How do message content and product type influence the effectiveness of location-based advertising?

(1)

HOW DO MESSAGE CONTENT AND PRODUCT

TYPE INFLUENCE THE EFFECTIVENESS OF

LOCATION-BASED ADVERTISING?

Master Thesis

Student: Marjolein Heijne

Student number: 11414510

Program: MSc. Business Administration

Track: Digital Business

Under the supervision of: Dr. J. Li

(2)

2 Statement of originality

This document is written by Student Marjolein Heijne who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document are original and no sources other than those mentioned in the test and the references have been used.

The faculty of Economics and Business is only responsible for the supervision of completion of this work, not for the content.

(3)

3 Acknowledgement

This thesis is the final step toward achieving my master’s degree in Business Administration and specialization in Digital Business at the University of Amsterdam. Writing this thesis has been a very interesting experience, in which I put much effort and finally gained a lot. I would like to take this opportunity to thank my supervisor Jing Li for supporting me in every possible way she could! She enthused and guided me through this important process of writing my master thesis with her expertise and positive criticism. It has been a great pleasure to work with her. Furthermore, I would also like to thank my family and friends for supporting me and having faith in me.

I hope you enjoy reading this thesis.

Kind regards, Marjolein Heijne 23rd of June 2017

(4)

4 Abstract

Purpose – This study aims to examine the effects of message content (brand advertising and brand promotion) and product type (hedonic and utilitarian) on the effectiveness of location-based advertising (LBA) on mobile devices. While LBA continuously advances on a

technological level, research on its effectiveness is scarce. In addition, little is known about the effect of regulatory fit on consumer behavior. This study therefore aims to explore whether message content is related to LBA’s effectiveness, and whether product type is a moderator of this relationship.

Design – This is an empirical study that adopts an experimental design based on an online survey. A between subject design with four conditions was adopted, two conditions of message content (brand advertising and brand promotion) and two conditions of product type (hedonic and utilitarian).

Findings – The results of this study reveal that message content has significant influence on LBA’s effectiveness. Purchase attitude, attitude toward the ad, and attitude toward the product are more positive for promotional LBA than for advertising LBA. Product type does not have an impact on the relationship between message content and LBA’s effectiveness.

Contribution – The contribution of this paper is three-fold. First, few studies have considered the role of message content on LBA’s effectiveness. This study showed that message content is important to take into account, as it affects intention to buy the product. The results suggest that the focus of marketers who wish to increase short-term buying behavior should be on promotional LBA and not advertising LBA. Promotional LBA was shown to be more effective in increasing attitude toward the product. The trade-off between advertising and promotional content appears to be relevant, important, and under researched. Moreover, this study responds to the need to explore which products or services are most suitable to be advertised with LBA on mobile devices, making valuable contributions to both mobile

(5)

5 marketing and consumer behavior research. Finally, although LBA adds several important opportunities for marketers, research on its effectiveness is scare. This study addresses the need for more in-depth research into the effectiveness of location-based advertising.

Keywords – mobile advertising, location-based advertising, advertising effectiveness, attitude,

(6)

6 Table of Contents Statement of originality 2 Acknowledgement 3 Abstract 4 1. Introduction 9 1.1 Research question 10 2. Literature Review 13 2.1 Mobile Advertising 13

2.2 Location-Based Advertising (LBA) 13

2.3 LBA Message Content 15

2.4 Product Type 16

2.5 The Effectiveness of LBA 16

3. Conceptual Framework and Hypotheses 19

3.1 Conceptual Framework 19 3.2 Hypotheses 19 3.2.1 Message Content 20 3.2.2 Product Type 21 3.2.3 Control Variables 23 4. Method 24 4.1 Research Design 24

4.2 Data Collection and Sample Description 25

4.3 Measures 26

4.3.1 Independent Variables 28

4.3.2 Dependent Variables 29

(7)

7

4.4 Method 30

5. Data Analysis and Results 32

5.1 Preliminary Analysis 32

5.1.1 Validity and Reliability 32

5.1.2 Correlations 34

5.2 Hypotheses Testing 35

5.2.1 The Effect of Message Content on LBA’s Effectiveness 35 5.2.2 The Moderating Effect of Product Type on LBA’s Effectiveness 36

5.3 Summary of results 38

6. Discussion and Conclusion ` 40

6.1 Theoretical and Practical Implications 42

6.2 Limitations and Future Research 43

6.3 Conclusion 45

7. References 46

8. Appendix 51

Appendix A – Survey Scenarios 51

Appendix B – Survey 53

List of Tables

Table 1: Demographic Summary 26

Table 2: Summary of Measures 27

Table 3: Participants per Scenario 32

Table 4: Explanatory Factor Analysis 33

Table 5: Reliability of Scales 34

(8)

8

Table 7: ANCOVA H1 36

Table 8: Descriptive Statistics of ANCOVA H1 36

Table 9: ANCOVA H2 38

Table 10: Results of Hypotheses Testing 38

List of Figure

(9)

9 1. Introduction

On January 9th 2007, Apple introduced the iPhone, a small and lightweight handheld mobile phone. Steve Job, Apple’s CEO, promised: “This will change everything” (Apple, 2017). His forecast was not an exaggeration. Merely ten years later, Apple’s iPhone is considered as a massive breakthrough as it has completely redefined mobile computing, which triggered the worldwide spread of mobile devices. Statistics show that the number of mobile devices is higher than the amount of human beings (Boren, 2014). The impact of mobile is far-reaching. Sales figures from Forrester show how mobile phones influence the shopping experience. More than $1 trillion of total retail sales in 2015 were influenced by mobile phones, with most of this coming from in-store transactions and further growth expected (Forrester, 2016).

The rapid adoption of mobile devices by consumers caused the incorporation of mobile advertising as an important component of a company’s overall communication strategy. Mobile advertising entails the communication of products or services to mobile device and smartphone consumers (Techopedia, 2017).With consumers around the globe glued to their phones, the future of mobile advertising looks bright. Firms are increasingly allocating more budgets to their mobile platform. It is forecast that mobile advertising will nearly double over the next few years, hitting $195.55 billion to account for 70% of total media spending globally by 2019 (eMarketer, 2015).

Given the outburst of mobile advertising, marketers are continuously searching for innovative means to exploit this coveted media platform (Limpf & Voorveld, 2015).

Location-based advertising (LBA) seems to hold the greatest promise of mobile advertising according to advertisement professionals (van't Riet et al., 2016). LBA can be defined as a subset of location-based marketing (LBM), that uses positioning technologies to target consumers based on their proximity to places of relevance and interest (Unni & Harmon,

(10)

10 2007). Nowadays, most smartphones are equipped with location-tracking technologies, such as the global positioning system (GPS), which enables marketers to pinpoint users’ real-time locations and deliver location-specific ads on their mobile devices (Lee, Kim, & Sundar, 2015). Location-based ads have a tendency to be highly individualized, with instant,

personalized, and location-aware aspects that adapt to the specific profile of each ad receiver (Bruner & Kumar, 2007; Dhar & Varshney, 2011). The use of context-awareness and

personalization entails that location-based ads are likely to be perceived as more appealing and persuasive, and thus used as a fundamental element of successful mobile advertising (Lee et al., 2015).

