AR you ready? The Effects of Augmented Reality on Brand Attitude and Purchase Intention

AR you ready?

The Effects of Augmented Reality on Brand Attitude and Purchase Intention

Jingjing Wang 12011088

Master’s Thesis

Graduate School of Communication Master’s Programme Communication Science

Supervisor: Sophie Boerman Date of completion: June 28, 2019

Abstract

Augmented reality (AR) is an emerging technology that enables people to interact with virtual objects in an unprecedented way. AR experience is seamlessly interwoven with the physical world such that it is perceived as an immersive aspect of the real environment. Marketeers of the shopping industry seem to be especially interested in AR marketing as it makes the online shopping experience far more immediate than before. However, little is known about whether the use of AR is effective in affecting consumers’ brand perceptions and further arousing their interests in the products. This study adopted an experimental design (N = 83) to explore the effects of AR technology (present or absent in two IKEA apps) on brand attitudes and purchase intention based on the Technology Acceptance Model (Davis, 1989). Our study results indicated that the perceived ease of use was a key factor that

mediates the effects of AR on purchase intention: non-AR app was perceived as easier to use than the AR app, consequently, people in the non-AR condition liked the brand more and wanted to buy the products more. Perceived usefulness, however, did not mediate the effects of AR on brand attitudes or purchase intention. Therefore, AR technology is a double edged sword which requires thought-through planning to make the best of its unique attributes.

Keywords: Augmented Reality, Technology Acceptance Model, perceived usefulness, perceived ease of use, brand attitude, purchase intention, IKEA

Introduction

In the summer of 2016, a mobile game named Pokémon Go had impressed the game industry. It had more than 500 million downloads in two months (Takahashi, 2016) and generated revenues of $470 million in 82 days (Minotti, 2016). What makes it stand out from the rest? Many people attribute its success to the innovative application of augmented reality technology.

Augmented reality (AR), which describes the visual alignment of virtual content with real-world contexts (Carmigniani & Furht, 2011; Javornik, 2016; Scholz & Smith,

2016), is a topic of immense interest for mobile marketing (Shankar et al., 2016) and retailing research (Grewal et al., 2017). Leading brands like IKEA, Wayfair, and Sephora have all introduced mobile AR shopping apps that enable consumers to virtually “try out” products on their own bodies or in their own spaces. The market size for AR was 640.2 million in 2015 and is expected to generate $120 billion in revenue by 2020 (Merel, 2015).

A literature review by Javornik (2016) indicated that virtual experience positively impacts brand attitude (Kim & Lennon, 2008), consumer intentions towards the purchase (Jin, 2009; Gabisch, 2011), and willingness to pay for both search and experience products (Li & Meshkova, 2013). However, Goel and Prokopec (2009) also pointed out that compared to ordinary websites, virtuality application may result in lower informativeness and consumer trust. The mixed results of previous studies have suggested that AR is a double edged sword which requires thought-through planning to make the best of its unique attributes. Given the popularity and huge potential of AR, it is worth investigating how consumers respond to this new technology and what factors contribute to the successful application of AR.

This study will explore the effects of AR-apps mainly based on the framework of the Technology Acceptance Model (TAM; Davis, 1989). The TAM is evaluated as a superior model in terms of conciseness and predictability in many fields, which explains people’s

acceptance of technology rooted in the theory of reasoned action (TRA; Ajzen & Fishbein, 1977). The TAM suggests that perceived usefulness (PU) and perceived ease of use (PEOU) are two major determinants of individual attitudes, including the attitude to construct

intention to use the technology and cause behavior (Davis, 1989). Several studies have found a positive effect of AR on perceived ease of use because its 3D presentation facilitates visualization of objects and its visual attractiveness positively affect simplicity of

engagement (Chung, Han, & Joun, 2015; Kim & Forsythe, 2008; Mayer & Moreno, 2003). AR-experience was also perceived as more useful than non-AR experience because it offers more in-depth information (Chen & Tan, 2004; Huang & Liu, 2014; Mahony, 2015) and enables people to have a more direct product experience (Mahony, 2015; Rese, Baier, Geyer-Schulz, & Schreiber, 2017) than non-AR experience. Therefore, AR was believed to be exerting a positive effect on perceived ease of use and perceived usefulness, and based on the propositions of the TAM (Davis, 1989), perceived ease of use and perceived usefulness would lead to more positive attitudes and purchase intention. Therefore, this study aims to shed light on the following research question:

RQ: Does AR influence perceived usefulness (PU) and perceived ease of use (PEOU), and do PU and PEOU mediate the effects of AR on brand attitude and purchase intention?

Theory Effects of AR on perceived usefulness

AR app (vs. non-AR app) has a positive effect on perceived usefulness of the app because AR-experience offers more in-depth and quality information (Chen & Tan, 2004; Huang & Liu, 2014; Mahony, 2015) and enables people to have a more direct product experience (Mahony, 2015; Rese, Baier, Geyer-Schulz, & Schreiber, 2017) than non-AR

experience. Perceived usefulness is originally defined by Davis (1989) as the prospective user's subjective probability that using a specific application system will increase his or her job performance within an organizational context. However, the goal of Davis’ study (1989) is to better understand why people (especially professionals) accept or reject computers. Therefore, in order to incorporate the concept of perceived usefulness into our study, we need to take into account the following limitations of the original definition: 1) it focuses on

professionals only, but we are interested in the general consumers; 2) it restricts usefulness to job-related benefits, but we are interested in usefulness in a broader sense. To better map this concept onto our study, we’ll adopt an adapted definition of perceived usefulness, which refers to “the degree to which a person believes that using a particular system would be beneficial and advantageous” (Manis & Choi, 2018). The informativeness and more direct product experience offered by AR-experience makes it beneficial and advantageous for

consumers, therefore AR-apps are expected to be perceived as more useful than non-AR apps. AR experience is regarded as more useful than non-AR experience in terms of

information depth and quality (Chen & Tan, 2004; Mahony, 2015). According to Mahony (2015), AR technology turns a two-dimensional portray of an object into three-dimensional, thus the image-based stimuli become animated which would increase the volume of

information derived by the recipient. The users of the AR app can obtain more

comprehensive information about the products than those of the non-AR app. Furthermore, the quality and usefulness of the information is determined by the degree to which consumers can use the information to predict their satisfaction with the product prior to the actual

purchase. Relevant and useful product information enables consumers to increase clarity about the product and to come to a satisfactory product choice (Chen & Tan, 2004). In our case, the AR app (IKEA Place) offers the product information in a more personalized way because the consumers are capable of adjusting the 3D model of that products according to

their own needs, and information such as size and color are generally regarded as the key and most relevant information of online shopping experience. However, the non-AR counterpart (IKEA Store) offers each user with the same set of pre-selected product images, which are less personalized and informative than those in the AR app. Therefore, the AR app

outperforms the non-AR app in terms of informativeness because it provides a more in-depth and relevant product information which helps with predicting product satisfaction.

