 2003). Moreover, women are more re-
cause negative evaluations and behavioral intentions (Gazley et al., sponsive to direct product experiences than men, and perceive a higher
2015; Huhn et al., 2011; McCoy et al., 2008; van Doorn and Hoekstra, risk of shopping online compared to men (which provide more indirect
2013). Intrusiveness could also negatively affect willingness to share product experiences; Garbarino and Strahilevitz, 2004; Hasan, 2010).
personal data. For example, intrusiveness of mobile apps (oper- Therefore, AR might be especially useful for women to help visualize
ationalized as how much personal data the app was collecting), has products. We expect that this group of young women could be potential
been found to negatively affect personal data disclosure (Wottrich et al., users of new technologies in e-commerce, and therefore, their reactions
2018). Moreover, ‘having to give too much personal information’ is to AR will provide valuable, relevant insights for practice and an ap-
considered one of the biggest drawbacks of using AR apps, according to propriate study sample. Additional requirements were that all partici-
a consumer survey (Dacko, 2016). Therefore, we expect that the per- pants had a laptop/computer with built-in camera or a webcam, be-
ceived intrusiveness will also decrease the willingness to share personal cause this was needed in the AR and non-AR self-condition.
data. We hypothesize: From the full responses (N = 183), 50 participants had to be re-
moved from the data because they failed the manipulation check, which
Hypothesis 6 ((The Perceived Intrusiveness Mediation Hypothesis):).
means they applied the makeup in a different online product pre-
Perceived intrusiveness negatively mediates the effect of online product
sentation than they were instructed.1 Additionally, one participant was
presentation on (a) brand attitude, (b) purchase intention, and (c)
willingness to share personal data.
1
We excluded participants that used a different online product presentation
condition then they were assigned to, as this leads to self-selection and does not
2.5. Positive versus negative effects assure random assignment to conditions. Moreover, participants could have had
several reasons for not using the assigned condition: (1) because participants
Lastly, we will explore whether the hypothesized positive mediating did not read the instructions carefully, (2) because they did not fulfill the re-
effects (H2 and H4) are stronger than the hypothesized negative effects quirements to have a camera which made them unable to use the AR condition
(H6). Equity theory (Adams, 1963) states that consumers make a or non-AR self condition, or (3) because they were not willing to use their

4
A.R. Smink, et al. Electronic Commerce Research and Applications 35 (2019) 100854

Table 1 Table 2
Sample characteristics. Measurement items with factor loadings and descriptives per variable.
Variables Sample Items INF ENJ INT BA PI WSPD
(N = 132)
The information about the lipstick colors I tried was…
Age 23.57 (2.15) complete* 0.91
Gender detailed* 0.90
Female 100% credible* 0.83
Education understandable* 0.90
Secondary vocational 0% helped me find answers on questions 0.91
Pre-university 6.8% about the different colors of
Higher professional 25.8% lipstick*
University 67.4% helped me make decisions on the color 0.89
Cosmetics use of lipstick*
Sometimes 9.1%
Trying the colors of lipstick on the website was…
Daily 90.9%
enjoyable* 0.95
Brand familiarity
fun* 0.97
Unfamiliar 7.6%
pleasant* 0.97
Moderately familiar 38.6%
interesting* 0.93
Familiar 53.8%
Online shopping experience 5.44 (1.63) Trying the colors of lipstick on the website was…
Brand attitude (pre-measure) 5.04 (0.98) interfering* 0.95
invasive* 0.95
intrusive* 0.97
unpleasant* 0.97
excluded because of extreme values on more than three variables and
disturbing* 0.97
was considered an outlier. Finally, 132 participants qualified for testing
the hypotheses. All participants were women and aged between 18 and I think the brand Sephora is…
bad – good 0.92
30 years old (M = 23.57, SD = 2.15). The participants were overall unattractive – attractive 0.94
highly educated, familiar with the Sephora brand, experienced online negative – positive 0.92