Consequently, profitable expectations draw attention to LBA as an area of research. When evaluating current publications on LBA, one can conclude that research on LBA is scarce (Banerjee & Dholakia, 2012). More research is desperately needed since there have barely been in-depth examination of LBA (Unni & Harmon, 2007). In addition, LBA on mobile devices is (still) in the hands of technically oriented scientists. While LBA has its roots in technically oriented disciplines, it is critical to examine marketing-related issues of LBA (Bauer & Strauss, 2016). Furthermore, it is important for practitioners to have actionable guidelines to harness the effectiveness of LBA amidst it growing popularity.

1.1 Research question

The purpose of this study is to provide a complete understanding of LBA by dissecting the effectiveness of LBA. Specifically, this research explores the following question:

“How do message content and product type influence the effectiveness of location-based advertising (LBA)?”

A noted element of advertising is message content. The holy grail of reaching the right person with the right message at the right time has greatly concerned marketing researchers through

(11)

11 the years. If the content of an ad is highly relevant to the consumer, LBA can serve effectively as a trigger for action (Unni & Harmon, 2007). For example, Barwise and Strong (2002) suggest that “mobile advertising would be more effective for frequently bought, low-priced products than for more expensive products” (Barwise & Strong, 2002). One could therefore ask if it will apply likewise for location-based ads. Many have therefore cited the need to examine which products or services are most suitable to be advertised with LBA on mobile devices (Bauer & Strauss, 2016; Bruner & Kumar, 2007; Unni & Harmon, 2007; Xu, Luo, Carroll, & Rosson, 2011). Prior research found that mobile advertising is more effective for habitually bought, low-priced products than for high-priced products (Unni & Harmon, 2007). The question arises whether this holds true for LBA.

The proposed model in this research focuses on consumers’ attitudes toward purchasing advertised products as a measure of advertising effectiveness. Although other measures of advertising effectiveness exist, attitudes and/or intentions to purchase are

frequently used in advertising research. Moreover, increasing favorable product attitudes is an example of a common campaign objective in practice (Bart, Stephen, & Sarvary, 2014). This is also in line with related research on the effectiveness of online (nonmobile) advertisements (Goldfarb & Tucker, 2011).

Researching the proposed model can result in both theoretical and practical contributions. Examination of (potential) influencers of LBA’s effectiveness is relevant because LBA represent potentially powerful ways to enable marketers to reach and interact individually with consumers in new and innovative ways whenever and wherever they are ready to buy. As consumers increasingly use location-based services and time-sensitive offerings, investigating the effectiveness of LBA is critical for the growth of the digital advertising industry. Given the fact that LBA is mainly explored from a technical perspective

(12)

12 (Bauer & Strauss, 2016), this research will to close a gap by further examining the user

perspective. Furthermore, this study bundles interdisciplinary research from the fields of marketing, psychology, and technology.

The remainder of this paper is organized as follows. The next chapter reviews the literature about location-based advertising (effectiveness), message content, and product type. Subsequently, the third chapter introduces the conceptual framework and hypotheses. The next chapter describes the research methodology followed by the results of data analyses. Lastly, chapter 6 provides a careful discussion and conclusion of the results, including academic relevance, managerial implications, limitations and suggestions for future researches.

(13)

13 2. Literature Review

This chapter reviews the literature on location-based advertising (LBA) and related concepts.

2.1 Mobile Advertising

Mobile advertising entails the communication of products or services to mobile device and smartphone consumers (Techopedia, 2017). A multitude of mobile advertising types have been identified by Zoller and Oliver (2011): messaging, display advertising, in-application advertising, and spot advertising around mobile TV and video. Furthermore, companies are focusing their efforts on branded mobile websites, branded applications and mobile search. Utilizing bar codes and coupons for rewards and redemptions are another common type of mobile advertising (Zoller & Oliver, 2011).

Technology has allowed the toolbox of mobile advertising to enlarge. To maximize return on investments, marketers nowadays have various tools to carefully segment and target their customers. Positioning technologies (e.g. GPS on smartphones) are frequently named as important means that enable marketers to incorporate real-time, location-based data to target consumers based on their proximity to places of relevance and interest (Unni & Harmon, 2007). Mobile advertising that utilizes and is tailored to the geographic location of the consumer is generally termed mobile location based advertising ((Bruner & Kumar, 2007; Unni & Harmon, 2007; Xu et al., 2011).

2.2 Location-Based Advertising (LBA)

Location-based advertising (LBA) has emerged as an increasingly attractive marketing platform for companies (Lee et al., 2015). Bruner and Kumar (2007) define LBA as marketer-controlled information customized for recipients’ geographic positions and received on mobile communication devices. The theory behind LBA is not new; it has been marketers’

(14)

14 area of practice for a long time. Who cannot bring to mind roadside billboards telling you how to get to a store (e.g. “Turn left to Ikea”)? In recent years, digital signs, which allow for

dynamic multimedia presentations, whereby content can be changed quickly by remote access, have increasingly replaced static billboards (Bauer, Dohmen, & Strauss, 2011). However, LBA on mobile devices adds several important opportunities for marketers. Bauer and Strauss (2016) suggest: “First, it addresses consumers (i) individually, (ii) based on their current location, and (iii) dynamically in real-time, which is an important distinction to locally installed signs. Second, LBA on mobile devices provides flexibility concerning content. Misplaced, out-of-date or expired information may be replaced quickly by remote access.”

In their paper, Bauer and Strauss (2016) analyze the research field of LBA. Their investigation revealed ten subfields of research. While they focus on the applied

methodologies in several publications, their analysis still offers a good overview of researched aspects related to LBA. Findings indicate that consumers may perceive LBA positively if it is being sent in the right format to the right person at the right time. Time, profile, interest, and preferences count for the most frequently used attributes used in LBA research as adaptation criterion in addition to ‘location’ (Bauer & Strauss, 2016). In this way, Banerjee and Dholakia (2008) reveal that consumers consider LBA most useful in leisure time. Consumers are more willing to respond to the offer when they are ready for consumption such as while they are shopping. Likewise, location based ads generated more willingness for immediate responses to the ad in public locations than private locations, such as being at home or at work.

In addition, mobile advertising emphasize the importance of perceived benefit.

Marketers should focus on delivering properly targeted LBA, which is relevant and beneficial in the eyes of the consumer (Unni & Harmon, 2007; van't Riet et al., 2016). Consumers may strongly value receiving personalized messages through LBA. On the other hand, there are obvious privacy concerns regarding LBA. Since the mobile phone is a very personal device

(15)

15 and due to advanced location tracking technologies, LBA on mobile devices can be viewed as an intrusion into a consumer's personal space (Limpf & Voorveld, 2015; Unni & Harmon, 2007; Xu et al., 2011).