The usefulness of AR also lies in its capability of offering consumers with more direct and uninterrupted interactions with abundant product information (Alba et al., 1997). Product experience can be categorized as either direct (e.g. trying a product in store) or indirect (e.g. watching an ad), and the virtual experience falls between direct and indirect on the

experience spectrum (Schwartz, 2011). With virtual experience, “customers can interact with the products in a way that is similar to what they can expect in the real world, but the virtual experience can also be slightly different in order to catch the users’ attention and persuade them to try the product in reality” (Fiore et al., 2005; Papagiannidis et al., 2014). In our study, the AR app (IKEA Place) enables consumers to have a more direct product experience by offering them personalized pre-purchase evaluations through virtually “trying out” the products online, but the non-AR app (IKEA Store) offers the consumers no option of trying out the products in any form prior to the actual purchase. Therefore, the more direct

customer-product interaction offered by the AR app helps the consumers to “better evaluate the online product prior to purchase, thereby reducing the likelihood of poor choice” (Kim & Forsythe, 2008). While the users of the non-AR app follow a buy-experience-think consumer decision journey (McKinsey, 2015), the AR app offers the consumers with an alternative experience-buy-think option, which better taps into the needs of contemporary consumers. Especially for furniture, the product category of our study focus, the product size and its fitfulness to the environment are very important in consumer decision making, however, it is

very difficult to present such key characteristics through 2D images and text descriptions. Therefore, in the context of online furniture shopping, AR app is more useful than non-AR app because “it provides proxy sensory experiences that can serve as a surrogate for direct examination when evaluating furniture online” (Kim & Forsythe, 2008). Hence, the

following hypothesis:

H1a: Compared with non-AR apps, AR-apps lead to higher perceived usefulness of the app.

Effects of AR on Perceived Ease of Use

AR app (vs. non-AR app) has a positive effect on perceived ease of use because AR-experience is less likely to result in cognitive overload and its 3D view is easier to use than the 2D view of non-AR experience, which would contribute to higher perceived ease of use (Davis, 1989; Kim & Forsythe, 2008; Raska & Richter, 2017). Perceived ease of use refers to the degree to which the prospective user expects the target system to be free of effort (Davis, 1989). An effort is an unlimited resource that could be allocated by an individual to a specific activity (Radner & Rothschild, 1975). There have been mixed results regarding the perceived ease of engaging in AR-experience: on the one hand, there are elements of AR which are currently inherent to the medium and are thus difficult to simplify, such as “the need to download an extraneous application or the need to hold the device appropriately” (Mahony, 2015); on the other hand, AR-experience makes it easier to visualize decision outcomes and visualization makes an event look more real (Macinnis & Price, 1987). This study leans towards the latter stance more because we believe that AR experience has a lower likelihood of causing cognitive overload and induces stronger telepresence than non-AR experience, which makes the AR app easier to use than the non-AR app (Davis, 1989; Kim & Forsythe, 2008; Raska & Richter, 2017). Also, more practically, the participants in our study do not

have to spare any effort of downloading an app because we’ve prepared the app for them, so the claim by Mahony (2015) that “some elements of AR requires extraneous effort” does not apply to our study that much.

Compared with AR users, it is less likely for the AR users to experience cognitive overload because the AR-experience moves some essential processing from verbal channel to visual channel (Mayer & Moreno, 2003). According to the Cognitive Theory of Multimedia Learning (Mayer & Moreno, 2003), humans process separate information processing channels for verbal and visual material, and cognitive overload occurs when the total processing exceeds the learner’s cognitive capacity. In the context of our study, the users of the non-AR app (IKEA Store) have to combine the text product information (verbal) and the product images (visual) together in order to visualize how the piece of furniture looks like in real settings, this would potentially lead to cognitive overload because the text and images are not presented altogether simultaneously. However, the users of the AR app (IKEA Place) can easily use the AR function to view the furniture as well as how it fits into the space by

focusing on the visual information only, and they do not have to process both verbal and visual information at the same time. Therefore, AR-experience saves the users the effort of processing multi-modality information and is expected to be easier to engage with than the non-AR experience.

Besides, the 3D view of AR experience is easier to use than 2D view of non-AR experience (Kim & Forsythe, 2008). As a core characteristic of AR, virtual annotations represent an important part of AR (Billinghurst & Kato, 2002). With virtual annotations, AR is responsive in a way that users are able to have control over the technology and lead a two-way communication between themselves and the technology (Song & Zinkhan, 2008; Sundar, 2009; van Noort et al., 2012). To be more specific, while the non-AR app (IKEA Store) in our study only offers static 2D product images which can only be passively consumed by the

app users, the AR app (IKEA Place) present the users with a 3D model of the product which they can proactively interact with. Therefore, the AR app is easier to use because it helps consumers visualize the product with an interactive 3D presentation. Hence, the following hypothesis:

H1b: Compared with non-AR apps, AR-apps lead to higher perceived ease of use of the app.