shoppers and positive about Sephora. All sample characteristics are
After trying the lipstick on the website, the chance that I will buy the lipstick is…
summarized in Table 1. improbable – probable 0.98
unlikely – likely 0.99
3.1.1. Stimulus material small – big 0.99
For the experimental task, we used an existing online application To be able to use this website to try cosmetics in the future, I am willing to give the following
from cosmetics brand Sephora (Sephora Virtual Artist; https:// information:
sephoravirtualartist.com), which enabled participants to use one of My name* 0.69
My address* 0.75
the three online product presentations: apply make-up products on a My date of birth* 0.81
model, on a picture of themselves or by using AR. The lay-out and My e-mail* 0.67
features of the application were exactly the same and only differed in My phone number* 0.78
the way the products were presented. For the online product pre- My hobbies* 0.78
My income* 0.78
sentation on a model (non-AR model), participants could choose be-
My media use* 0.79
tween four models and apply make-up on a picture of the chosen model. My purchase history of last month* 0.70
For the online product presentation on a picture of the self (non-AR EV 4.75 3,63 4.62 2.57 2.91 5.08
self), participants could take or upload a picture of their own face in the R2 0.79 0.91 0.92 0.86 0.97 0.56
application and apply make-up on their picture. For the online product α 0.95 0.97 0.98 0.92 0.99 0.90
M 4.66 5.15 2.85 5.13 4.00 3.09
presentation with AR, participants could virtually apply the make-up in SD 1.26 1.35 1.46 0.89 1.67 1.29
real time on their own face by using their live webcam or camera.
Note: INF = Perceived Informativeness, ENJ = Perceived Enjoyment,
3.1.2. Procedure INT = Perceived Intrusiveness, BA = Brand Attitude, PI = Purchase Intention,
An invitation with an URL to a Qualtrics questionnaire was dis- WSPD = Willingness to Share Personal Data; *Measured on 7-point Likert Scale.
tributed via e-mail and social media. The invitation instructed partici-
pants to only perform the study on a laptop or computer with a they chose, and were asked whether they had any problems with the
webcam, which was required for both the AR and the non-AR self- application. Participants that reported problems with the app (being
condition. In the questionnaire, participants first read a factsheet about unable to perform the experimental task) were excluded from further
the study and had to agree with the informed consent to start the study. participation. Then, participants filled in questions to measure the de-
Then, participants were randomly assigned to one of the three condi- pendent variables, mediators, and control variables. Lastly, they filled
tions: online product presentation on a model (N = 52), on a picture in a manipulation check asking how they had applied the make-up to
(N = 42), or AR (N = 38). They first had to read the instructions to check whether they had used the assigned condition. After filling out
perform the experimental task using the Sephora application. the questionnaire participants were thanked for their participation and
Participants were instructed to choose the assigned online product debriefed about the purpose of the study.
presentation from a menu. The experimental task was the same for each
condition: participants had to try several lipsticks using the Sephora 3.2. Measures
application and choose the lipstick that they liked best. After per-
forming the experimental task, participants had to fill in which lipstick The variables from the conceptual model; perceived informativeness
(Ahn et al., 2004), perceived enjoyment (Hassanein and Head, 2005),
(footnote continued) perceived intrusiveness (Li et al., 2002b), brand attitude (Li et al., 2002a),
camera in the AR condition or non-AR self condition, possible due to privacy purchase intention (Bearden et al., 1984), and willingness to share personal
concerns, and instead used the non-AR model condition. As this could have data (Rifon et al., 2005); were all measured with multiple items from
influenced reliability and validity of the results we excluded these participants. existing, validated scales and can be found in Table 2. The items were

