Effects of visibility of packaging brand elements on consumer attitude, quality perception and purchasing behavior in mobile online shopping context

(1)

Effects of Visibility of Packaging Brand

Elements on Consumer Attitude, Quality

Perception and Purchasing Behavior in

Mobile Online Shopping Context

Jiyoung Min

Student number: 11137630

MSc. in Business Administration – Marketing track

University of Amsterdam – Faculty of Economics & Business Master Thesis final version

First supervisor: Antoon Meulemans Second supervisor: Anouar El Haji January 27, 2017

(2)

Statement of originality

This document is written by Jiyoung Min who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Abstract

Does level of visibility of packaging brand elements on mobile phone screen affect consumers’ attitude, perception of quality and purchasing behavior? Using an auction research platform against Dutch consumers with ground coffee products, this article examines the impact of packaging appearance on mobile screen on consumer tendencies and behaviors. The results reveal that there is a positive causal effect of mobile screen visibility of packaging brand elements on consumers’ attitude and quality perception towards the product but not on purchasing behavior. Furthermore, the moderating effect of brand familiarity could not be found for the visibility effect on any of the attitude, quality perception and behavior variables. However, it was found that, for the attitude towards product, the visibility effect is significant only for unfamiliar brand condition but not on familiar brand condition. This research sheds light on how appearance of branded products on mobile screen could affect consumer tendencies and behaviors, and what marketers should take into consideration when developing packaging design for products to be presented and/or sold online.

(4)

I. Introduction

In the last few years, consumers are increasingly being engaged with web shops. According to an annual survey in the U.S. of over 5,000 online shoppers by United Parcel Service Inc.(“UPS Study,” 2016), the year 2016 was the first year that consumers bought more of their purchases on the web than in stores, making 51% of their purchases compared to 48% in 2015 and 47% in 2014. In the Dutch market, online retail industry has increased by 12.1% since 2015, reaching over 18 billion euros (Ecommerce News, 2016). This shopping trend was even stronger on mobile devices, jumping from 41% to 44% since last year. The increasing level of adoption of smartphones, usage and related services is boosting mobile e-commerce, and retailers are investing more in their digital operations (Ruddick, 2015).

In this unarguable trend of internet that is reshaping retail market (Klein, 1998), many brands and retailers are increasingly facing the challenges that come from the differences in shopping contexts between mobile online shopping environment and traditional bricks-and-mortar stores. Due to the fundamental difference of encountering with products on mobile screen with limited visual space, consumers’ ability to be able to interact with various branding cues presented on product packaging is naturally reduced compared to the shopping contexts of physical stores. It is especially important to consider because nowadays consumers are increasingly engaging with mobile phones in their decision-making process, and not only in eventual purchasing phase but also in online search phase in which they form initial attitudes and evaluate and compare product qualities. Therefore, it became imperative for brands to consider their appearance not only on shelf, but also on mobile screen.

Despite the considerable impact that mobile online shopping holds on how people shop and interact with brands nowadays, very little has been learned about the differences in consumer tendencies and behaviors in online landscape, and more specifically, how different presentations of products and brands influence consumer decision making, especially on mobile phones (Kim & Lennon, 2008).

There are many studies done about the general impact of packaging or product visual elements on consumers’ attitude, evaluations, judgements and behaviors. Visually appealing packages have a positive effect on consumers’ attention level (Underwood, Klein, & Burke, 2001), valuation (Ghoshal, Boatwright, & Cagan, 2009; Underwood et al., 2001), quality perception (S.T. Wang, 2013), evaluation, persuasion and behavior (Negm & Tantawi, 2015; Yeon Kim & Chung, 2011). About brand elements on pack in specific, previous studies found that the difference in visual cues on product packaging and

communications have varying effects on shoppers’ valuations and benefit associations (Ampuero & Vila, 2006; Sehrawet & Kundu, 2007), perception of product quality (Ampuero & Vila, 2006; Chowdhury &

(6)

2 Andaleeb, 2007; Insch & Florek, 2009), and even associations with specific positioning strategies

(Ampuero & Vila, 2006).

This paper examines how packaging appearance of branded products on smartphone screen affects consumers’ perceptions, attitudes and behaviors. In particular, the attention is given on ‘visibility’ of brand elements as we focus on one of the most prominent differences between offline vs mobile online situations – visual limitations on mobile setting that come from having to view products on small screen. In this study, visibility refers to clarity and readability of brand attributes on pack, and is defined in terms of the two constructs: 1) The extent of readability of text elements: how easily product information can be read, that is, how product information on packaging is presented in an easily readable style in terms of size, font style, font color, layout, etc., and 2) the extent of readability of non-textual and graphical elements, that is, how design aspects of packaging are displayed in a clear and discernable style in terms of size, lines, shapes, patterns, color, layout, etc.

This paper suggests that level of visibility of brand elements designed on packaging positively affects consumers’ attitude towards product, perception of product quality and purchase behavior, the degrees of which would differ depending on level of familiarity with the brand. The research question is formulated as the following:

How does visibility of brand elements on packaging affect consumers’ attitude towards product,

perception of product quality and purchasing behavior in mobile online shopping context, and how is the relationship moderated by brand familiarity?

This study makes several important contributions. First, considering that not much study has been done on performance of brand elements on packaging and on online setting in mobile online shopping context, it sheds light on consumer tendencies and behaviors in this new shopping environment that is

differentiated with offline settings. Secondly, by measuring the respondents’ willingness to pay for the products in the experiment, the current study measures the actual behaviors than hypothetical intentions, therefore provide more realistic and valid results. Lastly, this study provides relevant insights for the brands and retailers that solely or at least partly sell their packaged goods through mobile e-commerce about how brand performance can be limited in mobile shopping situations, and thus, what they should take into account when designing product packages to be showcased on mobile screen. It is particularly relevant in the context of new product developments or introduction of foreign brands in domestic markets, as the study examines the differential impacts depending on familiarity with brand.

The following chapter explores some of the studies and theories, along which the hypotheses of this study are introduced. Subsequently, the methodology of this research will be explained, which is followed by

(7)

3 the results and the discussion. The paper ends with the key contribution of this study made to both

academic and practical fields, and the limitations as well as the suggestions for future research.

II. Literature review

In terms of information theoretical view, products and brands is comprised of a range of cues, such as brand name, logo, packaging, color, price, etc., each of which takes part in forming diverse impressions that will subsequently contribute to shaping certain evaluations and behaviors (Jacoby, Olson, & Haddock, 1971). Packaging, among those extrinsic cues that are external to physical product itself, performs a crucial role in key communications and salesmanship at physical point of purchase, in which consumers spend less than 12 seconds on average to make decisions (Dickson & Sawyer, 1990). It is found that majority of shoppers counts on packaging to make their buying decisions at the point of purchase (Wells, Farley, & Armstrong, 2007).