Effects of AR on Brand Attitude

Perceived usefulness and perceived ease of use positively influence brand attitude, and AR app (vs. non-AR app) has an indirect positive effect on brand attitude via perceived usefulness and perceived ease of use. Brand attitude is defined as “consumers’ willingness to constantly respond and show a desirable or undesirable reaction toward a particular brand” (Yim et al., 2014). Davis (1989) pointed out that perceived ease of use and perceived

usefulness are two significant determinants of attitude formation. The more useful a

technology is perceived as, the more likely the users would develop positive attitudes towards that technology (Davis, 1989). Besides, the easier a technology is to interact with, the greater should be the user’s sense of efficacy (Bandura, 1982) and personal control (Lepper, 1985) regarding his or her ability to carry out the sequences of behavior needed to perform the required task (Davis, 1989). Several studies have applied the propositions of the TAM to the VR/AR industry and confirmed that perceived usefulness and perceived ease of use are important factors in predicting attitudes towards using virtual try-on (Manis & Choi, 2018; Kim & Forsythe, 2008). It is worth noting that the original TAM only proposed perceived usefulness and perceived ease of use as predictors of attitudes towards that technology, however, our study are more interested in the effects of these two factors (i.e. PU and PEOU) on brand attitudes, instead of attitudes towards AR technology. Therefore, we also looked at

some other research (Elliot & Speck, 2005; Kim, Kim, & Park, 2010; Raska & Richter, 2017) which found a direct effect of perceived usefulness and perceived ease of use on brand

attitude.

Perceived usefulness and perceived ease of use can facilitate a customer’s purchase decision process (Kim, Kim, & Park, 2010) and to improve attitudes toward that brand (Elliot & Speck, 2005). First, utilitarian value (i.e. perceived usefulness) positively affect brand attitude (Raska & Richter, 2017). Previous research has pointed out the importance of functional motivations in online shopping (Childers et al., 2001), and perceived usefulness played a key role in consumers’ functional motivations. Online consumers are willing to collect information regarding specific product over the web, and they would evaluate the brand more positively if they find the online shopping experience offered by the brand as useful. Therefore, if the users perceive the IKEA shopping app as useful, they would tend to develop positive attitudes towards IKEA as well. Secondly, research has confirmed that perceived ease of use is an important factor in predicting brand attitudes (Davis et al., 1992; Heijden, 2000). According to Rogers (1995), perceived ease of use positively relates to individual’s willingness to engage with a brand and to further foster positive brand attitudes. Therefore, if the users perceive the IKEA shopping app as easy to use, they would like the brand IKEA more and evaluate IKEA more positively.

In summary, by combining the propositions of the TAM (Davis, 1989) and findings of previous research (Elliot & Speck, 2005; Kim, Kim, & Park, 2010), we are hypothesizing an indirect effect of AR technology on brand attitude via perceived usefulness and perceived ease of use. Hence the following hypothesis:

H2: AR-apps (compared with non-AR apps) positively affect brand attitude via a) perceived usefulness; b) perceived ease of use of the app.

Simple Mediation Effect of AR on Purchase Intention

Perceived usefulness and perceived ease of use positively influence purchase intention, and they play a mediating role in the effects of AR on purchase intention. Purchase intention refers to “the likelihood that consumers will plan or be willing to purchase a certain product or service in the future” (Wu, Yeh, & Hsiao, 2011). Firstly, previous research has confirmed that information helpfulness (i.e. perceived usefulness) positively influences consumers' purchase intention (Filieri, McLeay, Tsui, & Lin, 2018). Filieri et al. (2018) argued that “if users believe that the information provided by the brand will help them to become familiar with a product/service and evaluate its quality and performance before purchase, they will be more likely to purchase that product/service.” Besides, perceived usefulness not only lies in the quality and depth of information, but also in how such information is presented. A core and inseparable part of any app is its UX design (i.e. interface usability or its aesthetics), namely, perceived usefulness is also demonstrated through the flexibility and usability of that app (Stoyanova, Brito, Georgieva, & Milanova, 2015). Research has shown that purchase intention has a strong relation with interface usability (Stoyanova et al., 2015). Therefore, the more helpful the information offered by the app and the higher usability of the app, the more likely the users tend to have stronger interests in the products and thus higher purchase intention. Secondly, several studies have emphasized the importance of developing easy-to-use apps (i.e. perceived ease of easy-to-use) to increase intention to buy (Filieri et al., 2018;

Stoyanova et al., 2015). A study by Manis and Choi (2018) indicated that a consumer’s perceptions of ease of use are positive predictors of purchase intention and suggested firms with technology in pre prototype or development stages to refine the product to be easier to use. Complexity, the antithesis of ease of use, would reduce consumers' willingness to engage with a brand. For instance, if an app is designed in a way that the product information is hard to find or to be processed, this would lead to consumer reactance of using this app.

Consequently, the consumers might develop negative feelings of the brand and may even exclude them from the consideration set, thus lower purchase intention. Therefore, the easier to use an app is, the higher the purchase intention is for the product in that app. In summary, based on our propositions of H1, we therefore hypothesize the following:

H3: AR-apps (compared with non-AR apps) positively affect purchase intention via a) perceived usefulness; b) perceived ease of use of the app.

Serial Mediation Effect of AR on Purchase Intention

Brand attitude has a positive influence on purchase intention, and together with perceived usefulness and perceived ease of use, they play a mediating role in the effects of AR on purchase intention. A key principle of TRA is that attitudes fully mediate the effects of beliefs on intentions. Similarly, the TAM (Davis, 1989) pointed out that people form intentions to perform behaviors towards which they have positive affect. Several studies have identified a positive effect of brand attitude on purchase intention (Park, Jeon, & Sullivan, 2015; Stoyanova et al., 2015; Wu et al., 2011), and consumers who have favorable attitudes toward online shopping are less likely to abort intended transactions (Cho & Cheon, 2004). Hence, further developed on our propositions of H3, we hypothesize the following:

H4: AR-apps (compared with non-AR apps) positively affect purchase intention indirectly through a) perceived usefulness; b) perceived ease of use, which consequently influences brand attitude.

App type (AR vs. non-AR) Perceived usefulness Perceived ease of use H2a H2b Brand attitude Purchase Intention H3a H1a H1b H4 H3b

Figure 1. Conceptual Model

Method Design and Participants

We adopted a one-factor (app type: AR vs. non-AR) between-subject experimental design. A total of 85 participants were recruited through convenience sampling. They were either students at the University of Amsterdam or employees at Elsevier, who were chosen because they were in line with IKEA’s target group: students, employees, professionals (Dudovskiy, 2019). Participants were 60.2% female and 39.8% male, aged between 20 and 64 (M = 31.23, SD = 10.97). 86% of the participants have at least completed a bachelor’s degree, and 59% of the participants are employed. They were randomly assigned to either control or treatment condition. Given that the effects of augmented reality (AR) is the focus of this study, people who were assigned to the treatment condition but did not use the AR function as instructed were excluded from the statistical analyses. Together, 2 participants were excluded. The final sample size is 83. There are 43 participants in the AR app condition and 40 participants in the non-AR app condition.