2.3 LBA Message Content

The content of the message is a factor that takes center stage in the design of effective ads. A marketing practitioner must understand how to form the right message to stand out in the constantly evolving world of e-commerce. Unni and Harmon (2007) identify two broad categories of LBA message content; namely brand advertising and brand promotion. Brand advertising is used by companies to raise consumers awareness of the brand. While it is unclear to what extent advertising is effective in generating sales (Sethuraman, Tellis, & Briesch, 2011), brand advertising has been found to increase brand awareness and brand salience. On the other hand brand promotion, which provides information about available promotional activities or an explicit incentive, can be relevant for consumers (Chen & Hsieh, 2012; Unni & Harmon, 2007). Both types of content could affect consumers’ desire to purchase the product after exposure to the ad.

The effects of advertising and promotion on brand equity and brand profitability have gained much attention in marketing literature (Jedidi, Mela, & Gupta, 1999; Yoo, Donthu, & Lee, 2000). However, there is little to no research which addresses the trade-off between advertising and promotion on the subject of mobile advertising. An interesting question therefor is whether brand advertising or brand promotion is the most influential in convincing consumers to take a purchasing action after viewing a location-based ad. It is crucial to examine these effects since brand promotion is more expensive to marketers (Unni & Harmon, 2007).

(16)

16 2.4 Product Type

Another factor that seems to influence advertising effectiveness is the type of product that is promoted in the ad. Previous research studies frequently made a distinction between utilitarian and hedonic products (Geuens, De Pelsmacker, & Faseur, 2011). Those products for which consumption is cognitively driven, instrumental, and goal oriented, are commonly classified as “utilitarian products”. They serve practical or functional tasks. Examples of utilitarian categories include office supplies, home appliances and garden equipment. On the other hand, hedonic products are consumed for luxury purposes and are associated with an affective and sensory experience of entertaining and pleasure (Bart et al., 2014). DVD’s, apparel and cosmetics are examples of hedonic categories.

There is evidence to suggest that mobile advertising would be more effective for frequently bought, low-priced products than for more expensive products (Barwise & Strong, 2002). It is unclear, however, whether these results generalize to LBA. Hence, a significant body of professional publications emphasizes the urge to shed more light on which products or services are most suitable to be advertised with LBA on mobile devices (Bauer & Strauss, 2016; Bruner & Kumar, 2007; Unni & Harmon, 2007; Xu et al., 2011). Research in this direction will be very valuable for the advertising industry.

2.5 The Effectiveness of LBA

A great deal of research work has been done to scrutinize mobile advertising effectiveness’s in general (Hühn, Khan, Lucero, & Ketelaar, 2012). Multiple measurement methods of advertising effectiveness exist. Effectuating favorable product attitudes and increasing purchase intentions are examples of common campaign objectives in practice. Hence, attitudes and intentions are frequently employed constructs in advertising/marketing literature (Bart et al., 2014). Attitude represents a key determinant of behavioral intention.

(17)

17 Whereas attitudes can be considered as the evaluation for or against some object, intentions represent someone’s motivation to exert effort to carry out a behavior.

Bellman et al. (2011) add that an attitude is easier to alter than an intention. An enhancement in attitude can be a signal of a short-lived effect or an early indicator of a behavioral change that will take place later. The relationship between consumer attitude and consumer behavior has been found in many studies (e.g. (Bauer & Strauss, 2016). Attitude is named as a significant predictor for intention to use and purchase intention. These behavioral intentions can be considered as conative effects. The hierarchy of the advertising effects model suggests that a consumer will go through a sequence of mental stages from cognitive, affective to conative, closely following the typical attitude structure components (Xu, Oh, & Teo, 2009).

Several related key antecedents of mobile advertising effectiveness are found in the literature. First, the overall positive or negative evaluation of the ad in the eyes of the receiver is captured by attitude toward the ad. This construct had often been applied for assessing the effects of advertisements and for identifying further factors which influence advertising effectiveness (MacKenzie, Lutz, & Belch, 1986). Likewise, like manner, attitude toward the ad positively affects response to mobile advertising (Yang, Kim, & Yoo, 2013).

Second, drawing on the advertising literature, product-related aspects can partially explain differences in campaign effectiveness. For instance, research has found that product attribute beliefs mediated attitude formation (Mitchell & Olson, 2000). Extensive literature on advertising effects provides support that campaign effectiveness seems to vary considerably according to the advertised product (Bart et al., 2014).

Aside from attitude toward the ad, it is pivotal to assess a user’s positive attitude toward a technology, since this will increase their intentions to use the technology (Xu et al., 2009). Yang et al. (2013) found that LBA acceptance positively affects (a) attitude and (b)

(18)

18 response to mobile advertising. Taken together, these studies suggest that higher levels of antecedent variables such as purchase attitude, attitude toward the ad, attitude toward the product, and attitude toward LBA indicate effective advertising.

(19)

19 3. Conceptual Framework and Hypotheses

This chapter elaborates on the relationships and effects between previously introduced subjects. It starts with the conceptual framework. Hypotheses are addressed afterwards.

3.1 Conceptual Framework

The main question is visualized in Figure 1, which can be used to understand the subject matter.

Figure 1: Conceptual Framework

The model shows that message content is expected to have a direct impact on ad effectiveness. It is presumed that the type of product advertised in the ad moderates this relationship. Among the outcomes, this study focuses on a four key variables of effectiveness —that are, purchase attitude, attitude toward the ad, attitude toward the product, and attitude toward LBA.

3.2 Hypotheses

This paragraph introduces the hypotheses. This study uses the regulatory focus theory (RFT) to motivate the hypotheses. In line with the RFT, people can be categorized based on their regulatory orientation in pursuing a goal. A distinguish can be made between prevention focused and promotion focused. A prevention focus stresses safety and security and means

(20)

20 minimizing negative outcomes. Contrary to a prevention focus, a promotion focus involves maximizing positive outcomes and emphasizes hope and achievement. RFT posits that people make choices that are consistent with their regulatory orientation in goal pursuit. Regulatory fit leads to greater customer engagement. Engaged customers tend to value and pay more for products. With regard to the effectiveness of LBA, it is relevant to note that promotional oriented customers are more likely to adopt nontraditional marketing channels (Kushwaha & Shankar, 2013) such as LBA.

3.2.1 Message Content

Marketers aim to achieve a regulatory fit since it directly affects consumers’

willingness to pay for a product and their attached value to a product. Properly targeted LBA would deliver only relevant messages in the right context to the consumer. As discussed previously, a marketing message can be divided into brand advertising and brand promotion based on its content. There is a strong link between brand promotion and the orientation of promotion focus. Promotion focused people are more likely to be driven by the need to signal advancement. They are driven by positive outcomes; therefore they are more sensitive to price promotions, in order to obtain profit (Kushwaha & Shankar, 2013). Promotional LBA, in contrast with advertising LBA, contains a concrete incentive. For that reason, brand promotion can positively affects promotion focused consumers’ purchase decision. Since brand advertising does not provide a concrete reason why the product should be purchased at that moment, it is less likely to affect promotion focused consumers’ purchase behavior. Taking into account that promotional oriented customers have more potential for LBA compared to prevention oriented customers, the current study hypothesizes that the advertising effect of promotional LBA is stronger than the effect of advertising LBA.