1. General impact of packaging

Packaging is a fundamental part of promotion that is more than physical container. Packages that are visually appealing can not only convey a substantial promotional value but also communicate important information, such as usage, features, benefits, etc. (Wells et al., 2007). Because of its substantial impact on how people feel, judge and behave about products and brands, packaging is sometimes even called the ‘‘fifth P’’ (Keller & Lehmann, 2006; Nickels & Marvin, 1976).According to an empirical study,

packaging pictures make consumers more focus on the brand as extrinsic cues of packages help shoppers infer intrinsic product attributes (Underwood et al., 2001). Thus, packaging can be a crucial tool for brands to communicate subliminally with their customers (Thalhammer, 2007).

Positively perceived packaging can have a direct impact on consumers’ perception, evaluation, persuasion and ultimately behavior (Negm & Tantawi, 2015; Yeon Kim & Chung, 2011). Studies reveal that more appealing packages have a positive impact on how consumers make valuation on products and form attitudes (Ghoshal et al., 2009) and the attitudes formed as such can sway how they perceive product quality and favor a brand over another (S.T. Wang, 2013). In an aesthetic perspective, well executed visual design can bring affective responses (Mzoughi & Abdelhak, 2011), which subsequently influences consumers to react more favorably to the promotional material as well as the brand (Mick & Buhl, 1992).

(8)

4

2. Impact of packaging brand elements

Such effects are driven by various brand elements presented on pack, such as logo, colors, claims, icons, graphical forms and images. Visual cues are categorized into visual imagery and text based imagery (Köksal, 2013), which, in harmony with one another, form different perceptions, evaluations, intentions and behaviors. Especially, this impact is found to be the most visible in low involvement situations (Silayoi & Speece, 2004).

Consumers perceive different values and benefits from the visual cues on pack depending on how colors, typography, graphical forms and images are presented (Ampuero & Vila, 2006; Sehrawet & Kundu, 2007). An array of cues that consist of a product package can perform as indicators of product quality (Ampuero & Vila, 2006; Insch & Florek, 2009). Product attributes, brand image, country-of-origin information could form different levels of perceived quality (Chowdhury & Andaleeb, 2007). Beyond mere valence or level of thoughts, varying dimensions of design variables can have impact in more complex levels, such as being associated with different positioning strategies in consumers’ minds (Ampuero & Vila, 2006).

If such brand elements on pack can influence consumers’ thoughts and actions to certain ways, they should have differential effects depending on how they are presented. This implies that presenting brand elements on pack in a clear manner is crucial to achieve an intended impact on consumers.

3. Consumer tendencies and behaviors in mobile e-commerce context

In an information processing perspective, how a branded product is viewed and perceived depends on three fundamental factors: brand elements as a message sender, individual factors as a receiver (individual ability, motivation, mood, etc.) and contextual factors that enable smooth, or at all, flow of messages. There are many studies focusing on the relation between packaging and shoppers under the assumption that the conditional factors in the shopping environment are adequate, such as contextual conditions that allow to see the shopping items well enough not to impose difficulties in the shopping experience, or revealing the importance and impact of packaging presented on physical brick-and-mortar store context (Prendergast & Pitt, 1996; Rettie & Brewer, 2000). However, there has been hardly any research conducted about how such impact can be elicited differently depending on contexts, particularly, in a mobile e-commerce context.

Although the number of studies about consumer behavior in online shopping context are increasing, especially in the field of mobile e-commerce, the subject area remains largely uninvestigated due to the

(9)

5 fact that it has such a short history and has been developing in an incredibly fast speed that its substantial impact has been observed only in relatively recent years (Lu & Yu‐Jen Su, 2009).

Limitations of mobile environment on shoppers

Mobile online setting entails specific different contextual settings that are different from physical store environment for consumers to perceive, evaluate and make decisions on products and brands. Due to certain features that are inherent in smartphone devices, such as small screen size, visibility of products and packaging can be limited and thus, the extent to which packaging can have impact on consumers can be restricted in mobile contexts. For consumers, it means that they have to bear higher interaction costs to process the equal amount of information as from a desktop or a physical setting (Budiu, 2015). Thus, such limitations could hinder shoppers’ natural ability to freely interact, clearly judge and instantly form certain feelings about the product or brand.

According to a study, people’s ability to understand textual contents on mobile devices was significantly lower than on desktop computers (Singh, Sumeeth, & Miller, 2011). People had difficulties in processing information on smartphones in general, having their comprehension ability reduced to approximately half the level when using the iPhone sized screens compared to the desktop screens. This research provides several explanations for what could hinder such capability: first, people have confined visual space, therefore forced to view limited things. This naturally reduce their level of understanding the full merits of the visual material, and in such situations, people tend to rely on other sources of information to aid their understanding, such as their memories. Second, people have to move around the page by scrolling and zooming in and out, which takes time, and therefore, is cognitively tiring and deters attention to the visual material in front (Singh et al., 2011).

Factors in e-commerce settings that could affect shoppers

Researchers have studied different factors in (mobile) e-commerce settings that influence consumers in general, focusing on specifications of websites, shoppers’ motivations, or tendencies in their shopping behaviors. Some of the important factors in mobile shopping experiences that affect consumers’

satisfaction and behaviors include quality and accessibility of information, website quality, seamless user interface and security perceptions (Bai, Law, & Wen, 2008; C. Park & Kim, 2003). According to a study that focused on service quality of mobile online services, positively perceived service quality affects both perceived value and customer satisfaction, which subsequently affects purchase intention. Additionally, content quality appeared to be one of the most influential factors on such attitudinal and behavioral tendencies (Kuo, Wu, & Deng, 2009).

(10)

6

4. Current study

The above studies, although not directly related to visibility of products or brands, imply that being able to process clear information about product benefits or features and seamless experiences coming from unhindered visual conditions on mobile screen could significantly contribute to shoppers’ positive evaluations and behaviors (C. Park & Kim, 2003). When consumers cannot properly see the visual components as intended and therefore, are not sufficiently engaged with them, it is inferable that it would reduce desired level of consumers’ positive thoughts, evaluations and/or behaviors.

Impact of visibility of packaging brand elements on consumers

In the Elaboration Likelihood Model (Petty & Cacioppo, 1986), it is discussed that one of the most influential factors that determine whether an individual would follow more or less elaborate manner to process information is his ability to sufficiently process a message. When a person’s cognitive ability is confined due to certain external factors, especially in a situation where the level of involvement is low (low motivation), he is more likely to follow peripheral route of processing information. That is, an individual would more rely on non-central matters, such as general impressions or positive and negative cues, rather than true merits of the subject matter.