5
A.R. Smink, et al. Electronic Commerce Research and Applications 35 (2019) 100854

adjusted to fit the context of this experiment. All variables formed re- Table 4
liable scales. Means and standard deviations per condition.
Additionally, several control variables were taken into account: use Variables Non-AR model Non-AR self AR (N = 38)
of cosmetics (daily/sometimes/no), online shopping experience, mea- (N = 52) (N = 42)
sured on a scale from 1 (very unexperienced) to 7 (very experienced),
a a b
Perceived informativeness 4.38 (1.40) 4.37 (1.09) 5.37 (0.95)
brand familiarity with Sephora (familiar/moderately familiar/un- a a b
Perceived enjoyment 4.84 (1.45) 4.88 (1.25) 5.88 (1.04)
familiar), and a pre-measure of brand attitude, measured on a scale Perceived intrusiveness 1.90 (0.97) a
2.90 (1.35) b
4.08 (1.21) c

from 1 (very negative) to 7 (very positive). Background variables age, Brand attitude 5.11 (1.03) a
5.02 (0.89) a
5.29 (0.68) a

a a b
gender, and education were also measured. The sample characteristics Purchase intention 3.48 (1.48) 3.87 (1.26) 4.87 (1.99)
a b c
are presented in Table 1. Willingness to share personal 2.52 (1.05) 3.15 (1.16) 3.81 (1.39)
data
Lastly, to check whether the experimental task was performed
correctly, we checked whether participants used the correct assigned Note: Table presents means per condition and standard deviation between
online product presentation. Participants were asked how they tried the brackets; abcMeans with different superscripts in the same row differ sig-
lipstick in the Sephora application, (1) on a model, (2) using a picture of nificantly in a post hoc Bonferroni tests at p < .05. Means with the same su-
myself, or (3) through a live connection with my webcam. As men- perscripts in the same row do not differ significantly.
tioned, participants that failed this check were excluded from further
analysis (N = 50). the non-AR self-condition (b = 1.00, SE = 0.27, p = .000), and the non-
AR model condition (b = 1.00, SE = 0.25, p = .000). Hence, the per-
4. Results ceived informativeness of AR hypothesis (H1a) was confirmed by the
data. Additionally, we expected that the non-AR self condition would
4.1. Control variables score higher than the non-AR model condition (H1b). However, there
was no difference between the non-AR self and the non-AR model
First, we checked whether the three experimental conditions dif- condition (b = 0.00, SE = 0.25, p = .994). Thus, the perceived in-
fered on any of the control variables. As can be seen in Table 3, cos- formativeness of the consumer’s own face hypothesis (H1b) was not
metics use, brand familiarity, education, online shopping experience, supported by the data.
and brand attitude (pre-measure) did not differ between the experi- Consequently, we expected that perceived informativeness would
mental conditions. However, there was a significant difference with positively mediate the effect of online product presentation on brand
respect to age, with the AR condition (M = 24.42, SD = 1.98) scoring attitude (H2a), purchase intention (H2b), and willingness to share
significantly higher than the picture condition (M = 23.00, SD = 2.04), personal data (H2c). The indirect effects that relate to these mediation
and the model condition (M = 23.40, SD = 2.20). A correlation ana- hypotheses can be found in Table 5. Each indirect effect is shown per
lysis showed that age did not correlate with the dependent variables or dependent variable, and for each of the three comparisons between the
mediators. Therefore, we decided not to include age as a covariate conditions (AR vs non-AR model, AR vs non-AR self, and non-AR model
when testing the hypotheses. vs non-AR self).
The results show that perceived informativeness had a positive di-
4.2. Hypotheses rect effect on brand attitude (b = 0.18, SE = 0.08, p = .029). However,
perceived informativeness did not mediate the effect of online product
To test the hypotheses, we used the PROCESS macro for SPSS presentation on brand attitude, as shown by the non-significant indirect
(version 3.2) from Hayes (2018) to calculate the direct and indirect effects in Table 5. The perceived informativeness mediation hy-
effects of the conceptual model (model 4). For each dependent variable pothesis was therefore rejected for brand attitude (H2a). However, the
(brand attitude, purchase intention, and willingness to share personal effect of online product presentation on purchase intention and will-
data), we calculated a parallel mediation model, with online product ingness to share personal data was positively mediated by perceived
presentation as independent variable, and perceived informativeness, informativeness, but only when comparing the AR condition to both
enjoyment and intrusiveness as mediators. Bootstrapping with 5000 non-AR conditions (self and model). The direct effects of perceived
samples was used to calculate 95% bias-corrected bootstrap confidence informativeness on purchase intention (b = 0.61, SE = 0.15, p = .000)
intervals (BCBCI) for the indirect effects (Hayes, 2018). The means per and willingness to share personal data (b = 0.39, SE = 0.12, p = .002)
condition can be found in Table 4 and all indirect effects from the were also significant. The indirect effects for comparing the two non-AR
mediation models can be found in Table 5. conditions were not significant. Thus, the perceived informativeness
mediation hypothesis was confirmed for purchase intention (H2b)
and willingness to share personal information (H2c), for the AR con-
4.2.1. Perceived informativeness hypotheses
dition (as opposed to the two non-AR conditions). In other words, the
First of all, we expected that online product presentation in AR
AR online product presentation led to a higher perceived informative-
would enhance perceived informativeness as opposed to both non-AR
ness as opposed to the two non-AR online product presentations, which
online product presentations (H1a). The data showed that the AR
consequently enhanced purchase intention and willingness to share
condition scored higher on perceived informativeness compared to both
personal data.
Table 3
Control variables.
4.2.2. Perceived enjoyment hypotheses
Variables χ2 (df) F(df) p With regard to perceived enjoyment, we expected that AR would
Cosmetics use 2.17 (2) 0.338 enhance perceived enjoyment, as opposed to the non-AR conditions
Brand familiarity 7.74 (4) 0.102 (H3a), and again a difference between the non-AR self and non-AR
Education 4.17 (4) 0.384 model condition (H3b). We found that the AR condition scored higher
Online shopping experience 0.59 (2, 129) 0.557 on perceived enjoyment than the non-AR self-condition (b = 1.00,
Brand attitude (pre-measure)** 2.56 (2, 119) 0.082
Age 4.98 (2, 129) 0.009*
SE = 0.29, p = .001), and the non-AR model condition (b = 1.03,
SE = 0.27, p = .000). The perceived enjoyment of AR hypothesis
Note: * Significant p < 0.05; ** The brand attitude pre-measure has missing (H3a) was accepted. There was no difference on perceived enjoyment
values and therefore a lower df is reported. between the two non-AR conditions (self and model; b = 0.03,