(21)

21 In line with this, Unni and Harmon (2007) report that “LBA tends to work best when consumers have a tangible offer or a marketing message on which they can act upon” (Unni & Harmon, 2007). This is due to the fact that consumers mentally construe promotional offers more concretely, which in turn, increases their involvement and purchase intent (Luo, Andrews, Fang, & Phang, 2013). This corresponds with promotional LBA where the ad contains an explicit promotional offer. Therefore, a stronger advertising effect after exposure to a promotional LBA is expected compared to advertising LBA, where the ad does not contain a concrete cue. This leads to the following hypotheses:

H1a: Purchase attitude would be greater for promotional LBA than for advertising LBA.

H1b: Attitude toward the ad will be greater for promotional LBA than for advertising LBA.

H1c: Attitude toward the product will be greater for promotional LBA than for advertising LBA.

H1d: Attitude toward LBA will be greater for advertising LBA than for promotional LBA.

3.2.2 Product Type

The advertising effect of message content may also be influenced by the type of product advertised. In line with prior research (Bart et al., 2014; Kushwaha & Shankar, 2013), this study draws a distinction between utilitarian products and hedonic products. Utilitarian products are typically linked to functional and instrumental attributes. It is relatively easy to compare and evaluate these products. Consumers who are shopping for utilitarian products make habitual purchases and goal-directed choices. They are cognitively involved and prefer value efficiency, which is often associated with a greater prevention focus. Goal-oriented

(22)

22 shopping behavior associated with hedonic product categories shows more experimental consumption, visible in unplanned purchases and variety seeking behavior. Hedonic products are viewed as more risky. Consumers are uncertain about future preferences, which leads to variety seeking as a choice heuristic.

Differences between utilitarian and hedonic product types are likely to influence consumer shopping behavior. Hedonic products are often associated with a greater promotion focus, whereas utilitarian attributes correlate with a greater prevention focus (Kushwaha & Shankar, 2013). In this way, an ad of a hedonic product maps with the promotion-focused attributes of promotional LBA, providing a strong content-category fit. In contrast, utilitarian products mismatch with promotional LBA. These sorts of advertisements would therefore probably be less effective. The influence of product type is weaker on the effect of advertising LBA, assuming that this form of message content is less adequate in serving the needs of LBA customers.

If shoppers’ focus fits with their content preference based on the product category characteristics, they will likely experience greater regulatory fit. In turn, greater regulatory fit will lead to a more positive attitude and greater purchase intention of products. Thus, an ad with a greater regulatory fit will probably be more effective. Therefore, it is predicted that hedonic (relative to utilitarian) products are conducive for the effect of message content on advertising effectiveness. This expectation can be expressed more formally as follows:

H2a: The effect of message content on purchase attitude is greater for hedonic products than that for utilitarian products.

H2b: The effect of message content on attitude toward the ad is greater for hedonic products than that for utilitarian products.

H2c: The effect of message content on attitude toward the product is greater for hedonic products than that for utilitarian products.

(23)

23

H2d: The effect of message content on attitude toward LBA is greater for hedonic products than that for utilitarian products.

3.2.3 Control Variables

Prior research on advertising effects indicates that additional factors may have to be considered due to their potential influence on the key constructs in the present research model. Any attributes relating to personality, values, attitudes, interests, or lifestyles may influence consumers’ attitude following exposure to LBA. This study therefore controls for the effect of prior experience with LBA and purchase impulsiveness.

Purchase impulsiveness refers to a consumer's tendency to buy spontaneously, in other words, an unplanned decision to buy a product or service. Highly impulsive buyers are more likely to experience spontaneous buying stimuli; their shopping lists are more "open" and receptive to sudden, unexpected buying ideas (Rook & Fisher, 1995). As a result, impulsive buyers are more likely to respond affirmatively and immediately to their buying impulses when exposed to LBA. Therefore, this factor is included as a control variable in this study.

Previous experience and past usage has been found to be influential on both future intentions and behavior. It has been shown that people who have higher levels of prior experience with the mobile medium are more likely to adopt a new mobile-based service (Varnali, Yilmaz, & Toker, 2012). Likewise, it is predicted that as consumers become more familiar with LBA, their perceptions regarding LBA would be attenuated, and hence this may directly affect their attitude. Therefore, prior experience with LBA is included as a control variable in this research. Furthermore, marketing research often takes demographic factors such as gender, age, and education into account. This study does not expect a particular impact of demographics on the effectiveness of LBA but will include these variables as controls nonetheless.

(24)

24 4. Method

To investigate the influence of message content and product type on LBA’s effectiveness, quantitative data has been collected and analyzed to test the proposed hypotheses. This chapter provides descriptions of the research design, data collection and used sample, used method and scales, and the analysis method.

4.1 Research Design

An experimental approach was taken in order to research the relationships between variables. Experiments are better suited than observational studies to draw conclusion about causal relationships between variables. An experiment deliberately imposes a treatment on a group of objects or subjects in the interest of observing the response. Measuring LBA’s effectiveness by including users in the loop, complies with Bauer and Strauss’s (2016) call for more user-centric research in the field of LBA. Subjects are randomly assigned to the

treatment conditions (levels of the independent variable). The only differences in the groups would be due to chance. The hypotheses were tested by a lab experiment. Hereby, it is easy to randomly allocate participants to different treatments and it provides the highest level of control over variables and research context. As a consequence, this ensures a high internal validity (Demmers, 2017). The illustrated quantitative research strategy is in line with the research strategies adopted by researchers investigating advertising effectiveness’s (Bauer & Strauss, 2016). An online research tool, Qualtrics, was used to develop the experiment as well as the subsequent survey. Qualtrics is an industry-leading provider of online survey content and freely available to students of the University of Amsterdam.

The population for this study consists of the Dutch consumers who have a mobile phone, which is approximately 12.6 million people (Statista., 2017). Because the population is large and no sampling frame could be retrieved for it, non-probability sampling technique was

(25)

25 used. The respondents of the survey have been selected using a snowball sampling technique, in order to include a large size of respondents. This study also used a convenience sampling technique, since I started the survey by inviting the respondents who were easily to reach for me.

4.2 Data collection and Sample Description

The survey was distributed via Internet. The final questionnaire has been administered on the 19th of April. Participants received an invitation by email or were asked via a

Facebook post to fill in the questionnaire. An incentive has been used to increase the likelihood of response. Participants had a chance to win one of the two €10 vouchers at Bol.com (Dutch online retailer) if they wrote down their email address at the end of the questionnaire. On the 1st of May, all questionnaires that have been completed were returned. In total, the questionnaire had been active for 12 days. Participation was voluntary. In total 261 questionnaires were returned and 204 of them were used for analysis.