When visual elements of the promoted product cannot be viewed clearly, consumers might be more dependent on how they generally and instantly see and feel about the product. People are called to be “cognitive misers”, having a tendency to reduce mental effort (Morris, Woo, & Singh, 2005), especially when they have to make low-involvement decisions about a product. This tendency would lead them to process brand messages rather in peripheral routes, thus not by thoroughly examining pertinent visual aspects given on pack. In that sense, unclear visual cues are unlikely to generate positive attitudes, evaluations or purchase decisions.

In this current paper, it is suggested that level of visibility of brand elements on pack would be positively related to consumers’ perceptions, evaluations and behavior, and more specifically, higher visibility would have causal effects on those outcome variables. The hypotheses are formulated as the following:

H1: The higher the level of visibility of packaging brand elements on mobile screen is, the more positive consumers’ attitude would be.

H2: The higher the level of visibility of packaging brand elements on mobile screen is, the higher product quality consumers would perceive.

H3: The higher the level of visibility of packaging brand elements on mobile screen is, the more consumers are likely to purchase the product.

(11)

7

Differential effect for familiar and unfamiliar brand

Following the theory of Elaboration Likelihood model, when people take peripheral route of thinking in low-involvement situations, they tend to take mental shortcuts to process information, such as heuristics (Petty & Cacioppo, 1986). One of the shortcuts is relying on source credibility or previously gained knowledge or experience. Relying on such shortcuts is a less effortful and fairly reliable way to make quick judgements, and the decisions made as such are generally unrelated to objective merits of the subject (Petty & Cacioppo, 1986).

The impact of brand reputation or familiarity has been dealt with in many studies. A significant role of brand familiarity, prior experience and pre-purchase information was found on perceived risk (Beaver, Lambert, & Morse, 1980; Ha, 2002; J. Park & Stoel, 2005), brand attitude (Machleit & Wilson, 1988) and purchase intention (J. Park & Stoel, 2005). Similarly, brand familiarity boosts consumers to gain

confidence about the brand, which subsequently encourages them to have stronger intention to purchase the product (Laroche, Kim, & Zhou, 1996). These results are in line with the findings in neuromarketing in which it is explained that perception of stimuli can be affected by how familiar a person is with it (Bülthoff & Newell, 2006). Furthermore, brand reputation can play a powerful role in indicating shoppers about product quality (Allison & Uhl, 1964; Marquardt, Makens, & Larzelere, 1965; Selnes, 1993; Valenzi & Eldridge, 1973).

In a study conducted using virtual reality simulation, it was found that, only for unfamiliar brands, packaging imagery has an effect to enhance consumers’ attention to the brand (Underwood et al., 2001). The results suggest that consumers pay less attention in general to packaging communications when they already know the brand, while they could perceive packaging elements more sensitively when the brands are unfamiliar.

In a situation where level of involvement is low and ability to process is interrupted, if the brand is familiar and fairly reliable, consumers could update their prior knowledge and experiences they have gathered (Snyder & Stukas, 1999) and the impact of stimuli could be weakened (Britton & Tesser, 1982). As they would instantly engage in routine response when they recognize that the brand is familiar, the level of visibility of brand elements that contain more elaborate information and messages about the product would not matter as much as it would when they are exposed to unfamiliar brands.

Therefore, based on the reasoning above, this study posits the following hypotheses:

H4a: The impact of level of visibility of brand elements on attitude would be significantly stronger for unfamiliar brand than familiar brand.

(12)

8

H4b: The impact of level of visibility of brand elements on perception of quality towards the product would be significantly stronger for unfamiliar brand than familiar brand.

H4c: The impact of level of visibility of brand elements on purchasing behavior would be significantly stronger for unfamiliar brand than familiar brand.

The conceptual model is presented below: Figure 1. Conceptual model

III. Methodology

1. Data collection

For collecting data, Veylinx software was used as a primary data collection tool. Veylinx is a marketing research firm based in Amsterdam and its online auction platform enables to collect data from its own panel in the Netherlands. With the Veylinx research platform, it is possible to collect behavioral data as it involves actual purchase of the auction items, as well as other data through the survey section which appears following the auction. The survey questions are attached in the appendix 1.

After creating the visual stimuli, the auction was set up in the software in which each of the six stimuli was randomly presented to each of the six equally distributed groups of panels. After indicating the price that they are willing to pay for the auctioned item shown in the stimulus, they were asked to answer several questions to measure their attitude and quality perception towards the product they just bid on.

(13)

9 The survey created in English was subsequently translated in Dutch before running the auctions to the Dutch panel.

2. Sampling

The sampling frame is all the Dutch people above 18 years old, and the sample is the panel of Veylinx which consists of roughly 8945 people. Among this panel who received the auction invitations, 884 respondents have participated in the auction, making the response rate to 9.9%.

3. Experiment design

The experiment was conducted in a between-subjects 2x3 factorial design, with two levels of familiarity and three levels of visibility. The level of familiarity was manipulated by pre-determining each stimulus for familiar and unfamiliar branded product in the pretest phase. Familiar and unfamiliar branded products were subsequently manipulated to differentiate their visibility by enlarging or minifying the brand elements on pack.

In the pretest phase, firstly, three well-known brands sold in the biggest supermarket chain were selected for familiar brands and three relatively unknown brands sold online were selected for unfamiliar brand. Afterwards, 43 university graduate students were asked to indicate the level of familiarity with the brands in the scale of 1 o 10 for both ranges of familiar brands and unfamiliar brands. The brands that scored the highest and the lowest in the level of familiarity were chosen as the experiment items to be used for familiar and unfamiliar brand conditions respectively.

4. Product item selection

The study was conducted based on ground coffee products. This product category was chosen because it is 1) a staple product item that people consume often, thus appropriate product to induce enough purchase in order to measure it, and 2) a balanced product in terms of purchase motivation that it is both functional or emotional, that is, people buy it for functional reasons for daily consumption purpose but also for hedonic reasons to try new products and to engage in emotions during consumption. For the familiar brand Douwe Egberts was chosen, and for the unfamiliar brand Cafédirect Machu Pichu brand was chosen. The stimuli used in the experiments are shown in the appendix 2.

5. Description of the measurements

For the two predictor variables of attitude and quality perception, they were measured in a 5-point Likert scale, in which 1 indicated ‘strongly agree’ and 5 indicated ‘strongly disagree’ with the given statement.

(14)

10 For measuring attitude toward product, two constructs were asked in terms of positivity and attractiveness. They were modified from the measure developed by Chattopadhyay and Basu (1990), which has been adopted a lot to measure consumer attitude in various studies. For measuring quality perception, the level of quality the respondents perceive was asked in a direct manner.