Procedure

The data collection was conducted in Amsterdam during the weekdays from May 2, 2019 to May 20, 2019. The participants were approached in public areas of University of Amsterdam (UvA) or the Elsevier company building because there are ample space for testing the AR function. There are 45 UvA students and 38 Elsevier employees. The participants were randomly assigned to one of the experimental conditions through the following procedure: I myself was the researcher and I recruited the participants by firstly presenting them with a printed factsheet and informed consent form. If they agreed to

participate, they were asked to pick one paper note out of an envelope containing 100 notes in total. On each paper note, there is a participant number and the corresponding condition. Participant number 1 to 50 were assigned to the treatment group while participant number 51 to 100 were assigned to the control group. After signing the informed consent, both groups were given an instruction asking them to search for a chair called “PELLO” through an IKEA app and to spend 2 to 3 minutes on the product page. The IKEA app specified in the

treatment condition is IKEA Place (AR-app), and that for the control condition is IKEA Store (non-AR app). Also, besides browsing the product page, participants in the treatment

condition were also explicitly asked to use the AR function (i.e. “try it in your place” function). After finishing using the app, participants were asked to fulfill an online

questionnaire during which their perceived usefulness and perceived ease of use of the app, brand attitude towards IKEA, purchase intention of the PELLO chair as well as demographics and several control variables were measured. At the end of the questionnaire, the participants were thanked and debriefed about the experiment.

Measures

Perceived Ease of Use

We measured perceived ease of use of the app by asking participants to indicate on a 7-point scale (1 = extremely disagree, 7 = extremely agree) to what extent they agreed with five statements (Gefen et al., 2017; Venkatesh & Davis, 2000). The statements are as follows: “My interaction with the IKEA app is clear and understandable”, “I find it easy to get the IKEA app to do what I want it to do”, “Using the IKEA app does not require a lot of mental effort”, “The IKEA app is flexible to interact with”, and “The IKEA app is easy to use.” Factor analysis revealed that the items load on one factor (eigenvalue = 3.51, explained

variances = 70.28%, Cronbach’s alpha = .89). The means of these five items is used as measurement of perceived ease of use (M = 5.51, SD = 1.19).

Perceived Usefulness

We measured perceived usefulness of the app by asking participants to indicate on a 7-point scale to what extent they agreed with five statements (Gefen et al., 2017; Venkatesh & Davis, 2000; Rese et al., 2017). The statements are as follows: “The IKEA app enhances shopping effectiveness”, “The IKEA app is useful for shopping”, “The IKEA app helps me make better purchase decisions”, “The IKEA app helps me visualize what the actual piece of product is”, “The IKEA app provides useful product information.” Factor analysis revealed that the items load on one factor (eigenvalue = 2.58, explained variances = 51.66%,

Cronbach’s alpha = .75). The means of these five items is used as measurement of perceived usefulness (M = 5.72, SD = 0.78).

Brand Attitude

To measure brand attitude, we adopted a validated scale by Spears and Singh (2004). Participants were asked to indicate their overall feelings towards the brand IKEA using a 7-point semantic differential scale with 5 items (“Unlikable/likable”, “Unpleasant/pleasant”, “Unfavorable/favorable”, “Unappealing/appealing”, “Good/bad”). Factor analysis revealed that the items load on one factor (eigenvalue = 3.37, explained variances = 67.41%,

Cronbach’s alpha = .87). The means of these five items is used as measurement of perceived usefulness (M = 5.80, SD = 0.75).

Purchase Intention

Purchase intention is measured by asking people to what extent do they intend (1 = extremely unlikely, 7 = extremely likely) to buy the PELLO chair they just searched for

(Spears & Singh, 2004). The higher the score is, the more likely people are going to buy that chair (M = 3.94, SD = 1.68).

Control Variables

To ensure that the effects of AR were not caused by other differences between the experimental groups, a number of control variables were measured. Participants were asked whether they have had any experience (1 = yes, 2 = no) with the IKEA brand, the IKEA app, and AR technology. 94.1% of the participants have shopped at IKEA before, but only 18.8% of the participants have ever used the app they were assigned to, and 62.4% of the

participants have experience with AR technology before. They were also asked about their familiarity (1 = never heard of it, 2 = heard of it, 3 = know a little, 4 = know a fair amount, 5 = know it well) with the IKEA brand (M = 4.53, SD = 0.72), the IKEA app (M = 2.13, SD = 1.12), and AR technology (M = 3.11, SD = 1.01). Their home ownership status (1 =

personally owned, 2 = rented) and preferred shopping method of furniture (1 = online shopping, 2 = in-store shopping) were also measured. 65.1% of the participants are renting houses while 34.9% of the participants own the house themselves. 78.8% of the participants prefer in-store shopping for furniture while 21.1% prefer buying furniture online. The shopping frequency of furniture (M = 2.07, SD = 0.46) was measured by asking participants how often they bought furniture (1 = never, 2 = yearly, 3 = quarterly, 4 = monthly, 5 = weekly). In addition, age, gender, employment status and education were measured (the descriptive statistics were mentioned in the participants section).

Results Randomization Check

The experimental groups did not differ with respect to gender, χ2 (1) = 0.89, p = .347; age, F = 0.67, p = .419; education, χ2 (4) = 2.34, p = .674; employment, χ2 (3) = 0.92, p

= .821; home ownership, χ2 (1) = 0.22, p = .637; furniture shopping frequency, χ2 (2) = 0.24, p = .887. With regard to previous relevant experience, the experimental groups do not

significantly differ in terms of brand use, χ2 (1) = 0.30, p = .586; AR use, χ2 (1) = 0.001, p = .978; brand familiarity, χ2 (4) = 4.36, p = .360; AR familiarity, χ2 (4) = 0.76, p = .944. However, there are significant differences between the experimental groups in terms of participants’ app use, χ2

(1) = 4.64, p = .031; app familiarity, χ2 (4) = 12.27, p = .015; preferred shopping method, χ2 (1) = 6.85, p = .009. Therefore, app use, app familiarity and shopping method were included as covariates in all analyses to control for any confounding effects. Table 2 shows the descriptive statistics for all dependent variables/mediators for the experimental groups.