6
A.R. Smink, et al. Electronic Commerce Research and Applications 35 (2019) 100854

Table 5 but only for the difference between the AR condition versus both non-
Indirect effects of online product presentation on brand attitude, purchase in- AR conditions. There was no difference between the non-AR self and
tention and willingness to share personal data. non-AR model condition. As such, the perceived enjoyment media-
BC 95% CI tion hypothesis was accepted for brand attitude (H4a), but only for the
AR condition (as opposed to both non-AR conditions). The indirect ef-
B SE Lower Upper fects on purchase intention and willingness to share personal data, via
perceived enjoyment, were not significant. Thus, the perceived en-
Hypothesis 2 (The Perceived Informativeness Mediation Hypothesis)
H2a: Online product presentation → Perceived informativeness → Brand attitude joyment mediation hypothesis is rejected for purchase intention
Non-AR Model vs. Non- 0.00 0.05 -0.108 0.115 (H4b) and willingness to share personal data (H4c). In sum, online
AR Self product presentation in AR enhanced the perceived enjoyment as op-
Non-AR Model vs. AR 0.18 0.11 −0.012 0.428
posed to the online product presentations without AR, which conse-
Non-AR Picture vs. AR 0.18 0.11 −0.009 0.408
quently enhanced brand attitude.
H2b: Online product presentation → Perceived informativeness → Purchase intention
Non-AR Model vs. Non- 0.00 0.16 −0.341 0.317
AR Self
4.2.3. Perceived intrusiveness hypotheses
Non-AR Model vs. AR 0.61 0.20 0.263 1.056 With regard to perceived intrusiveness, we expected that perceived
Non-AR Picture vs. AR 0.61 0.20 0.278 1.048 intrusiveness would be higher in the AR condition as opposed to the
H2c: Online product presentation → Perceived informativeness → Willingness to share non-AR conditions (H5a), and we also expected a difference between
personal data the non-AR conditions (with the self condition scoring higher than the
Non-AR Model vs. Non- 0.00 0.10 −0.192 0.215 model condition; H5b). The results showed that the AR condition
AR Self
scored higher than the non-AR self-condition (b = 1.17, SE = 0.26,
Non-AR Model vs. AR 0.38 0.16 0.110 0.730
Non-AR Picture vs. AR 0.39 0.15 0.133 0.713
p = .000), and the non-AR model condition (b = 2.18, SE = 0.25,
p = .000). The perceived intrusiveness of AR hypothesis (H5a) was
Hypothesis 4 (The Perceived Enjoyment Mediation Hypothesis) accepted. We also found a significant difference between the non-AR
H4a: Online product presentation → Perceived enjoyment → Brand attitude picture and non-AR model condition (b = 1.00, SE = 0.24, p = .000).
Non-AR Model vs. Non- 0.01 0.08 −0.158 0.182
As such, the perceived intrusiveness of the consumer’s own face
AR Self
Non-AR Model vs. AR 0.28 0.12 0.087 0.556 hypothesis (H5b) is also accepted.
Non-AR Picture vs. AR 0.27 0.11 0.085 0.530 We expected that perceived intrusiveness would negatively mediate
H4b: Online product presentation → Perceived enjoyment → Purchase intention the effect of online product presentation on brand attitude (H6a),
Non-AR Model vs. Non- 0.00 0.06 −0.125 0.154 purchase intention (H6b), and willingness to share personal data (H6c).
AR Self As can be seen in Table 5, the indirect effects on brand attitude and
Non-AR Model vs. AR 0.19 0.16 −0.116 0.523 purchase intention, via perceived intrusiveness, were not significant.
Non-AR Picture vs. AR 0.18 0.15 −0.108 0.488
The perceived intrusiveness mediation hypothesis is rejected for
H4c: Online product presentation → Perceived enjoyment → Willingness to share brand attitude (H6a) and purchase intention (H6b). Against expecta-
personal data
tions, perceived intrusiveness had a positive effect on willingness to
Non-AR Model vs. Non- 0.00 0.03 −0.063 0.082
AR Self share personal data (b = 0.20, SE = 0.08, p = .013). Moreover, we also
Non-AR Model vs. AR 0.07 0.11 −0.153 0.286 found a significant and positive indirect effect of online product pre-
Non-AR Picture vs. AR 0.06 0.10 −0.151 0.278 sentation on willingness to share personal data, via perceived intru-
Hypothesis 6 (The Perceived Intrusiveness Mediation Hypothesis) siveness. These indirect effects were significant for the difference be-
H6a: Online product presentation → Perceived intrusiveness → Brand attitude tween the AR condition and both non-AR conditions, as well as the
Non-AR Model vs. Non- −0.05 0.06 −0.166 0.074 difference between the non-AR self and non-AR model condition. Thus,
AR Self
while the perceived intrusiveness increased for the non-AR self condi-
Non-AR Model vs. AR −0.10 0.13 −0.373 0.151
Non-AR Picture vs. AR −0.06 0.08 −0.226 0.072 tion (as opposed to non-AR model condition), and the AR condition (as
opposed to both non-AR conditions), this actually led to a higher
H6b: Online product presentation → Perceived intrusiveness → Purchase intention
Non-AR Model vs. Non- 0.18 0.12 −0.023 0.455 willingness to share personal data. Since the mediation effect was po-
AR Self sitive instead of negative, the perceived intrusiveness mediation
Non-AR Model vs. AR 0.40 0.24 −0.054 0.882 hypothesis is also rejected for willingness to share personal data (H6c).
Non-AR Picture vs. AR 0.22 0.14 −0.037 0.514 Lastly, we posed a research question (Research Question 1) on
H6c: Online product presentation → Perceived intrusiveness → Willingness to share whether the positive hypothesized mediating effects (via perceived in-
personal data formativeness and perceived enjoyment) would either be stronger or
Non-AR Model vs. Non- 0.20 0.10 0.029 0.435
weaker compared to the negative hypothesized mediating effect (via
AR Self
Non-AR Model vs. AR 0.54 0.20 0.073 0.846 perceived intrusiveness). However, in this study, we only found positive
Non-AR Picture vs. AR 0.24 0.12 0.036 0.499 mediating effects, and no negative mediating effects through perceived
intrusiveness. Therefore, the positive effects of AR outweighed the
Note. Unstandardized b-coefficients; BC 95% CI = 95% Bias corrected bootstrap potential negative effects in this study, because the latter were absent.
confidence interval using 5.000 bootstrap samples; bold means the indirect Implications are discussed.
effect is significant; N = 132.

5. Discussion
SE = 0.27, p = .899). Therefore, the perceived enjoyment of the
consumer’s own face hypothesis (H3b) was rejected.
5.1. Conclusions
Consequently, we expected that perceived enjoyment would also
positively mediate the effect of online product presentation on brand
This study adds to the literature on AR by (1) comparing the relative
attitude (H4a), purchase intention (H4b), and willingness to share
effect of online product presentation through AR versus two non-AR
personal data (H4c). First of all, perceived enjoyment had a positive
conditions (a picture of the self or a model), (2) considering three un-
direct effect on brand attitude (b = 0.27, SE = 0.08, p = .001).
derlying mechanisms that could yield both positive (perceived in-
Significant indirect effects showed that perceived enjoyment positively
formativeness and perceived enjoyment) and negative (perceived in-
mediated the effect of online product presentation on brand attitude,
trusiveness) consumer reactions, (3) considering the effect on brand