For the third predictor variable, purchasing behavior, it was measured by the respondents’ willingness to pay for the product on Veylinx auction platform. Measuring willingness to pay to predict buying behavior is based on the Vickery auction, in which bids are sealed and the highest bidder wins but pays the second highest bid (Vickrey, 1961). The price that the respondents bid on the product item represents the

maximum price that maximizes their chance to win but not to overpay more than what they are willing to pay, thus reflecting the closest amount that they would actually be paying in real purchase situations. Naturally, the higher the amount they are willing to pay for the product, the stronger indication it is about their intention and behavior of buying the product.

IV. Results

1. Reliability test

The attitude variable was measured in two items (“I feel positive about the product”, “the product is attractive to me”) and it has high reliability, with the Cronbach’s Alpha = 0.909. The corrected item-total correlations indicate that both of the items have a good correlation with the total score of the scale (0.834 for both of the items).

Table 1 reports the basic descriptive statistics and correlations of all the independent and dependent variables, in addition to the two control variables of age and gender.

Table 1. Means, Standard Deviations, Correlations

Variables M SD 1 2 3 4 5 6 1. Age 42.611 14.02 − 2. Gender .49 .500 .01 − 3. Visibility 4.169 .765 .104** .127** − 4. Attitude 3.437 .847 .035 .090** .349** (.909) 5. Quality perception 3.745 .784 .058 .140** .417** .709** − 6. Bid amount 4.983 4.770 -.102** -.079** .015 .260** .192** − Note. Statistical significance: *p <.05; **p <.01; ***p <.001

(15)

11

2. Descriptive statistics

Sample characteristics

Among the total invitations sent to 8945 people, 884 people participated in the experiment (response rate = 9.88%). After deleting the missing data, it resulted the sample size of 876 in total. This number of the total respondents were divided into approximately equal sized six groups and distributed to each of six stimuli (the sample size of each treatment is given below in the Table 2). As it will be explained later in the analysis section, the respondents who used other devices than mobile phones have not been filtered out for the analysis because in this experiment significant differences could not be found between mobile users and other device users in the dependent variables of this research. The respondents consist of 51% of male, and they were well diversified in the age groups of over 18 years old (M =42.6, SD =14.02).

Data characteristics

For both familiar (Douwe Egberts) and unfamiliar brand (Cafédirect), the respondents had more positive attitude and quality perception than negative. In general, both the attitude and the quality perception were more positive for the familiar brand than for the unfamiliar brand. However, the respondents generally bid higher for the unfamiliar brand than the familiar brand.

For the familiar brand, the respondents had the most positive attitude and the highest quality perception in the high visibility treatment (attitude: M =3.62, quality perception: M =4), whereas for the unfamiliar brand, the respondents elicited such attitude and perception in the basis treatment (attitude: M =3.42, quality perception: M =3.58). For both familiar and unfamiliar brand, the respondents had the least positive attitude and the lowest quality perception in the low visibility treatment.

For the purchase behavior, the respondents who bid for familiar brand were willing to pay the most in the basis treatment (M =5), while the respondents with unfamiliar brand were willing to pay the most in the high visibility treatment (M =5.28). Table 2 depicts the descriptive statistics per treatment of visibility, divided by familiar and unfamiliar brand conditions.

Table 2. Descriptive Statistics per treatments

N M (attitude) SD (attitude) M (quality perception) SD (quality perception) M (bid amount) SD (bid amount) Familiar brand Basis 162 3.54 0.941 3.96 0.814 5 4.45 Low visibility 138 3.46 1.01 3.84 0.87 4.93 4.42

(16)

12 High visibility 141 3.62 0.83 4 0.71 4.88 4.5 Unfamiliar brand Basis 158 3.42 0.78 3.58 0.7 4.86 5.19 Low visibility 140 3.21 0.78 3.5 0.75 4.97 5.02 High visibility 137 3.36 0.62 3.58 0.7 5.28 5.06

Testing for normality

In order to check whether the data is normally distributed before conducting analysis, Kolmogorov-Smirnov and Shapiro-Wilk tests were conducted for the dependent variables, attitude and quality

perception. Bid amount is excluded from this test as it reflects the respondents’ willingness to pay for the product, thus the values can be naturally arbitrary. The result shows that the variables are not normally distributed (p > 0.05). However, from checking the skewness and kurtosis, for both variables the values were relatively close to 0, and they were normally distributed in terms of kurtosis.

As the sample size is large (n =876) and the values for skewness and kurtosis are within an acceptable range according to empirical criteria, normality was assumed for the further analysis. Table 3 below shows the results of the normality test for both variables of attitude and quality perception.

Table 3. Normality test

Kolmogorov-Smirnov Shapiro-Wilk

Statistic p Statistic P-value Skewness Kurtosis

Attitude 0.178 .000 .920 .000 -.228 .252

Quality perception 0.281 .000 .843 .000 -.488 .727

3. Visibility effects of brand elements on attitude, quality perception and behavior

To test the hypotheses, hierarchical multiple regression was used in order to examine the causal

relationship between level of visibility of brand elements on pack and levels of attitude towards product, perception of quality, and purchasing behavior, after controlling for age and gender.

Knowing from the previous experiments, roughly half of the Veylinx panel use mobile phone to participate in the auctions they receive and the other half use other online devices such as desktop computers or tablets. Although the current study seeks to find out the differential effects in mobile online

(17)

13 setting as opposed to offline environments, and not to online devices, it was checked whether there is a significant difference in the results between mobile phone users and other online device users. For this, it was asked in the survey which mode of device a respondent is using to participate in the auction. The results reveal that there was no significant difference between mobile phone users (n =460) and other device users (n =415) in their attitude (F (1, 873) =.969, p =.325), quality perception (F (1,873) =.710, p =.400) and purchasing behavior (F (1, 873) =1.296, p =.255). Therefore, for the further analysis, all the participants were included without screening out non-mobile phone users.

Attitude

First, the impact of visibility on attitude was examined. The results reveal that the level of attitude was significantly higher as the level of visibility increases (Table 4). In other words, the higher the visibility of brand elements on pack is, the more positive attitude people have towards the product. Thus, the

hypothesis 1 is supported.

In the first step of the hierarchical multiple regression in which two predictors of age and gender are entered, the model was statistically significant (F (2, 873) = 4.095; p < .05), and explained 0.9% of variance in the level of attitude. In the step 2 after the entry of visibility, the total variance explained by the model as a whole was 1.5% (F (3, 872) = 4.469; p < .05). The introduction of visibility explained additional 0.6% variance in the level of attitude, after controlling for age and gender (R² Change = .006; F (1,872) =5.177; p < .05). In the final model, gender and visibility variables were statistically significant, with gender recording a higher Beta value (ß =.090, p < .01) than visibility (ß =.076, p < .05). Gender was coded as 0 =Male, 1 =Female. It could be interpreted that when visibility level of brand elements

increases, the level of positivity in attitude towards product increases slightly, and women elicited more positive attitude than men.