TABLE 2

Descriptive Statistics for the Experimental Conditions

Variable AR app (IKEA Place) Non-AR app (IKEA Store)

Perceived usefulness 5.74 (0.67)a 5.69 (0.89)a

Perceived ease of use 5.00 (1.26)a 6.07 (0.81)b

Brand attitude 5.78 (0.75)a 5.82 (0.75)a

Purchase intention 3.81 (1.60)a 4.08 (1.77)a

Note. All variables are scaled from 1 to 7. N = 83: AR-app n = 43, non-AR app n = 40. a, b Means with a different superscript in the same row differ significantly at p < .05. Effects on Perceived Usefulness

To test the effects of app type (AR-app vs. non-AR app) on perceived usefulness (H1a), a one-way ANCOVA was conducted with app type as independent variable, perceived usefulness as dependent variable. App use, app familiarity, shopping method were included as covariates. No significant differences in perceived usefulness were found among two app type conditions, F (1,82) = 0.00, p = .997. People who are assigned to the AR app (M = 5.74, SD = 0.67) did not perceive the app as more useful than people who were assigned to the non-AR app (M = 5.69, SD = 0.89). Therefore, Hypothesis 1a was rejected.

Effects on Perceived Ease of Use

To test the effects of app type (AR-app vs. non-AR app) on perceived ease of use (H1b), a one-way ANCOVA was conducted with app type as the independent variable, perceived ease of use as the dependent variable. App use, app familiarity, shopping method were included as covariates. A significant effect of app type on perceived ease of use was found, F (1,82) = 20.54, p < .001. However, contrary to our expectations, people who were assigned to the non-AR app (M = 6.07, SD = 0.81) actually perceived the app as significantly easier to use than people who were assigned to the AR app (M = 5.00, SD = 1.26). Therefore, Hypothesis 1b was rejected.

Effects on Brand Attitude

To test the effects of app type on brand attitude mediated by perceived usefulness (H2a) and perceived ease of use (H2b), we used Hayes’ PROCESS macro model 4 (Hayes, 2012). App use, app familiarity, shopping method were included as covariates.

There was no significant total effect (total effect = .09, se = 0.19, p = .638) or direct effect (direct effect = -.06, se = 0.20, p = .751) of app type on brand attitude. There was also no significant indirect effect of app type on brand attitude via perceived usefulness (indirect

effect = .00, boot SE = 0.06, BCBCI = [-.11, .14]) or via perceived ease of use (indirect effect = .15, boot SE = 0.10, BCBCI = [-.05, .35]). The results from PROCESS macro model 4 indicated that app type was not a significant predictor of perceived usefulness, b = -0.00, se = 0.20, p = .997, but perceived usefulness was a significant positive predictor of brand attitude, b = 0.27, se = 0.12, p = .029. These results did not provide sufficient support for Hypothesis H2a, therefore H2a was rejected. Results also indicated that app type was a significant

predictor of perceived ease of use, b = 1.20, se = 0.26, p < .001, but perceived ease of use was not a significant positive predictor of brand attitude, b = 0.13, se = 0.09, p = .159. Therefore, there was no sufficient evidence to support H2b.

Effects on Purchase Intention

Simple Mediation. To test the effects of app type on purchase intention mediated by perceived usefulness (H3a) and perceived ease of use (H3b), we used Hayes’ PROCESS macro model 4 (Hayes, 2012). App use, app familiarity, and shopping method were included as covariates.

There was no significant total effect (total effect = .51, se = 0.41, p = .214) or direct effect (direct effect = .49, se = 0.48, p = .307) of app type on purchase intention. There was also no significant indirect effect of app type on purchase intention via perceived usefulness (indirect effect = .00, boot SE = 0.10, BCBCI = [-.28, .16]) or via perceived ease of use (indirect effect = .02, boot SE = 0.27, BCBCI = [-.53, .54]). The results from PROCESS macro model 4 indicated that neither perceived usefulness (b = 0.38, se = 0.29, p = .197) nor perceived ease of use (b = 0.02, se = 0.21, p = .940) was a significant predictor of purchase intention. These results violated the assumptions of Hypothesis H3a and H3b, therefore both H3a and H3b were rejected.

Serial Mediation. To test the effects of app type on purchase intention mediated by perceived usefulness and brand attitude (H4a), we conducted a serial multiple mediation analysis (PROCESS macro model 6) with perceived usefulness as the first mediator and brand attitude as the second mediator. App use, app familiarity, shopping method were included as covariates.

No significant indirect effects were found for all comparisons. The hypothesized serial mediation effect of app type on purchase intention via perceived usefulness and brand attitude was not significant, indirect effect = .00, boot SE = 0.04, BCBCI = [-.09, .10]). Our study results indicated that perceived usefulness is a significant positive predictor of brand attitude, b = 0.37, se = 0.10, p = .001; but brand attitude is not a significant predictor of purchase intention, b = 0.50, se = 0.26, p = .062. Therefore, we failed to find support for H4a.

Hypothesis 4b proposed a serial mediation effect of app type on purchase intention, mediated by perceived ease of use and brand attitude. To test H4b, we conducted a serial multiple mediation analysis with perceived ease of use as the first mediator and brand attitude as the second mediator. Though no significant total or direct effect of app type on purchase intention was found, a significant indirect effect of app type on purchase intention via perceived ease of use and brand attitude was found, indirect effect = .16, boot SE = 0.09, BCBCI = [.02, .36]. Results from PROCESS macro model 6 indicated that app type was a significant predictor of perceived ease of use, b = 1.20, se = 0.26, p < .001; perceived ease of use is also a significant positive predictor of brand attitude, b = 0.24, se = 0.08, p = .002; and brand attitude is a significant positive predictor of purchase intention, b = 0.56, se = 0.26, p = .033. However, this significant indirect effect of app type on purchase intention via

perceived ease of use and brand attitude was in the opposite direction as we expected, therefore H4b was rejected.