7
A.R. Smink, et al. Electronic Commerce Research and Applications 35 (2019) 100854

responses (brand attitude and purchase intention) and willingness to studies on personalized advertising (Gazley et al., 2015; McCoy et al.,
disclose personal information, and (4) exploring whether the positive 2008; van Doorn and Hoekstra, 2013) and did not negatively affect
effects of AR outweigh the potential negative effects. Four main con- consumer reactions as a consequence. Possibly, the positive effects of
clusions emerge from the research. online product presentation with AR on willingness to share personal
First, this study advanced the literature on AR by filtering out the data can partly be attributed to the sample used, since young people are
effect of seeing the self (compared to a model), versus the effect caused more used to share personal data to enable use of mobile applications
by AR (compared to the two non-AR conditions). In previous studies (Rice and Pearce, 2015), and could have lower level of privacy concerns
that compared effects of online product presentation with AR to non-AR in general (Hilken et al., 2017; Poushneh, 2018). Thus, while online
online product presentations, the conditions differed on both the product presentation involving the self (a picture or AR) can potentially
technology (AR or non-AR) and the presence of one’s own face or not raise the intrusiveness level, it seems from these results that these costs
(Baek et al., 2018; Verhagen et al., 2014). Results from this study are considered acceptable by users, and are therefore willing to disclose
showed that the effects on perceived informativeness and enjoyment personal information to use the technology.
were solely caused by AR (as opposed to both non-AR conditions). This brings us to our fourth conclusion, which is that within this
Online product presentation on a picture of the consumer’s own face study, the benefits of using AR for online shopping seem to outweigh
did not have an ameliorating effect compared to online product pre- the potential costs. As stated by equity theory (Adam 1963) and privacy
sentation on a picture of a model. We expected that seeing the self in calculus theory (Culnan and Armstrong, 1999), consumers make a cost-
the non-AR condition would serve as a self-reference cue which would benefit analysis when using, in this case, AR technology. As we only
increase processing of the content and the attribution of positive affect found positive mediating effects on consumer reactions, the benefits
towards this self-relevant information, as is posed in self-referencing (perceived enjoyment and perceived informativeness) of using online
theory (Kuiper and Rogers, 1979; Maslowska et al., 2016). However, in product presentation with AR outweighed the potential negative effect
this study context, it seems that self-referencing theory cannot be ap- of intrusiveness (which was absent).
plied to a picture of the self as a self-reference cue in order to enhance
informativeness and enjoyment. Rather, because AR enables an inter- 5.2. Limitations and future research
active and real-time view of the product on the consumer’s own face
(Azuma, 1997), it led to a more informative and enjoyable shopping There are some limitations which should be kept in mind when
experience, and these effects were not confounded by the presence of interpreting these results. This study focused on an existing make-up
ones’ own face. application, examined in a potential user group of AR applications
Second, both perceived informativeness and enjoyment positively (18–30 years old women). This limits the generalizability over a more
mediated the effects of online product presentation on brand responses diverse user population, including older people and men, which could
and willingness to share personal data, but the underlying mechanisms have differently affected the results. Regarding age, research has shown
seem to induce either behavioral or affective consumer reactions. More that young people are especially tech savvy and tend to have less
specifically, informativeness affected behavioral responses (purchase technology anxiety compared to older people (Rice and Pearce, 2015;
intention and willingness to share personal data), while enjoyment Vishwanath and Goldhaber, 2003). As a consequence, older people
enhanced affective responses (brand attitude). Thus, enhanced in- generally have more difficulties in using and adopting new technolo-
formativeness seems to lead to a cognitive process, in which the en- gies. Therefore, it is likely that they perceive AR applications as less
riching information offered through AR led to a decision-making pro- useful and enjoyable than younger people, and at the same time as more
cess in whether or not to buy the product and disclose personal intrusive. Regarding gender, research has shown that women find direct
information. At the same time, enhanced enjoyment seems to lead to an product experiences more important then men, and as such perceive a
affective process, in which the positive affect is transferred to the at- higher risk of shopping online (which provides a more indirect product
titude towards the brand. This is in line with research on branded apps, experience; Garbarino and Strahilevitz, 2004; Hasan, 2010). Therefore,
which showed that consumers’ use of information focused branded apps AR might be especially beneficial during online shopping for women (as
enhanced cognitive responses, while use of entertainment focused opposed to men) as it could reduce the risk of shopping online and
branded apps enhanced affective responses (van Noort and van provide a more realistic experience (Hilken et al., 2017). Thus, it could
Reijmersdal, 2019). Possibly, AR apps have more affective versus be- be that the positive effects found in this study would be weaker, or even
havioral effects depending on the specific goal to use AR apps, which negative when considering a more diverse population of both older
could be a more hedonic and/or utilitarian goal. For example, quali- people and men. Studies focusing on AR have mostly used young, stu-
tative research from Scholz and Duffy (2018) in which users of an AR dent samples, and differences with regard to sociodemographics have
makeup app were interviewed, illustrated that for some users, shopping been largely neglected (Suh & Prophet, 2018). Therefore, future re-
with AR was a purely hedonic, fun, and playful experience, without the search should consider more diverse populations, and examine whether
intention of actually buying something. However, other users reported gender and age influences the effectiveness and use of (different) AR
more on the use of AR for a more functional purpose (usefulness) to applications.
actually buy a product. This could be an interesting angle to explore in In relation to this, our study only showed positive effects of AR, and
a quantitative fashion in future research. no potential negative effects. However, this cost-benefit analysis could
Third, with regard to intrusiveness, both AR, and the presence of have different outcomes for people with different characteristics.
one’s own face increased the perceived level of intrusiveness of online Whether benefits of AR outweigh costs or otherwise, could partly be
product presentations. Thus, online product presentation on a picture of dependent upon the level of privacy concerns (Dinev and Hart, 2006;
one’s own face enhanced intrusiveness (compared to online product Wottrich et al., 2018), technology-related characteristics such as tech-
presentation on a model), and online product presentation in AR had an nology innovativeness (Huang and Liao, 2015), and demographics that
additional enhancing effect on intrusiveness (compared to both non-AR influence technology adoption as outlined in the previous paragraph
product presentations). While we expected that intrusiveness would (i.e., age; Rice and Pearce, 2015). Therefore, future studies could con-
lead to reactance and therefore would negatively affect consumer re- sider these personal characteristics as potential moderators to assess for
actions, we did not find any indirect effects on brand responses. whom AR has more positive, weaker, or even negative effects.
Contrary to our expectations, a higher perceived intrusiveness actually Part of our sample had to be excluded because they used a different
led to a higher willingness to disclose personal information. Although product presentation mode than they were assigned to. While this on-
the conditions that involved the presence of one’s own face raised in- line experimental set-up provided a more natural setting that enabled
trusiveness, this did not seem to lead to reactance, as has been found in participants to try the application in their own time and in their own