Table 4. Hierarchical Regression Model of Attitude

R R² R² Change B SE ß t Step 1 .096 .009* Age .002 .002 .034 1.012 Gender .152 .057 .090*** 2.667 Step 2 .123 .015** .006* Age .002 .002 .035 1.035 Gender .152 .057 .090** 2.673

(18)

14

Visibility .081 .036 .076* 2.275

Note. Statistical significance: *p <.05; **p <.01; ***p <.001

Differential effect between levels of visibility:

In order to further examine whether the difference of the effect of visibiility is significant between all levels of visibility, one-way ANOVA was conducted (Table 5). The result shows that the difference of the effects between levels are significant (F (2, 873) =3.235, p <.05; M(low) = 3.33, M(basis) = 3.48, M(high) = 3.49) as aligned with the result above, but it is not the case for between pair-wise levels. This can be due to that pair-wise test is less sensitive than the test for the whole model. However, what is noticeable is that the difference in the impact between the low and the high visibility was close to a significant level (p =.060) as well as between the low and the basis (p =.075), but clearly not between the basis and the high visibility (p =.985). This implies that packaging brand elements being less visible than average might have a significant impact on consumers to have less positive attitude, while better visible than average might not.

Table 5. One-way ANOVA results and descriptives of visibility on attitude

SS DF MS F Sig. Visibility 4.620 2 2.310 3.235 .040 Error 623.427 873 .714 Total 628.047 875 Mean SD n Low visibility 3.3309 .91232 278 Basis 3.4813 .86627 320 High visibility 3.4928 .74438 278 Total 3.4372 .84721 876 Quality Perception

For the impact of visibility on quality perception, the results indicate that the level of quality perception was significantly higher as the visibility level increases (Table 6), thus the hypothesis 2 is supported. In

(19)

15 other words, there is supported that the higher the visibility of brand elements on pack is, the higher quality people perceived about the product.

In the first step, the model was statistically significant (F (2,873) =10.223; p <.001), and explained 2.3% of variance. In the step 2, the model was significant (R² =.027, F (3,872) =8.080; p <.001), and the introduction of visibility explained additional variance in the model with the tolerable significance level of 5.4% (R² Change =.004, F (1, 872) = 3.729, p =.054). In the final model, only gender variable was significant (ß=.140, p <.001): women have more positive attitude towards the product than men. The beta value of visibility was not significant with 5% significance level, yet it was very close (ß=.065, p =.054), therefore moderately significant. The results of the pair-wise test between the visibility levels were not significant.

Table 6. Hierarchical Regression Model of Quality Perception

R R² R² Change B SE ß t Step 1 .151 .023*** Age .003 .002 .056 1.680 Gender .219 .052 .140*** 4.181 Step 2 .164 .027*** .004 Age .003 .002 .057 1.700 Gender .219 .052 .140*** 4.187 Visibility .063 .033 .065 1.931

Purchasing Behavior

For the impact of visibility on purchasing behavior, the results show that the willingness to pay was not significantly higher as the visibility level increases (Table 7). In other words, visibility level of brand elements on pack does not significantly affect purchasing behavior. Thus, the H3 is rejected.

In the first step, the model was statistically significant (F (2,873) = 7.311; p <.05), and explained 1.6% of variance. In the step 2, although the model was significant (R² = 1.7%, F (3, 872) = 4.896; p =.775), the introduction of additional variable of visibility could not explain a significant additional variance in the model (R² Change = .000, F (1, 872) = .081, p =.775). In the final model, only age and gender variables were significant, with age recording higher Beta value (ß=-3.009, p <.01) than gender (ß=-2.322, p <.05).

(20)

16 That is, if age increases by one year, willingness to pay for the product decreases by -3.009. On average, women are willing to pay 2.322 euro less than men.

Table 7. Hierarchical Regression Model of Willingness to Pay

R R² R² Change B SE ß t Step 1 .128 .016** Age -.034 .011 -.101** -3.014 Gender -.744 .320 -.078* 02.323 Step 2 .129 .017** .013 Age -.034 .011 -.101** -3.009 Gender -.744 .320 -.078* 02.322 Visibility .057 .201 .010 .285

Differential effects of text visibility and design visibility:

Apart from the impact of total visibility level, it was additionally examined whether that impact is significant for visibility of both types of brand elements – textual elements and graphic elements. Does the impact of visibility level of brand elements we found above differ depending on text and design aspects of brand elements? For this text, the surveyed visibility variable was used which enabled to examine the distinct effects of textual and design elements on the dependent variables. As the impact of visibility was already found to be not significant for behavior, only attitude and quality perception was examined here.

In general, only design visibility was found to be significant (F (4, 856) = 3.717, p <.01). That is, the attitude towards product was significantly more positive when people perceived the design aspect of brand elements to be comparatively highly visible. However, this effect was significant only for the condition of facing with unfamiliar brand (F(4, 419) = 2.822, p <.05).

As a whole, the results indicate that the effects are significant for both text visibility (F (4, 856) = 3.087, p <.05) and for design visibility (F (4, 856) = 7.472, p <.001). However, this effect was again significant only for unfamiliar brand condition (F(4, 426) = 3.806, p <.01). In other words, only in the case that the

(21)

17 brand they see is unfamiliar, people perceived the product quality significantly higher when they

perceived the text and/or design brand elements to be relatively highly visible.

4. Moderating effect of familiarity with the brand

To examine the differential impact of visibility level on attitude between the conditions where the brand is familiar and unfamiliar, the process model 1 was conducted. For this, a dummy variable was created in which the three treatments of varying visibility levels under the familiar brand condition were coded as 0 and the other three treatments under the unfamiliar brand condition were coded as 1.

Attitude

The results show that the interaction effect on attitude towards product between familiarity and visibility is not significant. The regression coefficient for XM is 𝑐3 = 0.0044 but is statistically not different from

zero (t (872) = -0.0617, p =0.95 (Table 8). Therefore, the effect of visibility level on consumer attitude towards product was not statistically proven to depend on brand familiarity, thus the hypothesis 4a is not supported. The total regression model was significant, accounting for relatively marginal impact of 2% of the variance in attitude.

Although a significant interaction effect could not be found, as it can be seen from the graph 1, an important tendency was observed in the effects of visibility on attitude between familiar and unfamiliar brand conditions: firstly, under the familiar brand condition the level of positivity in attitude was in general higher than under the unfamiliar condition, and secondly, in the unfamiliar brand condition, the slope of the line between low and basis visibility is steeper than familiar condition, and also the line is bent downwards for the slope between basis and high visibility, differently from the consistently upward line in the familiar brand condition.