Conclusion and Discussion

This study investigated the effects of AR usage (present or absent) in mobile apps on one’s perceptions of the usefulness and ease of use of that app based on the framework of the TAM (Davis, 1989), and whether these perceptions mediate the effects of AR on brand attitude and purchase intention. Our findings are as follows: 1) Compared to the non-AR shopping app, the AR shopping app led to lower perceived ease of use, which consequently resulted in less favorable brand attitudes and lower purchase intention; 2) People did not perceive the AR shopping app as more or less useful than the non-AR shopping app.

The first purpose of our study was to examine whether there was an effect of AR usage in mobile apps on perceived usefulness and perceived ease of use. Interestingly, we did find an effect of AR on perceived ease of use, but it was contrary to our expectations: the non-AR app is perceived as easier to use than the AR app. Such results add confidence to the findings of a study by Mahony (2015), which indicated that individuals were unlikely to intuitively understand their own role in the AR interaction and suggested that sufficiently explicit guidance played a vital role in ensuring simplicity in AR engagement. One possible explanation is that the navigation of the AR app (IKEA Place) is more complex than the non-AR app (IKEA Store). Several participants mentioned that the non-AR app did not provide them with sufficient instruction throughout the app interaction. We’ve noticed that some

participants had no clue what they should do once they landed on the main interface as it only offered a camera view without explaining what that view was for. Also, some people pointed out that the way of adjusting the virtual furniture was different from what they were

expecting. People who were used to using two fingers for zooming and one finger for moving were confused when they found it was the other way around in the IKEA Place app, which further increased the complexity of engaging with the AR function.

Our second but most important goal was to further explore whether perceived ease of use and perceived usefulness played a mediating role in the relationship between AR and brand attitude, AR and purchase intention respectively. Despite that our hypothesis and results went towards different direction, our experiment did partially provide evidence for the TAM (Davis, 1989), because we found an indirect negative effect of AR usage in mobile shopping apps on purchase intention, via perceived ease of use and brand attitude; but we found no effect of perceived usefulness on brand attitude or purchase intention. Our study results demonstrated that people in the non-AR condition perceived the app as easier to use, so they liked the brand more and showed higher purchase intention for the product than people in the AR condition. As we found no direct effect of AR on purchase intention, we can conclude that an app increases people’s purchase intention only when they perceive the app as easier to use and have a more positive attitude towards that brand.

It is worth noting that our study suggested that perceived usefulness was neither positively nor negatively influenced by the use of AR technology, which is contrary to previous findings which pointed out that AR technology leads to higher perceived usefulness (Chen & Tan, 2004; Huang & Liu, 2014; Mahony, 2015; Rese, Baier, Geyer-Schulz, & Schreiber, 2017) than its non-AR counterparts. A possible explanation might be that the participants were asked to imagine that they were in need of a new chair, but they actually did not have such needs, or at least we cannot confirm to what extent were they convinced by such instruction. Perceived usefulness is, however, a relative concept which depends on the value of the information they received and how well does such information tap into their own needs. If participants were not high involved in the experiment, they would lack the motive to centrally process the usefulness of this app, therefore resulting in similar perceived usefulness among two experimental groups.

Practical Implications

The result of this research indicated that AR shopping app is not necessarily a more effective marketing tool than traditional shopping app in terms of increasing people’s brand perceptions and product interests, which may even backfire when designed improperly. Even though AR technology was theoretically believed to be user-appealing (Chung, Han, & Joun, 2015; Kim & Forsythe, 2008; Mayer & Moreno, 2003), it requires adequate instruction and well-thought design to take advantage of those unique attributes of AR. The most important takeaway from our study is that perceived ease of use plays a key and fundamental role in fluent user experience and app success. Without sufficient instruction and proper navigation, people would be overwhelmed by the various features of AR technology and perceived it as harder to use than its non-AR counterparts.

It should be flagged with the practitioners in the marketing industry that AR technology is not a guarantee for marketing success. They should carefully evaluate the advantages and potential threats of adopting AR technology in their campaign and spend their money wisely. As our study results indicated, if people find the app easy to use, they would like the brand more and want to buy the advertised products more. Therefore, the general suggestion is that perceived ease of use should be a priority of app designing, and the app designers should make an effort to make the app more intuitive and user-friendly. If

marketeers are with a limited budget and cannot guarantee simplicity of engagement or offer adequate customer support within that budget, it might be a good idea to keep things simple and opt for a more traditional but executable marketing strategy.

Limitations and Future Research

There are two limitations of our study: First, our experiment did not provide an ideal environment for testing the AR function. The main use case of IKEA Place is for people to

put virtual furniture in a home setting, but our experiment took place in some public areas which are very different from a personal space. Therefore, the validity of our study was harmed. To improve the validity of the experiment, it is preferable that a home-like lab-setting is adopted in future research to make participants feel more at home and take advantage of the AR function to a greater extent.

Second, even though we tried to limit the differences among two groups to the presence or absence of AR function only, there are some features (check Appendix A for complete comparison) in two apps that are essentially different and cannot be ruled out, which poses a threat to the internal validity of our experiment. The most important of all is that the navigation of the IKEA Store app is clearer and more straightforward than the IKEA Place app. This would lower the perceived usefulness in the AR app condition and shortens the differences between the two experimental groups. Besides, the IKEA Store app enables users to buy products directly within the app, whereas the IKEA Place app is for browsing only and no purchase function is provided, therefore the IKEA Store app is more useful in terms of making actual purchase.

References

Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, N.J.: Prentice-Hall, 1980.

Alba, J., Lynch, J., Weitz, B., Janiszewski, C., Lutz, R., Sawyer, A., & Wood, S. (1997). Interactive home shopping: Consumer, retailer, and manufacturer incentives to participate in electronic marketplaces. Journal of Marketing, 61(3), 38.

doi:10.2307/1251788

Bagozzi, R. P. (1981). Attitudes, intentions, and behavior: A test of some key hypotheses. Journal of Personality and Social Psychology, 41(4), 607–627. doi:10.1037/0022-3514.41.4.607

Bandura, A. (1982). Self-efficacy mechanism in human agency. American Psychologist, 37(2), 122–147. doi:10.1037/0003-066X.37.2.122

Billinghurst, M., & Kato, H. (2002). Collaborative augmented reality. Communications of the ACM, 45(7). doi:10.1145/514236.514265

Chen, L. D., & Tan, J. (2004). Technology adaptation in e-commerce: Key determinants of virtual stores acceptance. European Management Journal, 22(1), 74–86.

doi:10.1016/j.emj.2003.11.014

Childers, T. L., Carr, C. L., Peck, J., & Carson, S. (2001). Hedonic and utilitarian motivations for online retail shopping behavior. Journal of Retailing, 77(4), 511–535.

doi:10.1016/S0022-4359(01)00056-2

Cho, C. H., & Cheon, H. J. (2004). Why do people avoid advertising on the internet? Journal of Advertising, 33(4), 89–97. doi:10.1080/00913367.2004.10639175

Chu, K., & Yuan, B. J. C. (2013). The effects of perceived interactivity on e-trust and

Commerce Research, 14(1), 124–136.