8
A.R. Smink, et al. Electronic Commerce Research and Applications 35 (2019) 100854

environment, this also enabled participants to choose a different online acceptance of Internet shopping malls. Electron. Commer. Res. Appl. 3, 405–420.
product presentation mode than they were assigned to. Most of these https://doi.org/10.1016/j.elerap.2004.05.001.
Azuma, R., 1997. A survey of augmented reality. Presence Teleoperators Virtual Environ.
participants used the online product presentation with a picture of a 6, 355–385. https://doi.org/10.1.1.30.4999.
model instead of the AR condition or the condition in which partici- Baek, T.H., Morimoto, M., 2012. Stay away from Me. J. Advert. 41, 59–76. https://doi.
pants needed to upload or take a picture of their own face. While org/10.2753/JOA0091-3367410105.
Baek, T.H., Yoo, C.Y., Yoon, S., 2018. Augment yourself through virtual mirror: the im-
participating in this study required all participants to have a webcam or pact of self-viewing and narcissism on consumer responses. Int. J. Advert. 37,
built-in camera, it could have been the case that participants turned out 421–439. https://doi.org/10.1080/02650487.2016.1244887.
to not fulfill these requirements and therefore were only able to use the Bearden, W.O., Lichtenstein, D.R., Teel, J.E., 1984. Comparison price, coupon, and brand
effects on consumer reactions to retail newspaper advertisements. J. Retail. 60,
online product presentation with the model. Future studies could ac- 11–34.
count for this by choosing a more controlled experimental setting. Blake, B.F., Neuendorf, K.A., Valdiserri, C.M., 2003. Innovativeness and variety of
Additionally, it could be that some participants were not willing to use Internet shopping. Internet Res. 13, 156–169. https://doi.org/10.1108/
10662240310478187.
their camera and therefore chose to use another online product pre-
Brehm, J.W., 1966. A theory of psychological reactance. Academic Press. Oxford,
sentation. A reason for this could be that these participants had higher England. https://doi.org/10.1002/hrdq.20027.
privacy concerns and were therefore not willing to give access to their Chen, K., Yen, D.C., 2004. Improving the quality of online presence through interactivity.
camera or a picture of their face. Therefore, it is important that future Inf. Manage. 42, 217–226. https://doi.org/10.1016/j.im.2004.01.005.
Culnan, M.J., Armstrong, P.K., 1999. Information privacy concerns, procedural fairness,
studies take into account variables that could influence whether people and impersonal trust: an empirical investigation. Organ. Sci. 10, 104–115. https://
are willing (or not) to use AR apps. doi.org/10.1287/orsc.10.1.104.
Lastly, this study considered intentions to buy products or disclosure Dacko, S.G., 2016. Enabling smart retail settings via mobile augmented reality shopping
apps. Technol. Forecast. Soc. Change 124, 243–256. https://doi.org/10.1016/j.
personal information. While this gives an indication of users’ responses techfore.2016.09.032.
towards AR applications and intended adoption, future studies could Dinev, T., Hart, P., 2006. An extended privacy calculus model for e-commerce transac-
take into account actual behavioral measures, such as real time buying tions. Inf. Syst. Res. 17, 61–80. https://doi.org/10.1287/isre.1060.0080.
Fang, H., Zhang, J., Şensoy, M., Magnenat-Thalmann, N., 2014. Reputation mechanism
behavior, actual personal data disclosure, and providing numbers on for e-commerce in virtual reality environments. Electron. Commer. Res. Appl. 13,
whether AR helps to reduce return rates. Moreover, actual user data 409–422. https://doi.org/10.1016/j.elerap.2014.08.002.
could also inform AR literature on the type of people that use and adopt Garbarino, E., Strahilevitz, M., 2004. Gender differences in the perceived risk of buying
online and the effects of receiving a site recommendation. J. Bus. Res. 57, 768–775.
AR. https://doi.org/10.1016/S0148-2963(02)00363-6.
Gazley, A., Hunt, A., McLaren, L., 2015. The effects of location-based-services on con-
5.3. Implications sumer purchase intention at point of purchase. Eur. J. Mark. 49, 1686–1708. https://
doi.org/10.1108/EJM-01-2014-0012.
Hamilton, R.W., Thompson, D.V., 2007. Is there a substitute for direct experience?
This study provides relevant insights to retailers and marketers comparing consumers’ preferences after direct and indirect product experiences. J.
wishing to invest in AR apps. The results indicate that for a potential Consum. Res. 34, 546–555. https://doi.org/10.1086/520073.
user group of AR applications, AR helps providing an online ‘try before Hasan, B., 2010. Exploring gender differences in online shopping attitude. Comput.
Human Behav. 26, 597–601. https://doi.org/10.1016/j.chb.2009.12.012.
you buy’ experience, delivering an informative and enjoyable product Hassanein, K., Head, M., 2005. The impact of infusing social presence in the web inter-
experience which can positively affect brand attitude and purchase face: an investigation across product types. Int. J. Electron. Commer. 10, 31–55.