Table 8. Interaction effect of familiarity on attitude

ß SE t p Alpha 𝑖1 3.3762 .1085 31.1075 <.001 Visibility (X) 𝑐1 .0820 .0503 1.6310 .1033 Familiarity (M) 𝑐2 -.1988 .1533 -1.2961 .1953 Visibility * Familiarity (XM) 𝑐3 .0044 .0712 -.0617 .9508 R² =.0208, p <.001, F (3, 872) = 6.1806

(22)

18

Effect SE t p

Familiar .0820 .0503 1.6310 .1033

Unfamiliar .0776 .0505 1.5381 .1244

Graph 1. Differential impacts of visibility on attitude depending on brand familiarity

To investigate this tendency further, one-way ANOVA test is conducted for the familiar and unfamiliar conditions separately (Table 9). The results indicate that the effect of visibility on attitude is significantly different between the varying levels of low, basis and high visibility in the unfamiliar brand condition (F (2,432) = 3.205, p <.05), whereas it is not in the familiar brand condition (F(2,438) = 1.083, p =.340). For the pair-wise test, the Tukey post-hoc tests revealed that the level of positivity of attitude was

significantly higher for the basis visibility group compared to the low visibility group (p <.05). However, there was no statistically significant difference between low-high visibility groups and basis-high

visibility groups.

Table 9. One-way ANOVA results and descriptives of visibility for attitude

Familiar brand Unfamiliar brand

SS DF MS F Sig. SS DF MS F Sig. Visibility 1.878 2 .939 1.083 .340 3.461 2 1.731 3.205 .042 Error 379.887 438 .867 233.288 432 .540 Total 381.765 440 236.749 434 3.2 3.3 3.4 3.5 3.6 3.7

Low visibility Basis High visibility

Familiar Unfamiliar

(23)

19 Mean SD N Mean SD N Low visibility 3.4565 1.01355 138 3.2071 .78426 140 Basis 3.5432 .94145 162 3.4177 .77964 158 High visibility 3.6206 .83002 141 3.3613 .62051 137 Total 3.5408 .93148 441 3.3322 .73858 435

Furthermore, in examining the sole effect of familiarity on attitude, there was a significant difference in the level of positivity in attitude between familiar and unfamiliar brand conditions (Table 10: F(1, 874) = 13.469, p <.001). It means that the level of attitude towards product is significantly higher when people perceived the brand familiar than unfamiliar.

Table 10. One-way ANOVA results and descriptives of familiarity for Attitude

SS DF MS F Sig. Familiarity 9.532 1 9.532 13.469 .000 Error 618.515 874 .708 Total 628.047 875 Familiarity Mean SD N Familiar 3.5408 .93148 441 Unfamiliar 3.3322 .73858 435 Total 3.4372 .84721 876 Quality Perception

The results show that the interaction effect on quality perception between familiarity and visibility is not significant. The regression coefficient for XM is 𝑐3 = -0.0375 but is statistically not different from zero (t

(872) = -0.5778, p =0.5635 (Table 11). Therefore, the effect of visibility level on quality perception towards product does not depend on brand familiarity, thus the hypothesis 4b is not supported. However, the total regression model is significant, accounting for 6.3% of the variance in quality perception towards product. In the further investigation, the results show no significant conditional effect for either familiar or unfamiliar brand condition.

(24)

20 Table 11. Interaction effect of familiarity on quality perception

ß SE t p Intercept 𝑖1 3.7746 .1087 34.7102 <.001 Visibility (X) 𝑐1 .0796 .0477 1.6680 .0957 Familiarity (M) 𝑐2 -.3045 .1450 -2.1003 .0360 Visibility * Familiarity (XM) 𝑐3 -.0375 .0648 -.5778 .5635 R2=.0630, p<.001, F (3, 872) = 19.5475

Conditional effect of Visibility (X) on Quality Perception (Y) at levels of Familiarity (M)

Effect SE t p

Familiar .0796 .0455 1.7475 .0809

Unfamiliar .0421 .0457 .9214 .3571

Graph 2. Differential impacts of visibility on quality perception depending on brand familiarity

However, although an interaction effect of familiarity could not be found, in general there was a

significant difference in the level of quality perception between familiar and unfamiliar brand conditions

3.5 3.6 3.7 3.8 3.9 4

Low visibility Basis High visibility

Familiar Unfamiliar

(25)

21 (Table 12: F (1, 874) = 54.621, p <.001), as noticeable in the graph 2. It means that people perceive product quality significantly higher when they perceived the brand familiar than unfamiliar.

Table 12. One-way ANOVA results and descriptives of familiarity for Quality Perception

SS DF MS F Sig. Familiarity 31.658 1 31.658 54.621 .000 Error 506.573 874 .580 Total 538.232 875 Familiarity M SD N Familiar 3.9342 .80212 441 Unfamiliar 3.5540 .71759 435 Total 3.7454 .78430 876 Purchasing Behavior

The moderation effect of familiarity on the relationship between visibility level and purchasing behavior was not tested as it was revealed from the previous investigation that the relationship is not significant. Thus, it is naturally concluded that the hypothesis 4c is not supported.

A one-way ANOVA test was conducted to test whether the sole effect of familiarity exist on purchasing behavior (Table 13). The result shows that there was no significant effect between familiar and unfamiliar brand conditions on the respondents’ purchasing behavior (F (1, 874) =.086, p =.769). It means that people were not significantly willing to pay more for either familiar or unfamiliar brand.

Table 13. One-way ANOVA results and descriptives of familiarity for Willingness to Pay

SS DF MS F Sig.

Familiarity 1.970 1 1.970 .086 .769

Error 19913.343 874 22.784

Total 19915.312 875

(26)

22 Familiar 4.9360 4.44454 441 Unfamiliar 5.0308 5.08490 435 Total 4.9831 4.77078 876

V. Discussion

1. Findings

Visibility effects of packaging brand elements on attitude, perception and behavior

This research investigates the effects of varying visibility levels on mobile screen of packaging brand elements on consumers’ attitude, quality perception and purchasing behavior, with the datasets of the two coffee brands that are obtained using a real auction platform against 876 Dutch consumers.

Contributing to unknown impact of mobile screen visibility of packaging brand elements on online shoppers, the current study demonstrates that the effects of mobile visibility level of brand elements on pack are significant for attitude and quality perception, but not for purchasing behavior. These results hold even after controlling for age and gender.