Chung, N., Han, H., & Joun, Y. (2015). Tourists’ intention to visit a destination: The role of augmented reality (AR) application for a heritage site. Computers in Human

Behavior, 50, 588–599. doi:10.1016/j.chb.2015.02.068

Daugherty, T., Li, H., & Biocca, F. (2008). Consumer learning and the effects of virtual experience relative to indirect and direct product experience. Psychology and Marketing, 25(7), 568–586. doi:10.1002/mar.20225

Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User acceptance of computer technology: A comparison of two theoretical models. Management Science, 35(8), 982–1003. doi:10.1287/mnsc.35.8.982.

Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1992). Extrinsic and intrinsic motivation to use computers in the workplace. Journal of Applied Social Psychology, 22(14), 1111–1132.doi:10.1111/j.1559-1816.1992.tb00945.x

Dudovskiy, J. (2019). IKEA segmentation, targeting and positioning: Targeting

cost-conscious customers - research-methodology. Retrieved June 16, 2019, from https://research-methodology.net

Elliott, M. T., & Speck, P. S. (2005). Factors that affect attitude toward a retail web site. Journal of Marketing Theory and Practice, 13(1), 40–51.

doi:10.1080/10696679.2005.11658537

Filieri, R., McLeay, F., Tsui, B., & Lin, Z. (2018). Consumer perceptions of information helpfulness and determinants of purchase intention in online consumer reviews of services. Information and Management, 55(8), 956–970.

doi:10.1016/j.im.2018.04.010

Fiore, A. M., Kim, J., & Lee, H. H. (2005). Effect of image interactivity technology on consumer responses toward the online retailer. Journal of Interactive Marketing,

19(3), 38–53. doi:10.1002/dir.20042

Fishbein, M. (2009). An integrative model for behavioral prediction and its application to health promotion (2nd Edition). In R. J. DiClemente, R. A. Crosby, & M. C. Kegler (Eds.), Emerging Theories in Health Promotion Practice and Research (pp. 215– 234). San Francisco, CA: Jossey-Bass.

Gao, Q., Rau, P. L. P., & Salvendy, G. (2009). Perception of interactivity: Affects of four key variables in mobile advertising. International Journal of Human-Computer

Interaction, 25(6), 479–505. doi:10.1080/10447310902963936

Ha, S., & Stoel, L. (2009). Consumer e-shopping acceptance: Antecedents in a technology acceptance model. Journal of Business Research, 62(5), 565–571. doi:10.1016/j.jbusres.2008.06.016

Hauser, J. R., & Shugan, S. M. (2008). Intensity measures of consumer preference. Operations Research, 28(2), 278–320. doi:10.1287/opre.28.2.278

Heijden, H. V. D., Verhagen, T., & Creemers, M. (2003). Understanding online purchase intentions: Contributions from technology and trust perspectives. European Journal of Information Systems, 12(1), 41–48. doi:10.1057/palgrave.ejis.3000445

Hernandez, B., Jimenez, J., & Martín, M. J. (2009). Adoption vs acceptance of e-commerce: Two different decisions. European Journal of Marketing, 43(9), 1232–1245. doi:10.1108/03090560910976465

Hsieh, J. K., Hsieh, Y. C., & Tang, Y. C. (2012). Exploring the disseminating behaviors of eWOM marketing: Persuasion in online video. Electronic Commerce Research, 12(2), 201–224. doi:10.1007/s10660-012-9091-y

introduction to theory and research. Contemporary Sociology, 6(2), 244. doi:10.2307/2065853

Huang, T. L., & Liu, F. H. (2014). Formation of augmented-reality interactive technology’s persuasive effects from the perspective of experiential value. Internet Research, 24(1), 82–109. doi:10.1108/IntR-07-2012-0133

Javornik, A. (2016). Augmented reality: Research agenda for studying the impact of its media characteristics on consumer behaviour. Journal of Retailing and Consumer Services, 30, 252–261. doi:10.1016/j.jretconser.2016.02.004

Jin, S. A. A. (2009). The roles of modality richness and involvement in shopping behavior in 3D virtual stores. Journal of Interactive Marketing, 23(3), 234–246. doi:10.1016/j.intmar.2009.04.005

Jin, S.-A. A., & Bolebruch, J. (2009). Avatar-based advertising in second life: The role of presence and attractiveness of virtual spokespersons. Journal of Interactive Advertising, 10(1), 51–60. Retrieved June 16, 2019, from https://www.ebscohost.com/ Kim, M., & Lennon, S. (2008). The effects of visual and verbal information on attitudes and

purchase intentions in internet shopping. Psychology and Marketing, 25(2), 146–178. doi:10.1002/mar.20204

Kim, J., & Forsythe, S. (2008). Adoption of virtual try-on technology for online apparel shopping.Journal of Interactive Marketing, 22(2), 45–59. doi:10.1002/dir.20113 Kim, J. U., Kim, W. J., & Park, S. C. (2010). Consumer perceptions on web advertisements

and motivation factors to purchase in the online shopping. Computers in Human Behavior, 26(5), 1208–1222. doi:10.1016/j.chb.2010.03.032

critical review of the technology acceptance model. Information and Management, 40(3), 191–204. doi:10.1016/S0378-7206(01)00143-4

Lepper, M. R. (1985). Microcomputers in education: Motivational and social issues. American Psychologist, 40(1), 1–18. doi:10.1037/0003-066X.40.1.1