The questionnaire was available in English and Dutch. There was a note on the welcome page to make participants aware of the possibility to change their language preference. The survey consisted of three sections. In the first section, participants read an introduction with general information about the survey followed by one of the aforementioned scenarios. Questions about how they felt about the advertisement, the brand and the store followed each scenario. In addition, participants were asked about their attitudes toward the product, toward the ad and toward LBA. The third section dealt with questions related to the control variables. The questionnaire ended with asking the respondents about their

demographics. Since these questions were confidential, respondents were asked at the end of the survey in order to increase the response rate. All responses were processed anonymously. Appendix B shows the final version of the questionnaire in English.

(26)

26 A 2 (message content: brand advertising vs. brand promotion) x 2 (product type: utilitarian vs. hedonic) between-subjects experiment was conducted in order to examine the proposed hypotheses and research question. Subjects were randomly assigned to one treatment. All responses were collected in computer-based settings. 204 participants (62 males, 412 females) were recruited by e-mail and social media. Participants’ ages ranged from 17 to 81 years (M = 29.05, SD =11.73). A majority of the respondents’ highest achieved education level was University (Bachelor), consisting 46.6% of the total. Table 1 provides a

demographic summary.

Table 1: Demographic Summary

Gender Frequency % Male 62 30,4 Female 142 69,6 Age Frequency % <20 7 3,43 21-25 132 64,71 >26 65 31,86 Education Frequency % Secondary school 19 9.3 University (Bachelor) 95 46.6 University (Master) 86 42.2 PhD or other degrees 4 2.0 4.3 Measures

The survey was developed by using previously validated scales to prevent any reliability and validity issues. The items for all these measured variables were adopted from prior advertising research and slightly modified to fit the study’s context. To ensure that all participants were users of a mobile device, participants were asked whether they own a mobile phone (yes or no), with all 204 respondents indicating that they owned one. An instructed response item was used once to minimize careless responding. An overview of all

(27)

27 variables, their measures and measure levels can be seen in Table 2 Summary of Measures.

Table 2: Summary of Measures

Variable Measure Level Items

Message Content 1 = Advertising, 2 =

Promoting

Nominal See scenarios

Product Type 1 = Hedonic, 2 =

Utilitarian

Nominal See scenarios

Purchase Attitude

Kozup et al., 2003

7-points Likert Scale (1 = Totally disagree, 7 = Totally agree)

Interval/Ratio 1. How likely would you be to purchase the product, given the information shown?

2. Assuming you were interested in buying this product, would you be more likely or less likely to purchase the product, given the information shown? 3. Given the information shown, how likely is it that you would consider the purchase of the product, if you were interested in buying this product?

Attitude toward the Ad

MacKenzie et al., 1986

7-points Likert Scale (1 = Not attractive, 7 = Attractive)

Interval/Ratio 1. How would you describe your overall opinion of the advertisement?

7-points Likert Scale (1 = Irritating, 7 = Not irritating)

2. How would you describe your overall opinion of the advertisement?

7-points Likert Scale (1 = Not interesting, 7 = Interesting)

7-points Likert Scale (1 = Bad, 7 = Good)

Attitude toward the Product Trampe et al.,

2010

7-points Likert Scale (1 = Not functional, 7 = Functional)

Interval/Ratio 1. How would you describe your overall opinion of the product?

7-points Likert Scale(1 = Unfavorably, 7 = Favorably)

2. How would you describe your overall opinion of the product?

7-points Likert Scale (1 = Not interesting, 7 = Interesting)

7-points Likert Scale (1 = Bad, 7 = Good)

(28)

28

Attitude toward LBA

Tsang et al., 2004

7-points Likert Scale (1 = Totally disagree, 7 = Totally agree)

Interval/Ratio 1. In general, location-based advertising would probably be irritating.

2. In general, location-based advertising would be entertaining.

3. Location-based advertising will provide useful information.

4. In general, I would be favorable toward location-based advertising.

5. I will probably not pay attention to location-based advertising.

4.3.1 Independent Variables

All treatments were manipulated using scenarios of real life situations, developed by the researcher itself. Each scenario contained a section of text followed by a written on-screen push notification. Subjects were asked to imagine a situation that is verbally described in the online questionnaire. The set of stimuli consisted of four scenarios reflecting extremes in levels of message content and product type, which are shown in Appendix A. They are situations from everyday life that can easily be imagined by every subject. Based on previous research into factors influencing the impact of LBA (Banerjee & Dholakia, 2012; Luo et al., 2013), it has been decided to sketch a public location (low geographic distance) and a same day offer (low temporal distance) to increase the odds of advertising effectiveness'.

Unni and Harmon (2007) were among the first to present how initial consumer evaluations of mobile LBA differ between pull-based and push-based LBA. In pull-based LBA (i.e., overt-based, reactive LBA) information is only send to the consumer when it is explicitly requested for. In the push-based LBA (i.e. covert-based, proactive LBA), content is sent automatically to consumers based on screening their behaviors through real-time location tracking of their mobile devices (Limpf & Voorveld, 2015; Unni & Harmon, 2007; Xu, Luo, Carroll, & Rosson, 2011). Particularly for push-based LBA is the potential harm of a

(29)

29 wherever they are. However, a solution is provided by means of a permission-based approach, where users can set their own readiness to receive and can shield unsolicited push messages or “spamming”. This permission for push-based LBA is required by law in the USA and Europe (Xu et al., 2009). Therefore, permission-based LBA is the focus of this research. The push-based LBA is used in the experiment scenarios because push-push-based applications can exploit location-based information largely. The operationalization of the independent variables runs as follows:

 Message content: Types of message content were Brand Advertising and Brand Promotion.

o Brand Advertising was expressed by mentioning a recently launched collection/product (New flavors of X are available in store).

o Brand Promotion was expressed by mentioning a discount of 20% on a collection/product (Only today: 20% discount on X).

 Product type: Types of products were Utilitarian and Hedonic. o Utilitarian Product (washing capsules of PureWash) o Hedonic Product (ice creams of IceDelice

4.3.2 Dependent Variables

Purchase Attitude was measured with three items adopted from Kozup et al. (2003). This measure included items such as “How likely would you be to purchase the product, given the information shown?” Participants were asked to rate their degree of agreement on the statements about different possible barriers to mobile shopping. Items were rated on a scale from 1 (totally disagree) to 7 (totally agree). The following question was asked: “Please indicate on a scale from 1 to 7 that to extent do you agree or disagree the following statements (1=totally disagree, 7=totally agree)”.

(30)

30 Attitude toward the ad and attitude toward the product were assessed using an index of four adjectival items such as ‘‘interesting” and “good” adapted from MacKenzie et al. (1986) and Trampe et al. (2010).

Attitude toward mobile LBA was measured with four items adopted from Tsang et al. (2004). This measure included items such as “In general, location-based advertising would probably be irritating” and “In general, I would be favorable toward location-based

advertising”. Participants were asked to rate their degree of agreement on the statements about different possible barriers to mobile shopping. Items were rated, in the same manner as

Purchase Attitude, on a scale from 1 to 7.