These outcomes indicate that the better visible brand elements are on product packaging, the more

positive attitude people have about the product and also the higher quality they perceive about the product. Meanwhile, the non-significant results on purchasing behavior indicate that people are not significantly willing to pay more for the product when packaging brand elements are relatively more clearly visible than not. These findings suggest an important evidence that, in mobile online shopping context, visibility of visual elements on product packaging could have a crucial role in influencing shoppers’ attitude formation and quality perception about the product. When the brand elements on pack are not clearly visible on mobile screen, shoppers might not be able to understand the full merits of the product and have difficulty in engaging with the brand and the product sufficiently that could hinder their positive

evaluation and judgements. In such cases, when consumers confront with products they might be influenced more by their pre-formed knowledge or experiences rather than visual assets on product packaging that are created with substantial considerations and investments and intended to effectively. This result could have substantial negative implications to brand owners because building such visual assets requires extensive considerations and investments and these efforts might not illuminate with full effectiveness as intended.

(27)

23 Additionally, for attitude towards product, although a significant difference of the effects between the pairwise levels of visibility could not be found, the difference of the effects between the low and the high visibility levels and between the low and the basis visibility levels showed an important tendency. Consumers might not perceive or evaluate importantly when packaging brand elements are visible on mobile screen at either medium or high level, while they might when the brand elements are noticeably not well visible. This implies that there is a bottom threshold of visibility level at which consumers start to perceive or feel negatively and reflect on their attitude formation and evaluation. If a product is visible just well enough not to impose substantial difficulties to understand text elements to be sufficiently informed and to identify design elements to form certain emotional reactions, consumers might not realize those visual elements perceptually different, thus not hold significantly different stance in forming their attitude or evaluating quality. When visual elements are prominently unclear to sufficiently engage with them, on the other hand, although perceived level of visibility may differ depending on the screen size of shoppers’ different mobile phone models, it could have a clear deterrence on shoppers’ interaction with the brand communication contents and seamless shopping experience, which could lead to lesser positivity towards the product or the brand.

Investigating the types of brand elements more specifically, this effect on attitude is significant only for design aspects of visibility and not textual aspects, while on quality perception the effect is significant for both text and design aspects. It implies that, in forming attitude, graphic elements might be more

sensitively perceived than texts, while in evaluating quality of the product, both text and design could play an equally important role. Additionally, these significant effects existed only for the condition where people were unfamiliar with the brand. This finding is consistent with the results of an empirical study done by Underwood, Klein and Burke (Underwood et al., 2001) that packaging imagery increased shoppers’ attention to the brand only for low familiarity brands. The differential effect of familiarity with the brand is further discussed in the next section.

This finding can be explained considering the difference between how in general people form attitude and evaluate quality. In line with the dual processing theory (Cacioppo, Petty, & Morris, 1983), forming a mere valence of attitude involves a quick reaction based on abstract feelings and intuitive associations about the object, which may not necessarily require processing textual information but mainly rely on visual abstraction based on overall design. On the other hand, making an evaluation about quality of an object, one might need to engage more in effortful and deliberate thinking based on information rather than feeling, and often compare with other options or past experiences, which would naturally require more careful attention on textual contents to gather information to make any judgement.

(28)

24

Moderation effect of familiarity with brand

For the moderating role of brand familiarity, it was hypothesized that the impact of visibility level on consumers would appear weaker for familiar brands than unfamiliar brands. In an investigation of this interaction effect of brand familiarity in such relationship, a significant interaction effect between brand familiarity and visibility level of packaging brand elements was not found for any of the dependent variables. In other words, the effect of varying levels of mobile screen visibility of packaging brand elements on attitude, quality perception and purchasing behavior did not significantly differ between when the brand is perceived to be familiar and when it is perceived to be unfamiliar.

However, in the probing test for the attitude variable which was conducted with the separate dataset of familiar and unfamiliar brand conditions, it was revealed that the level of positivity in attitude was significantly different when brand is unfamiliar but not when brand is familiar. When brand is perceived to be unfamiliar, it is concluded that the attitude is significantly less positive when packaging brand elements are visible at a poor level than at a medium level, while the attitude is not significantly more positive when packaging brand elements are visible at a high level. When brand is perceived to be familiar, on the other hand, the attitude was not significantly different depending on the visibility levels. The possible explanations for the result that familiarity mattered in forming attitude but not on quality perception and purchasing behavior are as follows: as forming an attitude is a perceptual reaction that occurs almost instantly and thus involves relatively effortless processing, it may take a peripheral route which includes relying on external cues such as familiarity or reputation or source credibility, etc. On the other hand, as evaluating true merits of a product or making purchasing decision may require relatively more deliberate thinking and effortful considerations, thus taking central route to process the messages, one might tend to scrutinize all given visual messages relatively more thoroughly, rather than

automatically resort to external cues such as familiarity with the brand.

For the significant difference of the impact of familiarity found only between low and basis visibility levels, the possible explanations are: firstly, as in line with the discussion given above for the main effect of visibility, it could be because people can be affected by visibility only when it is prominently poor, and not affected much as long as they can read and see at a certain level. Secondly, as the packaging designs at the basis visibility level that are used in the experiments are the original designs of the brands that are developed and optimized to be aesthetically the most appealing and visually the most balanced based on the pre-researches and experiences. Manipulating and making them explicitly large and extended might have made the respondents perceive the designs sloppy, less refined and therefore, less favorable.

(29)

25 Aside from the interaction effect with visibility levels, the main effect of familiarity was significant on attitude and quality perception but not on behavior. In other words, people had significantly more positive attitude and higher quality perception about the product when they perceived the brand familiar than unfamiliar, but were not willing to pay significantly more for familiar brand than unfamiliar brand. The results suggest that firstly, being familiar with a brand and therefore, having prior knowledge and/or experiences does have a positive impact on how consumers feel and evaluate the quality about the product. This result is consistent with the findings of the literature regarding the impact of brand reputation, such as on perceiving product quality or reduced risk. For purchasing behavior, on the other hand, familiarity might not have a large impact on ultimate purchase decisions as there might be many other factors that could affect to lead to this final action, such as perceived necessity of the product, aversion to switch from the currently using brand, financial considerations, etc.