intentions. Since AR specifically led to beneficial effects during the https://doi.org/10.2753/JEC1086-4415100202.
Hayes, A.F., 2018. Introduction to Mediation, Moderation, and Conditional Process
online shopping experience (and not merely showing a picture of the Analysis, Second Edition : A Regression-Based Approach, second ed. The Guilford
consumers’ face), it is recommended to invest in AR applications to Press, New York.
induce positive affective and behavioral responses. Additionally, the Hilken, T., de Ruyter, K., Chylinski, M., Mahr, D., Keeling, D.I., 2017. Augmenting the eye
of the beholder: exploring the strategic potential of augmented reality to enhance
results show that app builders should focus on both hedonic and utili- online service experiences. J. Acad. Mark. Sci. 45, 1–22. https://doi.org/10.1007/
tarian aspects in the AR shopping experience, as these seem to induce s11747-017-0541-x.
either affective or behavioral effects. Lastly, as shopping with AR in- Hilken, T., Heller, J., Chylinski, M., Keeling, D.I., Mahr, D., de Ruyter, K., 2018. Making
omnichannel an augmented reality: the current and future state of the art. J. Res.
creases users’ willingness to disclose personal data, AR could not only
Interact. Mark. 12, 509–523. https://doi.org/10.1108/JRIM-01-2018-0023.
be beneficial for brand responses, but could also help marketers to use Holland, R.W., Wennekers, A.M., Bijlstra, G., Jongenelen, M.M., Van Knippenberg, A.,
this information and provide more personalized solutions to consumers. 2009. Self-symbols as implicit motivators. Soc. Cogn. 27, 579–600. https://doi.org/
10.1521/soco.2009.27.4.579.
In sum, the results of this study add to the AR literature by showing
Hopp, T., Gangadharbatla, H., 2016. Novelty effects in augmented reality advertising
that the positive effects on consumer reactions were caused by AR, and environments: the influence of exposure time and self-efficacy. J. Curr. Issues Res.
not merely by the presence of one’s own face. Moreover, by considering Advert. 37, 113–130. https://doi.org/10.1080/10641734.2016.1171179.
potential positive and negative mediating mechanisms, this study Huang, T.-L., Liao, S.-L., 2017. Creating e-shopping multisensory flow experience through
augmented-reality interactive technology. Internet Res. 27, 449–475. https://doi.
showed that AR mainly had positive effects on brand responses and org/10.1108/IntR-11-2015-0321.
willingness to share data, through increasing perceived informativeness Huang, T.L., Liao, S., 2015. A model of acceptance of augmented-reality interactive
and enjoyment. At the same time, no negative effects were found technology: the moderating role of cognitive innovativeness. Electron. Commer. Res.
15, 269–295. https://doi.org/10.1007/s10660-014-9163-2.
through perceived intrusiveness. Thereby, it demonstrates that in this Huhn, A.E., Ketelaar, P., Khan, V.-J., Nuijten, K., van Gisbergen, M., 2011. I localize, I
context, the benefits of using the technology, seemed to outweigh the advertise and I create effect? A study on the perceived ad intrusiveness of location
potential costs in a potential user group of AR applications. based advertising within a virtual supermarket. Tijdschr. Voor Commun. 39, 21–41.
Javornik, A., 2016. ‘It’s an illusion, but it looks real!’ Consumer affective, cognitive and
behavioural responses to augmented reality applications. J. Mark. Manag. 32,
Funding 987–1011. https://doi.org/10.1080/0267257X.2016.1174726.
Kaushik, K., Kumar Jain, N., Kumar Singh, A., 2018. Antecedents and outcomes of in-
formation privacy concerns: role of subjective norm and social presence. Electron.
This research did not receive any specific grant from funding
Commer. Res. Appl. 32, 57–68. https://doi.org/10.1016/j.elerap.2018.11.003.
agencies in the public, commercial, or not-for-profit sectors. Kim, J., Forsythe, S., 2008. Sensory enabling technology acceptance model (SE-TAM): a
multiple-group structural model comparison. Psychol. Mark. 25, 901–922. https://
doi.org/10.1002/mar.20245.
References
Klein, L.R., 2003. Creating virtual product experiences: the role of telepresence. J.
Interact. Mark. 17, 41–55. https://doi.org/10.1002/dir.10046.
Adams, J.S., 1963. Toward an understanding of inequity. J. Abnorm. Soc. Psychol. 67, Kuiper, N.A., Rogers, T.B., 1979. Encoding of personal information: self-other differences.
422–436. https://doi.org/10.1093/past/96.1.22. J. Pers. Soc. Psychol. 37, 499–514. https://doi.org/10.1037/0022-3514.37.4.499.
Ahn, S.J.(Grace), Phua, J., Shan, Y., 2017. Self-endorsing in digital advertisements: using Li, H., Daugherty, T., Biocca, F., 2002a. Impact of 3-D advertising on product knowledge,
virtual selves to persuade physical selves. Comput. Human Behav. 71, 110–121. brand attitude, and purchase intention: the mediating role of presence. J. Advert. 31,
https://doi.org/10.1016/j.chb.2017.01.045. 43–57. https://doi.org/10.1080/00913367.2002.10673675.
Ahn, T., Ryu, S., Han, I., 2004. The impact of the online and offline features on the user Li, H., Daugherty, T., Biocca, F., 2001. Characteristics of virtual experience in electronic