4.3.3 Control Variables

Purchase impulsiveness was operationalized with three items adopted from Rook and Fisher (1995). Prior experience with LBA was measured with: “How frequently have you used or received mobile location-based advertising during the past year?” Participants were asked to rate this item on a scale from 1 (Never) to 7 (More than 100 times).

The model also includes relevant demographic variables to describe and assess the approximate representativeness of the sample. Age was measured by asking respondents to indicate their year of birth. Participants’ gender was identified using the question “What is your gender?” Education was measured by questioning the participant’s highest completed level of education (or the program he or she is currently enrolled in). Participants were asked to rate this item on a scale from 1 (Primary school) to 5 (PhD or other degrees).

4.4 Method

This research used SPSS to analyze the collected data. Preliminary analyses consisted of cleaning data, checking for outliers, and recoding. Furthermore, an exploratory factor analysis and a reliability analysis have been conducted to check the validity and reliability of

(31)

31 variables. Hereafter, Pearson’s correlation coefficients were calculated to check the relations between variables. Next, the overall scores of variables were calculated and assumptions for ANCOVA were checked. All assumptions hold for the data. Thus, a one-way ANCOVA with backward elimination was run four times to test the influence of message content on purchase attitude, attitude toward the ad, attitude toward the product, and attitude toward LBA. Finally, a factorial ANCOVA with backward elimination was done four times to test the moderation effect of product type on the relationship between message content and advertising

effectiveness. A full model was tested including the control variables Gender, Age,

Education, Prior Experience with LBA and Purchase Impulsiveness. Effects were deleted one by one until a stopping condition (p < 0.30) was satisfied. At each step, the effect showing the smallest contribution to the model was deleted. The data analysis process and the results will be explained in more detail in the next chapter.

(32)

32 5. Data Analysis and Results

This chapter addresses all hypotheses previously mentioned, states the statistical tests that were used, the results and where necessary a clarification of those results.

5.1 Preliminary Analysis

The first step was to clean data from inadequate results. Online survey collected 261 questionnaires in 12 days. 57 of them were discarded because of a low completeness less than 100%. All respondents meet the requirement of having a mobile phone. The final database consisted of 204 respondents, N=204. The remaining questionnaires were all complete without any missing data. Table 3 displays the final breakdown of participants per scenario. The items “In general, location-based advertising would probably be irritating”, “I will probably not pay attention to location-based advertising” and, “I carefully plan most of1 my purchases” were recoded, because these three items were counter-indicative.

Table 3: Participants per Scenario

Hedonic Product Utilitarian Product

Brand Promotion 55 53

Brand Advertising 60 36

5.1.1 Validity and Reliability

An exploratory factor analysis and a reliability analysis have been conducted to check the validity and reliability of variables. First, a principal component factoring analysis (PCF) was used to check the discriminant and convergent validity. Eigenvalue is larger than 1. The Kaiser-Meyer-Olkin measure verified the sampling adequacy for the analysis, KMO: 0.879. Bartlett’s test of sphericity χ2

(120) = 1910,099, p < .001, indicated that correlations between items were sufficiently large for PCF. An initial analysis was run to obtain eigenvalues for each component in the data. Four components had eigenvalues over Kaiser’s criterion of 1 and in combination explained 70.727% of the variance. In agreement with Kaiser’s criterion,

(33)

33 examination of the scree plot revealed a levelling off after the fifth factor. Thus, four factors were retained and rotated with a Promax with Kaiser Normalization rotation. Table 4 shows the factor loadings after rotation. The items that cluster on the same factors suggests that factor 1 represents Attitude toward the Ad, factor 2 represents Attitude toward the Product, factor 3 represents Attitude toward LBA, and factor 4 represents Purchase Attitude. As the results suggest, the third item of LBAAtt shows high cross-loadings on the factor ProdAtt as well, this could be due to the content of the item. For that reason this item is excluded from further analysis.

Table 4: Explanatory Factor Analysis

Component

Item Factor 1 Factor 2 Factor 3 Factor 4

Purchase Attitude1 ,926

Attitude toward the Ad1 ,844

Attitude toward theProduct1 ,841

Attitude toward LBA1 ,876

Attitude toward LBA4 ,481 ,539

Eigenvalues 6,87 2,06 1,28 1,11

% of variance 42,94 12,88 7,97 6,94

Note: Factor loadings over .40 appear in bold. Recoded items are displayed with R.

Second, a reliability test is necessary for multi-item scales analysis in order to achieve a meaningful result. To verify if all the items in one scale measure the same, or if some questions should not be used for analysis, the Cronbach’s Alphas of the variables were tested. In this research, a Cronbach’s Alpha of > 0.7 was considered acceptable. Results showed that

(34)

34 Cronbach’s Alphas of all the variables were above the limit. The corrected item-total

correlations indicate that all the items have a good correlation with the total score of the scale (all above .30). Also, none of the items would substantially affect reliability if they were deleted. Table 5 shows the final reliability scores of all scales.

Table 5: Reliability of Scales

Construct N of items Cronbach's Alpha*

Purchase Attitude 3 0,856

Attitude towards the Ad 4 0,909 Attitude towards the Product 4 0,858

Attitude towards LBA 4 0,777

* Cronbach's Alpha should > 0.7

5.1.2 Correlations

Before the correlation test and further analysis, Message Content and Product Type were transformed into dummy variables by following the guidelines of Field (2013). Pearson’s correlation coefficients were calculated to check the relations between variables. The threshold is 0.5. Table 6 shows that all correlations are below 0.5. Pearson correlation revealed that independent variables were not significantly correlated with each other. Age has a significant moderate negative relationship with Gender (r = -.237, p < 0.05) and Education (r = -0.295, p < 0.05).

Table 6: Correlation Matrix

Variable Mean SD 1 2 3 4 5 6 7

1 Gender 1,70 0,46 1

2 Age 29,05 11,73 -,237** 1

3 Education 3,37 ,68 ,68 ,295** 1

4 Prior Experience with LBA

2,07 1,43 ,017 ,075 -,016 1

5 Purchase Impulsiveness 4,09 1,17 ,127 -,175* -,006 ,11 1

6 Message Content 1,53 ,50 ,082 -,062 -.062 -,101 ,02 1

7 Product Type 1,44 ,50 ,044 -,030 -,069 -,008 ,004 ,116 1 **. Correlation is significant at the 0.01 level (2-tailed).

(35)

35 5.2 Hypotheses Testing

5.2.1 The Effect of Message Content on LBA’s Effectiveness

To test the influence of message content on purchase attitude, attitude toward the ad, attitude toward the product, and attitude toward LBA, a one-way ANOVA was run separately for each dependent variable. The test results of hypothesis 1 are summarized in table 7 and 8.

Hypothesis 1a. A one-way ANCOVA was performed to investigate whether purchase

attitude is greater for promotional LBA than for advertising LBA. In the final model Age (F(1, 204) = 40.63, p < 0.01) and Education (F(1, 204) = 5.50, p = 0.02) were significant covariates. The effect of Message Content on Purchase Attitude was marginally significant,

F(1, 204) = 2.96, p = 0.09. Purchase attitude results to be greater among promotional LBA (M

= 4.39, SD = 1.39) comparing to advertising LBA (M = 4.00, SD = 1.33). Thus, hypothesis 1a is somewhat supported.