2. Practical Implications

Firstly, this research holds contribution in that it sheds light on how consumers can be affected by the level of visibility of packaging brand elements in mobile online setting. As there are still little studies done about consumer tendencies and behavior on mobile online shopping environments, this study adds an important insight in the studies of mobile online context. The matter of visibility might not be too critical factor to consider in offline environment, such as on shelf in stores, because shoppers can hold up physical goods and easily examine them with no limitation on their visual zone, whereas in mobile setting, it might be a significant holdback for consumers’ barrier-free shopping experiences. Therefore, it is important to know whether the level of visibility of packaging matters in mobile online context. Secondly, this research examines actual purchasing behavior of shoppers, therefore providing more realistic insights than measuring consumers’ purchase intentions. As it is difficult to measure shoppers actual buying behavior in experiment settings, and it is known that often there is substantial gap between their intentions and behaviors, it is often challenging to claim the findings based on intentions would be indeed the case in actual purchasing situations. By using the auction system of Veylinx, this current study investigated the real behaviors of the consumers, therefore providing a truer picture of the sought effect. Thirdly, this study provides an important marketing implication regarding what marketers should consider in designing packaging for branded goods that are sold not only in physical brick-and-mortar stores but also through online, and how they should design them in terms of clarity and readability of visual elements. The results suggest that, regardless of whether a brand is known to consumers or not, they will have more positive attitude about the product and perceive the product quality higher when the brand elements on packaging are clearly readable and recognizable. If packaging design of a product is too

(30)

26 intricate, detailed or small, the brand assets and product information that are designed to optimally

influence shoppers might not be as effective as intended. This learning is important not only for developing new products or introducing foreign brands in a domestic market, but also for already established brands. When updating packaging designs or redesigning them, marketers should consider that having its reputation built in consumers’ mind does not mean it will stay unaffected when visibility of the brand elements is reduced in rolling out the new designs.

To optimize promotional effect of packaged branded goods, marketers should 1) consider visibility issue of the brand assets when targeting mobile online shoppers and either incorporate such consideration in the initial design stage or develop a different version for online selling, and 2) when designing them, test thoroughly clarity and readability of the packaging design on mobile screen considering various visual elements, such as fonts and size of textual elements, graphic patterns, pictures, product images, etc.

3. Limitations and future research

One of the main limitations of this study is that the research setting is confined to the online settings only, therefore does not reveal the differential effects between online and offline situations. Due to constraints in time and resources, the effects that are studied could not be conducted in both mobile online and offline settings and therefore compared in order to see whether such effects found in mobile online setting in this current study are indeed different in offline settings. For example, if the test of the impact of visibility levels of packaging visual elements were tested in in-store environments as well, and resulted in the findings that such impact found in mobile setting are significantly weaker or doesn’t exist at all, it would confirm the existence of such effects specific to the mobile online setting and therefore strengthen the results from the experiments.

Secondly, the results may vary depending on product types or categories. The findings of this study are obtained from one product type and category – ground coffee. However, depending on several other influential factors, such as involvement level of a product (whether a product requires low-involvement decision or high-involvement decision), or purchase motivation (whether a product is for hedonic purposes or utilitarian purposes), the results may elicit different tendencies. Also, the results may be different for specific product kinds: for example, in a previous study, it is found that, for wine products, the ones with comparatively delicate and complex designs were perceived to be more sophisticated and associated to be of higher quality (Orth & Malkewitz, 2008).

Another limitation comes from the specific setting of Veylinx auction platform. Although Veylinx auction carries an important advantage that it enables measuring consumers’ purchasing behaviors, there

(31)

27 might be substantial difference in shopping experience between Veylinx auction and other usual online shopping websites, such as Amazon, ebay, etc. In the experiments of this study, only a single front of the pack image of each product was used and no other additional product information or claim was not provided separately in the ads, differently with the regular online shopping websites where, once you click a specific product and open a dedicated page for one product, usually detailed product information is provided with multiple product shots. Thus, in real mobile online shopping situations, consumers usually have more information to reflect on their feelings or judgements about the product they are shopping. Moreover, general looks or feeling of the shopping experiences or interacting with the websites are different with Veylinx platform. All of these contextual differences might be factors to influence consumers shopping tendencies and behaviors.

For future research, therefore, it is recommended to conduct experiments of same design in offline settings, such as on shelf in stores, in promotional stands, etc. in order to find out the effects are indeed specific to mobile online shopping environments and matter less in offline environments where

consumers are free from visual constraints as on mobile screen. Secondly, to broaden the scope of the results, similar studies can be done for different types and categories of products, i.e. hedonic vs. utilitarian products, symbolic vs. functional, low vs. high involvement products, etc. Thirdly, a further study can be performed using some of the usual online shopping websites in order to validate whether the findings are consistent in real online shopping contexts rather than in an experiment setting where

respondents are aware of participating in a research.

VI. Conclusion

The broad dissemination of smartphone usage and the rapid development mobile services triggered the growing trend of mobile online shopping in recent years. In this fast transition of shopping channels, it became crucial for brand owners and marketers to understand how brands and products appear not only on physical shelf but also on consumers’ mobile screen. Acknowledging limited visual space of mobile screen that poses challenge for consumers to freely interact with visual materials of branded goods, this current study attempts to investigate the hypothesized effects of mobile visibility of packaging brand elements on consumers’ attitude, quality perception and purchasing behavior, and how the effect differs by brand familiarity.

In this study, a deductive approach was adopted to test the hypotheses, in which the experiments were incorporated with the subsequent surveys. This current study used Veylinx auction platform for the data

(32)

28 collection in which Dutch consumers voluntarily participated and were engaged in real purchasing

activities, thus enabling to document the hypothesized impact on real consumer behaviors. The results suggest that the varying levels of mobile visibility of packaging brand elements are

significant for attitude and quality perception towards product, but not for purchasing behavior. While the effect on quality perception was significant for both textual and design visual elements, the effect on attitude was significant only for design aspects of visibility. Furthermore, for both attitude and quality perception, these effects existed only for the condition when people had previous knowledge or

experience with the brand. In the search of moderating role of brand familiarity, a significant interaction effect between brand familiarity and mobile visibility levels of brand elements was not found for neither of consumer attitude, quality evaluation or buying behaviors. Although it was not found that brand familiarity holds a strengthening or weakening power on visibility effect on consumers, it was found that visibility effect is significant only when brand is unfamiliar, but only between low-basis visibility levels. A sole effect of brand familiarity on consumers was significant for attitude and quality perception but not on behavior.

This study makes several contributions to the existing literature as well as to practice. First, it takes part in filling the gap in the relatively uninvestigated field of mobile online shopping and consumers’ tendencies and behaviors in it. Secondly, it sheds light on consumer behavior in mobile online context by measuring actual purchasing behaviors rather than mere intentions, therefore delivering a more realistic and valid perspective both theoretically and practically. Furthermore, this study carries critical implications for practitioners in marketing sector by providing the useful insights on what to consider when marketing packaged goods to be viewed and sold in mobile online settings, and how to design the packaging in terms of clarity and readability of visual elements.

Acknowledging the limitations stemming from the limited resources and research methodology, further studies are recommended in several directions: future research can be conducted in offline settings in order to directly compare with the results revealed in this current study done solely in online settings, and also on different types and categories of products to verify differential effect depending on product items. Additionally, to complement the shortcomings residing in the specific characteristics of the Veylinx auction platform used for the experiments, additional studies can be done on other online shopping websites which share common looks and interface styles as regular online shops.

Effects of visibility of packaging brand elements on consumer attitude, quality perception and purchasing behavior in mobile online shopping context