9
A.R. Smink, et al. Electronic Commerce Research and Applications 35 (2019) 100854

commerce: a protocol analysis. J. Interact. Mark. 15, 13–30. marketing and consumer-brand relationships. J. Retail. Consum. Serv. 44, 11–23.
Li, H., Edwards, S.M., Lee, J.-H., 2002b. Measuring the intrusiveness of advertisements: https://doi.org/10.1016/j.jretconser.2018.05.004.
scale development and validation. J. Advert. 31, 37–47. https://doi.org/10.1080/ Scholz, J., Smith, A.N., 2015. Augmented reality: designing immersive experiences that
00913367.2002.10673665. maximize consumer engagement. Bus. Horiz. 59, 149–161. https://doi.org/10.1016/
Li, T., Meshkova, Z., 2013. Examining the impact of rich media on consumer willingness j.bushor.2015.10.003.
to pay in online stores. Electron. Commer. Res. Appl. 12, 449–461. https://doi.org/ Shih, C.(Eric), 1998. Conceptualizing consumer experiences in cyberspace. Eur. J. Mark.
10.1016/j.elerap.2013.07.001. 32, 655–663. https://doi.org/10.1108/03090569810224056.
Maslowska, E., Smit, E.G., van den Putte, B., 2016. It is all in the name: a study of con- Suh, A., Prophet, J., 2018. The state of immersive technology research: a literature
sumers’ responses to personalized communication. J. Interact. Advert. 16, 74–85. analysis. Comput. Human Behav. 86, 77–90. https://doi.org/10.1016/j.chb.2018.04.
https://doi.org/10.1080/15252019.2016.1161568. 019.
Mathwick, C., Malhotra, N., Rigdon, E., 2001. Experiential value: conceptualization, Tokunaga, R.S., 2013. Engagement with novel virtual environments: the role of perceived
measurement and application in the catalog and Internet shopping environment. J. novelty and flow in the development of the deficient self-regulation of internet use
Retail. 77, 39–56. https://doi.org/10.1016/S0022-4359(00)00045-2. and media habits. Hum. Commun. Res. 39, 365–393. https://doi.org/10.1111/hcre.
McCoy, S., Everard, A., Polak, P., Galletta, D.F., 2008. An experimental study of ante- 12008.
cedents and consequences of online ad intrusiveness. Int. J. Hum. Comput. Interact. van Doorn, J., Hoekstra, J.C., 2013. Customization of online advertising: the role of in-
24, 672–699. https://doi.org/10.1080/10447310802335664. trusiveness. Mark. Lett. 24, 339–351. https://doi.org/10.1007/s11002-012-9222-1.
Morgan, B. 2017. iPhone 8 Augmented Reality feature to revolutionize customer ex- van Noort, G., van Reijmersdal, E.A., 2019. Branded apps: explaining effects of brands’
perience. < https://www.forbes.com/sites/blakemorgan/2017/09/11/iphone-8- mobile phone applications on brand responses. J. Interact. Mark. 45, 16–26. https://
augmented-reality-feature-to-revolutionize-customer-experience/# doi.org/10.1016/j.intmar.2018.05.003.
299efda76a6c > (accessed January 4, 2019). van Noort, G., Voorveld, H.A.M., van Reijmersdal, E.A., 2012. Interactivity in brand web
Overmars, S., Poels, K., 2015. Online product experiences: the effect of simulating sites: cognitive, affective, and behavioral responses explained by consumers’ online
stroking gestures on product understanding and the critical role of user control. flow experience. J. Interact. Mark. 26, 223–234. https://doi.org/10.1016/j.intmar.
Comput. Human Behav. 51, 272–284. https://doi.org/10.1016/j.chb.2015.04.033. 2011.11.002.
Pantano, E., Rese, A., Baier, D., 2017. Enhancing the online decision-making process by Verhagen, T., Vonkeman, C., Feldberg, F., Verhagen, P., 2014. Present it like it is here:
using augmented reality: a two country comparison of youth markets. J. Retail. creating local presence to improve online product experiences. Comput. Human
Consum. Serv. 38, 81–95. https://doi.org/10.1016/j.jretconser.2017.05.011. Behav. 39, 270–280. https://doi.org/10.1016/j.chb.2014.07.036.
Pavlou, P.A., Anderson, A.G., Liang, H., Xue, Y., 2007. Understanding and mitigating Vishwanath, A., Goldhaber, G.M., 2003. An examination of the factors contributing to
uncertainty in online exchange relationship: a principal-agent perspective. MIS Q. 31, adoption decisions among late-diffused technology products. New Media Soc. 5,
105–136. 547–572. https://doi.org/10.1177/146144480354005.
Poushneh, A., 2018. Augmented reality in retail: a trade-off between user’s control of White, T.B., Zahay, D.L., Thorbjørnsen, H., Shavitt, S., 2008. Getting too personal: re-
access to personal information and augmentation quality. J. Retail. Consum. Serv. 41, actance to highly personalized email solicitations. Mark. Lett. 19, 39–50. https://doi.
169–176. https://doi.org/10.1016/j.jretconser.2017.12.010. org/10.1007/s11002-007-9027-9.
Rice, R.E., Pearce, K.E., 2015. Divide and diffuse: comparing digital divide and diffusion Wottrich, V.M., van Reijmersdal, E.A., Smit, E.G., 2018. The privacy trade-off for mobile
of innovations perspectives on mobile phone adoption. Mob. Media Commun. 3, app downloads: The roles of app value, intrusiveness, and privacy concerns. Decis.
401–424. https://doi.org/10.1177/2050157915590469. Support Syst. 106, 44–52. https://doi.org/10.1016/j.dss.2017.12.003.
Rifon, N.J., Larose, R., Choi, S.M., 2005. Your privacy is sealed: effects of web privacy Yan Tam, K., Ying Ho, S., 2005. Web personalization as a persuasion strategy: an ela-
seals on trust and personal disclosures. J. Consum. Aff. 39, 339–362. https://doi.org/ boration likelihood model perspective web personalization as a persuasion strategy
10.1111/j.1745-6606.2005.00018.x. an elaboration likelihood model perspective. Inf. Syst. Res. 16, 271–291. https://doi.
Schlosser, A.E., 2003. Experiencing products in the virtual world: the role of goal and org/10.1287/isre.1050.0058.
imagery in influencing attitudes versus purchase intentions. J. Consum. Res. 30, Yim, M.Y.-C., Chu, S.-C., Sauer, P.L., 2017. Is augmented reality technology an effective
184–198. https://doi.org/10.1086/376807. tool for e-commerce? an interactivity and vividness perspective. J. Interact. Mark. 39,
Schoenbachler, D.D., Gordon, G.L., 2002. Trust and customer willingness to provide in- 89–103. https://doi.org/10.1016/j.intmar.2017.04.001.
formation in database-driven relationship marketing. J. Interact. Mark. 16, 2–16. Yim, M.Y.-C., Park, S.-Y., 2018. “I am not satisfied with my body, so I like augmented
https://doi.org/10.1002/dir.10033. reality (AR)”: consumer responses to AR-based product presentations. J. Bus. Res.
Scholz, J., Duffy, K., 2018. We ARe at home: how augmented reality reshapes mobile https://doi.org/10.1016/J.JBUSRES.2018.10.041.

10