Hypothesis 1b. A one-way ANCOVA was performed to investigate whether attitude

toward the ad is greater for promotional LBA than for advertising LBA. In the final model Age (F(1, 204) = 12.34, p < 0.01) was a significant covariate. There was a statistically significant effect of Message Content on Attitude toward the Ad, F(1, 204) = 4.87, p = 0.03. Attitude toward the Ad results to be greater among promotional LBA (M = 3.91, SD = 1.43) comparing to advertising LBA (M = 3.43, SD = 1.42). Thus, hypothesis 1b is supported.

Hypothesis 1c. A one-way ANCOVA was performed to investigate whether attitude

toward the product is greater for promotional LBA than for advertising LBA. In the final model Age (F(1, 204) = 10.88, p < 0.01) and Prior Experience with LBA (F(1, 204) = 5.61, p = 0.02) were significant covariates. The effect of Message Content on Attitude toward the Product was marginally significant, F(1, 204) = 3.79, p = 0.05. Attitude toward the Product results to be greater among promotional LBA (M = 4.28, SD = 1.25) comparing to advertising LBA (M = 3.94, SD = 1.21). Thus, hypothesis 1c is somewhat supported.

(36)

36

Hypothesis 1d. A one-way ANCOVA was performed to investigate whether attitude

toward the product is greater for promotional LBA than for advertising LBA. The effect of Message Content on Attitude toward LBA was statistically not significant, F(1, 204) = 0.69, p = 0.41. Thus, hypothesis 1d is not supported.

Table 7: ANCOVA H1

Source df F Sig.

Purchase Attitude Age 1 40,631 ,000

Education 1 5,504 ,020

Message Content 1 2,961 ,087

Attitude toward the Ad Age 1 12,339 ,001

Attitude toward the Product Age 1 10,874 ,001

Purchase Impulsiveness 1 2,885 ,091 Prior Experience with LBA 1 5,612 ,019

Attitude toward LBA Message Content 1 ,693 ,406

Table 8: Descriptive Statistics of ANCOVA H1

Mean SD N

Purchase Attitude Advertising 4,00 1,33 96

Promotional 4,39 1,39 108

Attitude toward the Ad Advertising 3,43 0,14 96

Promotional 3,91 1,43 108

Attitude toward the Product Advertising 3,94 1,21 96

Promotional 4,28 1,25 108

Attitude toward LBA Advertising 3,73 1,20 96

Promotional 3,59 1,15 108

5.2.2 The Moderating Effect of Product Type on LBA’s Effectiveness

To test whether the effect of Message Content on the Effectiveness of LVA is

moderated by Product Type, a factorial ANCOVA was conducted four times. Table 9 shows the test results.

(37)

37

Hypothesis 2a. It was hypothesized that the effect of Message Content on Purchase

Attitude is moderated by Product Type. In the final model Age (F(1, 204) = 40.10, p < 0.01) and Education (F(1, 204) = 6.63, p = 0.01) were significant covariates. The main effect of Message Content was marginally significant (F(1, 204) = 2.96, p = 0.09). The main effect of Product Type (F(1, 204) = 1.09, p = 0.30) and the interaction effect between Message Content and Product Type (F(1, 204) = 1.38, p = 0.24) were not significant. Hypothesis 2a is therefore not supported.

Hypothesis 2b. To test whether the effect of Message Content on Attitude toward the

Ad is moderated by Product Type a Factorial ANOVA was conducted. In the final model Age (F(1, 204) = 12.714, p < 0.01) was a significant covariate. The main effect of Message

Content was significant (F(1, 204) = 5.165, p = 0.02). The main effect of Product Type (F(1, 204) = 0.66, p = 0.42) and the interaction effect between Message Content and Product Type (F(1, 204) = 2.11, p = 0.15) were not significant. Hypothesis 2b is therefore not supported.

Hypothesis 2c. It was hypothesized that the effect of Message Content on Attitude

toward the Product is moderated by Product Type. In the final model Age (F(1, 204) = 10.61,

p < 0.01) and Prior Experience with LBA (F(1, 204) = 5.28, p = 0.02) were significant

covariates. The main effect of Message Content (F(1, 204) = 3.45, p = 0.07) and Product Type (F(1, 204) = 3.66, p = 0.06) were both marginally significant. The interaction effect between Message Content and Product Type (F(1, 204) = 1.57, p = 0.21) was not significant. Hypothesis 2c is therefore not supported.

Hypothesis 2d. To test whether the effect of Message Content on Attitude toward

LBA is moderated by Product Type a Factorial ANOVA was conducted. In the final model, the main effect of Message Content (F(1, 204) = 0.56, p = 0.54), Product Type (F(1, 204) = 0.00, p = 0.96), and their interaction effect (F(1, 204) = 1.03, p = 0.31) were not significant. Hypothesis 2d is therefore not supported.

(38)

38

Table 9: ANCOVA H2

Source df F Sig.

Purchase Attitude Age 1 40,100 ,000

Product Type 1 1,092 ,297

Message Content * Product Type 1 1,375 ,242

Attitude toward the Ad Age 1 12,714 ,000

Product Type 1 ,657 ,419

Attitude toward the Product Age 1 10,610 ,001

Purchase Impulsiveness 1 2,429 ,121 Prior Experience with LBA 1 5,281 ,023

ProdTyp 1 3,657 ,057

Attitude toward LBA Age 1 1,137 ,288

Message Content 1 ,563 ,454

Product Type 1 ,002 ,961

Message Content * Product Type 1 1,032 ,311 5.3 Summary of results

Table 10 concluded the test results of proposed hypotheses. Findings of data analysis will be carefully discussed in next chapter.

Table 10: Results of Hypotheses Testing

Hypotheses Test Result

H1a: Purchase attitude would be greater for promotional LBA than for advertising LBA.

Somewhat supported H1b: Attitude toward the ad will be greater for promotional LBA

than for advertising LBA.

Supported H1c: Attitude toward the product will be greater for promotional

LBA than for advertising LBA.

Somewhat supported H1d: Attitude toward LBA will be greater for advertising LBA than

for promotional LBA.

Not supported H2a: The effect of message content on purchase attitude is greater

for hedonic products than that for utilitarian products.

(39)

39 H2b: The effect of message content on attitude toward the ad is

greater for hedonic products than that for utilitarian products.

Not supported H2c: The effect of message content on attitude toward the product

is greater for hedonic products than that for utilitarian products.

Not supported H2d: The effect of message content on attitude toward LBA is

greater for hedonic products than that for utilitarian products.