Li, T., & Meshkova, Z. (2013). Examining the impact of rich media on consumer willingness to pay in online stores. Electronic Commerce Research and Applications, 12(6), 449– 461. doi:10.1016/j.elerap.2013.07.001

Macinnis, D. J., & Price, L. L. (1987). The role of imagery in information processing: review and extensions. Journal of Consumer Research,13(4), 473. doi:10.1086/209082 Mahony, S. O. (2015). A proposed model for the approach to augmented reality deployment

in marketing communications. Procedia - Social and Behavioral Sciences, 175, 227– 235. doi:10.1016/j.sbspro.2015.01.1195

Manis, K. T., & Choi, D. (2018). The virtual reality hardware acceptance model (VR-HAM): Extending and individuating the technology acceptance model (TAM) for virtual reality hardware. Journal of Business Research. Elsevier Inc. doi:10.1016/j.jbusres.2018.10.021

Mayer, R. E., & Moreno, R. (2003). Nine ways to reduce cognitive load in multimedia learning. Educational Psychologist, 38(1), 43–52. doi:10.1207/S15326985EP3801_6 McKinsey. (2015, October). The new consumer decision journey. Retrieved April 8, 2019,

from https://www.mckinsey.com

simulated products on purchase intention. Journal of Retailing and Consumer Services, 21(5), 877–887. doi:10.1016/j.jretconser.2014.02.010

Park, E. J., Kim, E. Y., Funches, V. M., & Foxx, W. (2012). Apparel product attributes, web browsing, and e-impulse buying on shopping websites. Journal of Business Research, 65(11), 1583–1589. doi:10.1016/j.jbusres.2011.02.043

Park, H. H., Jeon, J. O., & Sullivan, P. (2015). How does visual merchandising in fashion retail stores affect consumers’ brand attitude and purchase intention? International Review of Retail, Distribution and Consumer Research, 25(1), 87–104. doi:10.1080/09593969.2014.918048

Parra, J. F., & Ruiz, S. (2009). Consideration sets in online shopping environments: the effects of search tool and information load. Electronic Commerce Research and Applications, 8(5), 252–262. doi:10.1016/j.elerap.2009.04.005

Radner, R., & Rothschild, M. (1975). On the allocation of effort. Journal of Economic Theory, 10(3), 358–376. doi:10.1016/0022-0531(75)90006-X

Raska, K., & Richter, T. (2017). Influence of augmented reality on purchase intention: The IKEA case. Jönköping University. Retrieved April 9, 2019, from http://www.diva-portal.org/smash/get/diva2:1115470/FULLTEXT01.pdf

Rese, A., Baier, D., Geyer-Schulz, A., & Schreiber, S. (2017). How augmented reality apps are accepted by consumers: A comparative analysis using scales and opinions. Technological Forecasting and Social Change, 124, 306–319. doi:10.1016/j.techfore.2016.10.010

Integrated Approach to Communication Theory and Research (pp. 415–434). Routledge. doi:10.4324/9780203710753-35

Saltzer, E. B. (1981). Cognitive moderators of the relationship between behavioral intentions and behavior. Journal of Personality and Social Psychology, 41(2), 260–271. doi:10.1037/0022-3514.41.2.260

Schwartz, A. M. (2011). Augmenting purchase intent: An empirical study on the effects of utilizing augmented reality in online shopping. SSRN. doi:10.2139/ssrn.1858976 Song, J. H., & Zinkhan, G. M. (2008). Determinants of perceived web site Interactivity.

Journal of Marketing, 72(2), 99–113. doi:10.1509/jmkg.72.2.99

Stoyanova, J., Brito, P. Q., Georgieva, P., & Milanova, M. (2015). Comparison of consumer purchase intention between interactive and augmented reality shopping platforms through statistical analyses. In INISTA 2015 - 2015 International Symposium on Innovations in Intelligent SysTems and Applications, Proceedings. Institute of Electrical and Electronics Engineers Inc. doi:10.1109/INISTA.2015.7276727

Sundar, S. S., Bellur, S., Oh, J., Jia, H., & Kim, H. S. (2016). Theoretical importance of contingency in human-computer interaction: Effects of message interactivity on user engagement. Communication Research, 43(5), 595–625. doi:10.1177/0093650214534962

van Noort, G., Voorveld, H. A. M., & van Reijmersdal, E. A. (2012). Interactivity in brand web sites: Cognitive, affective, and behavioral responses explained by consumers’ online flow experience. Journal of Interactive Marketing, 26(4), 223–234. doi:10.1016/j.intmar.2011.11.002

Toward a unified view. MIS Quarterly, 27(3), 425. doi:10.2307/30036540 Warshaw, P. R. (2006). Predicting purchase and other behaviors from general and

contextually specific intentions. Journal of Marketing Research, 17(1), 26. doi:10.2307/3151113

Warshaw, P. R., & Davis, F. D. (2007). Self-understanding and the accuracy of behavioral expectations. Personality and Social Psychology Bulletin, 10(1), 111–118.

doi:10.1177/0146167284101013

Wu, P. C. S., Yeh, G. Y. Y., & Hsiao, C. R. (2011). The effect of store image and service quality on brand image and purchase intention for private label brands. Australasian Marketing Journal, 19(1), 30–39. doi:10.1016/j.ausmj.2010.11.001

Yim, M. Y. C., Sauer, P. L., Williams, J., Lee, S. J., & Macrury, I. (2014). Drivers of attitudes toward luxury brands: A Cross-National Investigation into the roles of Interpersonal Influence and Brand Consciousness. International Marketing Review, 31(4), 363–389. doi:10.1108/IMR-04-2011-0121

Zhang, S., Zhao, J., & Tan, W. (2008). Extending TAM for online learning systems: An intrinsic motivation perspective. Tsinghua Science and Technology, 13(3), 312–317. doi:10.1016/S1007-0214(08)70050-6

Appendix A

Comparison between the two apps

Function IKEA Store (non-AR app) IKEA Place (AR app)

Purchase option: “add to bag”

√ ×

In stock information √ ×

Specific location in store √ ×

link to the website page √ ×

Assembled size √ √

Package measurement √ √

Care instruction √ √

Assembly instruction √ √

Designer information √ √

Environment & Materials √ √

AR option: “try it in your place”

× √

Recommendation: “you may also like”

× √

Element of humanized voices