(40)

40 6. Discussion and Conclusion

The purpose of this research was to study how message content and product type influence the effectiveness of location-based advertising. It was hypothesized that message content would affect LBA’s effectiveness. This expectation was partially supported by the results, as purchase attitude, attitude toward the ad, and attitude toward the product were greater in the brand promotion condition as compared to the brand advertising condition. In addition, this study expected that message content would result in a greater attitude, but only for promoting ads. It was assumed that an ad of an hedonic product maps with the promotion-focused attributes of promotional LBA, providing a strong content-category fit, which will lead to a greater consumer’ attitude. This expectation was not borne out, as there were no effects of product type on the relationship between message content and attitude.

First, the results confirm that LBA is, as expected, an effective marketing tool. The increase of attitude toward the product, attitude toward the ad, and attitude toward purchase is noticeable. This is an important fact considering there was until now little known about the effect of LBA. Moreover, one can conclude that the average values of the attitude towards LBA are fairly neutral. People are not very enthusiastic about LBA yet, and there is therefore a process for marketers to convince consumers of the relevance of LBA. A positive side effect is that prior experience with LBA did only once significantly affect LBA's effectiveness. This suggest that someone who just starts receiving LBA can be directly influenced in a positive way.

The finding that message content focusing on brand promotion was related to a more positive attitude may be due to the strong link between the brand promotion and the

orientation of promotion focused consumers, which are more likely to adopt LBA. These findings are in line with research from Unni and Harmon (2007) and Luo et al. (2013) who report that LBA works best when the ad contains a concrete cue on which consumers can act

(41)

41 upon, as is the case with brand promotion. However, their study did not reveal significant differences for promotional and advertising LBA in perceived benefits, perceived value, and intentions to use LBA. It is interesting that there is a difference found in the current study. This may be explained by the fact that attitude is a more direct way of measuring intent on buying compared to perceived benefits, perceived value, and intentions to use LBA. It suggests that a direct way of measuring the intention to buy might be better to have an understanding of LBA's effectiveness. For example, promotional LBA may not affect the perceived value, but the intention to purchase may be influenced due to exposure to the promotional ad. Perhaps brand promotion leads to a higher purchase intention and brand advertising leads to a higher perceived value, but the latter may be observed in the long term.

A plausible explanation of why message content was unrelated to attitude towards LBA is that consumers already have a positive attitude towards LBA if they have given permission to receive LBA. The results suggest that the location-based ads meet the expectations of consumers, since their general attitude toward LBA is not affected. On the other hand, purchase attitude, attitude toward the ad, and attitude toward the product can be linked more specifically to one ad. Therefore, one can explain that the content of the ad has a direct influence on the aforementioned.

It further appears from the results that age plays an important role regarding attitude. It is striking that a higher age is coming with a more positive attitude towards LBA, while you may expect younger people to be more open to this relatively new marketing medium. Within the sample was a wide range of age, which makes it representative for the general population.

(42)

42 The finding that product type does not affect the relationship between message content and LBA’s effectiveness might mean that LBA is widely applicable to various products and the type of product is therefore less relevant. Apparently, there is no relationship between product category characteristics and a LBA shoppers’ focus. There was no support for the regulation focus theory on this aspect. This suggests that both hedonic and utilitarian products are suitable to be advertised with LBA on mobile devices. Another explanation for the results can be that other product-related aspects such as costs or tangibility might have a stronger influence on ad effectiveness. Likewise, Bart et al. (2014) highlight the level of involvement as an important influencer of mobile advertising effectiveness.

6.1 Theoretical and Practical Implications

The findings of this research have important implications for both theory and practice. From a theoretical standpoint, the results showing that fit between brand promotion content and the orientation of promotion focused consumers give further empirical support to the regulatory focus theory. Most importantly, the findings advance knowledge and

understanding of the effectiveness of LBA. This study is an extension of the knowledge about the influence of message content on LBA's effectiveness reported by Unni and Harmon (2007). Where in their studies no differences have been found between advertising and promotional LBA in several aspects, the current research shows that the type of message certainly can have a direct impact on LBA’s effectiveness. This allows LBA research to take the next step towards real consumer behavior. One can now no longer ignore the effect of message content and the size of interest must be determined by research with real purchase data.

From a practical standpoint, the present study’s findings have two important managerial implications. First, this study shows that LBA can be effective in influencing

(43)

43 consumers’ attitudes. LBA represents a fundamental element of successful mobile

advertising. The current research supports the arguments of marketers to deploy this tool and to allocate more budgets to this platform.

Second, this study shows that it is important for marketers to focus on the promotion of LBA itself. Consumers appear to have a relatively neutral attitude toward LBA. LBA in a new and innovative medium that can provide much more personalized ads, but consumers are still not completely convinced. It is up to the marketers to ensure that consumers are aware of the importance and personal relevance of LBA, in order to further increase LBA's

effectiveness.

Third, it appears that the effectiveness of LBA is not influenced by product type. This gives marketers a license to apply LBA for a wide range of product options, since it turned out to be an effective marketing tool.

6.2 Limitations and Future Research

Although the present study gave rise to important theoretical and managerial

implications, a number of limitations that indicate opportunities for future research have to be noted. More research is needed into the product dimension of LBA (Bauer & Strauss, 2016; Bruner & Kumar, 2007; Unni & Harmon, 2007; Xu et al., 2011). This study was limited to one general product characteristic (hedonic versus utilitarian). Future research could attempt to further disentangle moderating effects of factors as tangibility, involvement, costs, and frequency of buying on the relationship between message content and the effectiveness of LBA.

Another imitation is that this study is based on location-based notifications for which permission is given by the consumer. LBA, however, can come in many different formats. Location supplies marketers with a tremendous diversity of technologies, approaches, and strategies to communicate better by more targeted and personalized information toward the

(44)

44 consumer. Therefore, it is hard to make general statements about LBA’s effectiveness by inferring from this specific case. Future research into the effectiveness of other types of LBA will open the door to new insights for advertising agencies, marketers, and brands.

Third, a virtual reality setting would have allowed for a more real-life shopping experience and therefore increase the ecological and external validity. This would also allow the assessment of purchase behavior to examine LBA’s effectiveness. Although purchase attitude it is extensively used and considered as a validated precedent of advertising’s effectiveness (Bart et al., 2014; van't Riet et al., 2016), a positive attitude toward purchasing need not to result in actual purchasing. This limitation applies to most experimental study designs (Bryman & Bell 2007).

Moreover, the results of Jedidi et al. (1999) suggest that price promotion may have a negative effect on brand equity in the run. It is therefore warranted to examine the long-term effects of LBA (in its various manifestations) on brand equity. This can further expand the field’s understanding of the effectiveness of advertising.

Finally, the use of a convenience sample is a limitation of this study. Many students for the University of Amsterdam have participated in this study. As a consequence, the amount of well-educated people (Bachelor and above) in the sample was way above the Dutch average with 90.8% compared to 27.59% (CBS, 2017). Generalizability of the results to the general population is likely to be affected. Future research should examine LBA’s effectiveness with a more diverse sample.

Strong points of this study were valid and reliable scales, a good sample size, and a wide range of ages within the sample. The experimental setting allowed investigating the direction of the effects. In addition, the current study focused on different product types, where previous research had raised many questions about this aspect. Several types of attitudes were included in this research. This study was a first step in showing that message